Carbon stocks and greenhouse gas emissions in ...

Brazilian Agricultural Research Corporation (Embrapa) Embrapa Environment Ministry of Agriculture, Livestock and Food Supply

Carbon stocks and greenhouse gas emissions in Brazilian agriculture Magda A. Lima Robert M. Boddey Bruno J. R. Alves Pedro L. O. de A. Machado Segundo Urquiaga Technical Editors Embrapa Brasília, DF 2012

Copies of this publication may be obtained from: Embrapa Environment Rodovia SP-340, km 127,5 Caixa Postal 69 CEP 13820-000 Jaguariúna, SP Phone: +55 (19) 3311-2700 Fax: +55 (19) 3311-2640 www.cnpma.embrapa.br [email protected] Unit responsible for content Embrapa Environment Comitê de Publicações da Embrapa Meio Ambiente President Marcelo Augusto Boechat Morandi Members Adriana M. M. Pires Fagoni Fayer Calegario Lauro Charlet Pereira Aline de Holanda Nunes Maia Embrapa Technological Information Parque Estação Biológica (PqEB) Av. W3 Norte (Final) CEP 70770-901 Brasília, DF Phone: +55 (61) 3448-4236 Fax: +55 (61) 3448-2494 www.embrapa.br/liv [email protected] Unit responsible for publication Embrapa Technological Information Editorial coordination Fernando do Amaral Pereira Lucilene Maria de Andrade Juliana Meireles Fortaleza Editorial supervision Wesley José da Rocha Proofreading Maria Cristina Ramos Jubé Bibliographic standardization Iara Del Fiaco Rocha Translation TakeFive Translations & Training – Rio de Janeiro José Henrique Oliveira Moreira Donald Scott Alexander Graphic design, desktop publishing and cover Carlos Eduardo Felice Barbeiro 1st edition 1st printing (2012): 1,500 copies

All rights reserved. Unauthorized reproduction of this publication, or any part of it, constitutes a copyright infringement (Law 9,610/98). International Cataloging in Publication (CIP) Data. Embrapa Technological Information Carbon stocks and greenhouse gas emissions in Brazilian agriculture / Magda A. Lima ... [et al.], technical editors. – Brasília, DF: Embrapa, 2012. ___ pp.: ill. color. ; 16 cm x 22 cm. ISBN 978-85-7035-053-4 1. Environmental impact. 2. Environmental protection. 3. Soil. I. Lima, Magda A. II. Boddey, Robert M. III. Alves, Bruno J. R. IV. Machado, Pedro L. O. de A. V. Urquiaga, Segundo. VI. Embrapa Meio Ambiente. CDD 363.7 © Embrapa 2012

Authors Alexandre Berndt Biologist, Ph.D. in Applied Ecology, researcher at Embrapa Cattle – Southeast, São Carlos, São Paulo (SP) [email protected] Arminda Moreira Carvalho Agronomist, Ph.D. in Ecology, researcher at Embrapa Cerrados, Planaltina, Distrito Federal (DF) [email protected] Augusto Cesar Franco Ecologist, post-doctorate in Plant Ecophysiology, professor at Universidade de Brasília (UNB), Brasília, Distrito Federal (DF) [email protected] Beata Emoke Madari Agronomist, Ph.D. in Soil Science and Plant Nutrition, researcher at Embrapa Rice and Beans, Santo Antônio de Goiás, Goiás (GO) [email protected] Bruno José Rodrigues Alves Agronomist, post-doctorate in Soil Science, researcher at Embrapa Agrobiology, Seropédica, Rio de Janeiro (RJ) [email protected] Cimélio Bayer Agronomist, Ph.D. in Soil Science, associate professor at Universidade Federal do Rio Grande do Sul (UFRS), Porto Alegre, Rio Grande do Sul (RS) [email protected] Cintia Rodrigues de Souza Biologist, Master of Forest Science, researcher at Embrapa Western Amazon, Manaus, Amazonas (AM) [email protected] Cláudia Pozzi Jantalia Agronomist, Ph.D. in Plant Science, researcher at Embrapa Agrobiology, Seropédica, Rio de Janeiro (RJ) [email protected]

Claudio José Reis de Carvalho Agronomist, Ph.D. in Plant Ecophysiology, researcher at Embrapa Eastern Amazon, Belém, Pará (PA) [email protected] Débora Marcondes Bastos Pereira Milori Physicist, Ph.D. in Physics, researcher at Embrapa Agricultural Instrumentation, São Carlos, São Paulo (SP) [email protected] Elaine Cristina Cardoso Fidalgo Agronomist, Ph.D. in Agricultural Engineering, researcher at Embrapa Soils, Rio de Janeiro, Rio de Janeiro (RJ) [email protected] Elio Marcolin Agronomist, Master in Agricultural Engineering, researcher at Instituto Rio Grandense do Arroz (Irga), Cachoeirinha, Rio Grande do Sul (RS) [email protected] Eloisa Aparecida Belleza Ferreira Agronomist, Master in Agronomy, researcher at Embrapa Cerrados, Planaltina, Distrito Federal (DF) [email protected] Falberni de Souza Costa Agronomist, Ph.D. in Soil Science, researcher at Embrapa Acre, Rio Branco, Acre (AC) [email protected] Francisco das Chagas Leonidas Agronomist, Master in Agronomy, researcher at Embrapa Rondônia, Porto Velho, Rondônia (RO) [email protected] Haron Abrahim Magalhães Xaud Agronomist, Master in Remote Sensing, researcher at Embrapa Roraima, Boa Vista, Roraima (RR) [email protected] Helber Custódio de Freitas Meteorologist, Ph.D. in Applied Ecology, researcher at Universidade de São Paulo (USP), São Paulo, São Paulo (SP) [email protected] Helio Tonini

Forest Engineer, Ph.D. in Forest Engineering, researcher at Embrapa Roraima, Boa Vista, Roraima (RR) [email protected] Henrique Pereira dos Santos Agronomist, Ph.D. in Plant Science, researcher at Embrapa Wheat, Passo Fundo, Rio Grande do Sul (RS) [email protected] Humberto Ribeiro da Rocha Aeronautical Civil Engineer, Ph.D. in Meteorology, lecturer and professor at Universidade de São Paulo (USP), São Paulo, São Paulo (SP) [email protected] João Baptista Silva Ferraz Biologist, Ph.D. in Forest Sciences, researcher at Instituto Nacional de Pesquisas da Amazônia (Inpa), Manaus, Amazonas (AM) [email protected] João José Assumpção de Abreu Demarchi Agronomist, Ph.D. in Biological Sciences, researcher at Agência Paulista de Tecnologia dos Agronegócios, Nova Odessa, São Paulo (SP) [email protected] Jônatan Dupont Tatsch Meteorologist, Ph.D. in Meteorology, São Paulo, São Paulo (SP) [email protected] Julio Cezar Franchini dos Santos Agronomist, post-doctorate in Organic Chemistry, researcher at Embrapa Soybean, Londrina, Paraná (PR) [email protected] Ladislau Martin-Neto Physicist, Ph.D. in Applied Physics, researcher at Embrapa Agricultural Instrumentation, São Carlos, São Paulo (SP) [email protected] Luciano José de Oliveira Accioly Agronomist, Ph.D. in Soil, Water and Environmental Science, researcher at Embrapa Soils, Rio de Janeiro, Rio de Janeiro (RJ) [email protected] Luís Henrique de Barros Soares

Agronomist, Ph.D. in Molecular and Cellular Biology, researcher at Embrapa Agrobiology, Seropédica, Rio de Janeiro (RJ) [email protected] Luiz Fernando Carvalho Leite Agronomist, post-doctorate in Soils and Plant Nutrition, researcher at Embrapa Mid-North, Teresina, Piauí (PI) [email protected] Magda Aparecida de Lima Ecologist, Ph.D. in Geosciences and Environment, researcher at Embrapa Environment, Jaguariúna, São Paulo (SP) [email protected] Márcio dos Santos Pedreira Zootechnician, Ph.D. in Animal Production and Nutrition, teacher at Universidade Estadual do Sudoeste da Bahia, Itapetinga, Bahia (BA) [email protected] Marco Bindi Agronomist, Ph.D. in Forage Crop Modeling, teacher at Universitá degli Studi di Firenze, Centralino, Italy [email protected] Marco Moriondo Agronomist, Ph.D. in Forage Crop Modeling, teacher at Universitá degli Studi di Firenze, Centralino, Italy [email protected] Marcos Antonio Vieira Ligo Ecologist, Ph.D. in Applied Ecology, researcher at Embrapa Environment, Jaguariúna, São Paulo (SP) [email protected] Maria Conceição Peres Young Pessoa Bachelor of Applied Mathematics, Ph.D. in Automation, researcher at Embrapa Environment, Jaguariúna, São Paulo (SP) [email protected] Maria Lucia Meirelles Biologist, Ph.D. in Biology, researcher at Embrapa Cerrados, Planaltina, Distrito Federal (DF) [email protected] Maristela Ramalho Xaud Agronomist, Master in Remote Sensing, researcher at Embrapa Roraima, Boa Vista, Roraima (RR)

[email protected] Mauricio Rizzato Coelho Agronomist, Ph.D. in Soils and Plant Nutrition, researcher at Embrapa Soils, Rio de Janeiro, Rio de Janeiro (RJ) [email protected] Moisés Cordeiro Mourão de Oliveira Júnior Biologist, Master in Agronomical Statistics and Experimentation, researcher at Embrapa Eastern Amazon, Belém, Pará (PA) [email protected] Odo Maria Artur Siegmund Pedro Rudolfo Barão Primavesi Agronomist, Ph.D. in Soils and Plant Nutrition, retired researcher at Embrapa Cattle – Southeast, São Carlos, São Paulo (SP) [email protected] Omar Vieira Villela Agronomist, Master in Plant Science, researcher at Agência Paulista de Desenvolvimento do Agronegócio (Apta), Pindamonhangaba, São Paulo (SP) [email protected] Osvaldo Machado Rodrigues Cabral Meteorologist, Ph.D. in Meteorology, researcher at Embrapa Environment, Jaguariúna, São Paulo (SP) [email protected] Paulo Guilherme Salvador Wadt Agronomist, Ph.D. in Soils and Plant Nutrition, researcher at Embrapa Acre, Rio Branco, Acre (AC) [email protected] Pedro Luiz Oliveira de Almeida Machado Agronomist, Ph.D. in Soils and Plant Nutrition, researcher at Embrapa Rice and Beans, Santo Antônio de Goiás, Goiás (GO) [email protected] Robert Michael Boddey Agricultural Chemist, Ph.D. in Agriculture, researcher at Embrapa Agrobiology, Seropédica, Rio de Janeiro (RJ) [email protected] Roberval Monteiro Bezerra de Lima

Forest Engineer, Ph.D. in Forest Engineering, researcher at Embrapa Western Amazon, Manaus, Amazonas (AM) [email protected] Rogério Sebastião Corrêa da Costa Agronomist, Ph.D. in Biotechnology, researcher at Embrapa Rondônia, Porto Velho, Rondônia (RO) [email protected] Rosa Toyoko Shiraishi Frighetto Chemist, Ph.D. in Organic Chemistry / Natural Products, researcher at Embrapa Environment, Jaguariúna, São Paulo (SP) [email protected] Rosana Clara Victoria Higa Agronomist, Ph.D. in Forest Engineering, researcher at Embrapa Forestry, Colombo, Paraná (PR) [email protected] Segundo Sacramento Urquiaga Caballero Agronomist, Ph.D. in Soils and Plant Nutrition, researcher at Embrapa Agrobiology, Seropédica, Rio de Janeiro (RJ) [email protected] Steel Silva Vasconcelos Agronomist, Ph.D. in Forest Resources and Conservation, researcher at Embrapa Eastern Amazon, Belém, Pará (PA) [email protected] Vanda Gorete Souza Rodrigues Agronomist, Master in Tropical Agriculture, researcher at Embrapa Rondônia, Porto Velho, Rondônia (RO) [email protected] Vera Regina Mussoi Macedo Agronomist, Master in Soil Sciences, retired researcher at Instituto Rio Grandense do Arroz (Irga), Cachoeirinha, Rio Grande do Sul (RS) [email protected] Vinicius de Melo Benites Agronomist, Ph.D. in Soils and Plant Nutrition, researcher at Embrapa Soils, Rio de Janeiro, Rio de Janeiro (RJ) [email protected]

Presentation The increase of the world’s food production has resulted in the replacement of native vegetation with annual and permanent crops, and also with plant species used for animal feed. Nowadays, the demand for renewable energy has motivated the expansion of plantations for energy and biofuel production. This scenario of changes in soil use has been emblematic for Brazil, a country that has become one of the main leaders in the fight against world hunger due to its extensive cultivable areas and its consolidated agricultural economy. Until recently, very little was known about the impact of agriculture and livestock production on Brazilian greenhouse gas (GHG) emissions into the atmosphere, specifically CO2, CH4 and N2O. Actions carried out by Embrapa and most research institutions in the country regarding this issue were limited, and little information was being generated in Brazil about GHG emission factors for tropical regions, which is of great importance to develop inventories and draft public policies aiming to mitigate emissions. As a result, initial estimates of GHG emissions for Brazilian agriculture were calculated based on emission factors obtained from studies carried out on temperate climate agriculture, which entails many uncertainties. For this reason, in 2003, Embrapa created a multiinstitutional and multidisciplinary research group tasked with quantifying GHG fluxes in the various farming, livestock, forestry and agroforestry systems in the country, aiming not only to obtain a characterization, but also to identify management systems that contribute to mitigate the greenhouse effect. This publication presents the contributions of Embrapa and its partner institutions for the advancement of knowledge about GHG emissions in the Brazilian agricultural sector, the result of an integrated work effort that constituted the Agrogases research network – Carbon Dynamics and Greenhouse Gas Emissions in Brazilian Agricultural, Agroforestry and Forestry Systems. Among the results, the most prominent is the generation of specific values for CH4 and N2O emission factors for Brazilian conditions, which were then compared with the figures presented by the Intergovernmental Panel on Climate Change (IPCC). Methane emission factors were determined in irrigated rice production systems, with possible mitigation associated with the planting system. Other aspects worth noting were the positive nitrogen balance in agricultural systems, especially in no-tillage planting, and the rehabilitation of degraded pastures to promote sequestration of C in soils. In beef and dairy cattle production, emission factors for enteric methane were determined for various diet conditions, demonstrating that it is possible to significantly reduce the production of this gas per unit of meat produced by improving the quality of the forage. The introduction of forage legumes with an efficient system of obtaining N2 from the air through biological nitrogen fixation is a good strategy to improve not only the productivity of pastures, but also their quality. This strategy, combined with the adoption of the crop-livestock integration system, offers the best prospects for boosting grain and livestock production in a sustainable way, significantly contributing to mitigation of GHG emissions. The implementation of forest inventories, using remote sensing techniques and stoichiometric models, resulted in the assessment of the potential for carbon sequestration in various Brazilian forest systems (forests planted in the South and North regions of the country, Caatinga [arid shrubland] in the Northeast region) and agroforestry ecosystems. In this context, it is worth mentioning the major positive impact caused by reforestation with economically important

species, improving rural incomes and significantly contributing to carbon sequestration in the soilplant system. The use of robust simulation models for estimating carbon balance and greenhouse gas emission in sugarcane and grain production systems proved to be an essential tool in this topic, especially due to the need to adjust them to tropical conditions. With these results, the Agrogases network is meeting society’s needs for knowledge consolidated by agricultural research regarding the carbon sequestration potential of Brazilian soils, in forest plantations and in native vegetation, as well as regarding the greenhouse gas emission potential of the animal and plant production systems used in Brazil. Furthermore, the results indicate measures to mitigate GHG emissions, through optimized strategies and practices in land use and production systems.

– the Editors

Table of Contents Chapter 1 Carbon stock based on Brazilian soil survey: A contribution to the national inventory ................................................................... Chapter 2 Carbon stocks in Brazilian soils: quantity and mechanisms for accumulation and preservation ......................... Chapter 3 Carbon dynamics a humid area of the Cerrado ....................................................... Chapter 4 Biomass Stock in planted forests, agroforestry systems, secondary forests and Caatinga ......................................... Chapter 5 Emissions of nitrous oxide and nitric oxide from soils in agricultural systems .................................................................................. Chapter 6 Methane emissions in flooded rice cultivation ........................................................ Chapter 7 Implementation of a generic model for sugarcane crops in the Southeastern region of Brazil ........................................... Chapter 8 Greenhouse gas production in agricultural systems: Groundwork for an inventory of methane emission by ruminants .................. Chapter 9 Computer simulations for the study of carbon and nitrogen dynamics and greenhouse gas emissions in agricultural production systems ................ Chapter 10 Practices to mitigate greenhouse gas emissions in Brazilian agriculture ..................................................................................

Chapter 1

Carbon stock based on Brazilian soil survey: A contribution to the national inventory Elaine C. C. Fidalgo, Vinicius de M. Benites, Paulo G. S. Wadt, Mauricio R. Coelho, Beata E. Madari, Pedro L. O. de A. Machado

Abstract: Soils represent an important component in the biogeochemical cycle of carbon (C), storing about four times more C than plant biomass and almost three times more than the atmosphere. Concerns about global climate changes and the contribution of carbon accumulation for their mitigation have created a need for national estimates of carbon stocks in soils, to which Brazil has responded by using available databases and applying various methods. In order to contribute information on estimates of C in soils, the present study was conducted using data from a soil database called SigSolos, organized by Embrapa Soils to estimate soil C stocks in Brazil. The estimate of carbon in Brazilian soils was carried out considering the various types of soil under various land use systems and distributed through the various Brazilian biomes. The results obtained, although comparable to those obtained in other surveys, demonstrate the existence of gaps in the data and the need to build a soil database for all of Brazil. Keywords: soil carbon, soil density, pedotransfer function, soil database.

Introduction Concerns about global climate changes and the contribution of carbon accumulation for their mitigation have created a need for national estimates on carbon stocks in soils (ESWARAN et al., 1993). Soils represent an important component in the biogeochemical cycle of C, storing about four times more C than plant biomass and almost three times more than the atmosphere (IPCC, 2001). Brazil is one of the few countries known to produce estimates on soil carbon stocks through the use of various types of calculations and mappings (BATJES, 2005; BERNOUX et al., 2002; SCHROEDER; WINJUM, 1995). In order to contribute information on estimates of C in Brazilian soils, the present study was conducted as part of the Agrogases Project, aiming to estimate carbon stocks in Brazilian soils, using data from a soil database called SigSolos, organized by Embrapa Soils.

Estimate of soil density Soil density (SD) is essential for calculations of carbon stocks. SD values are necessary to convert the carbon content obtained as a percentage of the dry weight into carbon weight per unit area (HOWARD et al., 1995). The carbon stock (Ct) is then calculated as follows: Ct (t ha−1) = SD (g cm−3) * soil carbon concentration (%) * soil sampling depth (cm) Recent studies have indicated the importance of estimating soil carbon stocks when comparing various types of land use, considering the same soil mass (SISTI et al., 2004; SMITH et al.,

1998). There is, however, a serious shortage of SD data in many studies involving data on organic carbon, especially in reports of soil surveys carried out in Brazil. In Brazil, Bernoux et al. (1998) and Tomasella and Hodnett (1998) provided the first basis to predict SD based on soil attributes in the Amazon Basin. Bernoux et al. (1998), using a stepwise forward multiple regression procedure with progressive inclusion of variables, demonstrated that clay content, organic carbon and pH are the best predictors of SD, and Tomasella and Hodnett (1998) adjusted multiple linear regressions to estimate SD from sand, silt and clay content. Subsequently, results obtained by Bernoux et al. (1998) were used to estimate SD and carbon stock (carbon mass per unit area) for similar soils in all of Brazil (Bernoux et al., 2002). Pedotransfer functions (PTF) based on easily obtainable soil attributes, such as organic carbon or clay content, have strong potential to replace SD measurements when these cannot be obtained. However, as reported by De Vos et al. (2005), published PTFs show large differences in performance when applied in environments other than those in which they were adjusted. Based on this fact, within the Agrogases network, Benites et al. (2007) developed a statistical procedure to predict the SD of the Brazilian soils found in most biomes from easily obtainable soil attributes. The available SD values and those estimated by Benites et al. (2007) were used to estimate the carbon stock in Brazilian soils. The data used in this study was taken from a soil database called SigSolos, organized by Embrapa Soils (CHAGAS et al., 2004), which contains soil information for the period of 1958 to 2001. Procedures for quality control of data in this database were developed by establishing threshold values for each soil attribute and adjusting measurement units to the international system. The thresholds used made it possible to eliminate values outside the established ranges based on correlations between soil attributes, such as C versus N, pH versus Ca+Mg, and clay versus sand. This procedure also made it possible to identify and correct typographical errors in the original survey reports. To estimate the density of Brazilian soils, a set of soil profile data was organized from initial data on SigSolos, including only samples in which the soil density had been measured. This data comprised 363 profiles (1,542 horizons) collected in the majority of Brazilian biomes between 1975 and 2004, which had been published in various soil survey bulletins. The soil profile distribution for the various biomes is illustrated in Figure 1. It is important to point out that many of the profiles described in the database did not include their geographical coordinates. In those cases, for purposes of representation on the map, the coordinates used were those of the seat of the municipality in which the profile was collected.

CHAP 1 - FIGURE 1 Perfil de Solo

Soil Profile

Bioma

Biome

Amazónia

Amazon

Mata Atlântica

Atlantic Forest

Figure 1. Map of Brazil with the location of soil profiles that were used in this study and that contain information on soil density.

Source: Benites et al. (2006).

Descriptive statistics and multiple linear regression analysis (STATSOFT, 1999) were applied to all the corrected data. Multiple regression analysis with progressive inclusion of the variables entered into the model, avoiding the inclusion of collinear variables and variables with no significant predictive effect, was used for the exploratory analysis, relating SD to 17 soil attributes. This procedure allowed for the development of simpler and clearer predictive models, involving easily obtainable soil attributes. To select the best predictive models, an analysis was made on the residuals of the estimates of the various models generated, using a soil data matrix that was independent from the one used to develop the models. The models selected were the ones with the lowest estimation error, i.e., with the smallest difference between the estimated value and the value measured in laboratory. To predict the SD of surface horizons (up to 30 cm deep) and subsurface horizons (below 30 cm), variables in the dataset were taken into account in the multiple regression analysis. For surface horizons, the best SD prediction model was a function of total organic carbon (TOC), clay, sum of bases (SB), Fe2O3, P, Al and TiO2 (adjusted R2 = 0.74; number of observations = 300); whereas for subsurface horizons, the best model was a function of N, clay, SB, Ca+Mg, P, C:N and TOC (adjusted R2 = 0.77; number of observations = 213). The mean SD value for Brazilian soils was 1.36 g cm−3, ranging from 0.13 g cm−3 to 2.25 g cm−3. Regression functions are useful for predicting the SD of Brazilian soils from other attributes. In a first exploratory regression analysis using all the samples in the dataset (number of observations = 1,002), a model was developed, and SD could be predicted based on the following variables: N, clay (total and dispersed in water), SB, C:N, Al2O3 and Ca+Mg. These variables explained 70% of the variation in SD . The adopted model, which is a simplified regression model, uses only TOC, clay and SB, and it described 66% of variation in SD for all soils and all depths. Partitioning the dataset into groups of various depths and soil classes did not result in considerable improvements in SD prediction. When compared to three other existing regression equations to estimate SD, as proposed by Bernoux et al. (1998), Manrique and Jones (1991) and Tomasella and Hodnett (1998), the proposed model showed lower bias, higher precision and higher accuracy.

Estimate of carbon stock in Brazilian soils Schroeder and Winjum (1995) presented the first estimate of carbon stocks in Brazilian soils (71.4 Pg1 in the 0 cm to 100 cm layer) based on the predominant type of ecosystem. Under the United Nations Framework Convention on Climate Change (UNFCCC), Bernoux et al. (2002) developed an inventory of C stocks of soils under native vegetation in Brazil based on methodology suggested by IPCC/UNEP/OECD/OEA (IPCC et al., 1997). The estimate of C stocks in the layer up to 30 cm deep was obtained by considering a mapping of the various soil-vegetation associations. Regarding soil types, six categories were considered: mineral soils with high-activity clay, mineral soils with low-activity clay, sandy soils, volcanic soils, flooded soils and organic soils. The map legend for Brazilian soils of the time was divided into those six categories and the values 1

1 Pg = 1 billion tons.

of C in the soil were quantified in 5,585 horizons by wet combustion with dichromate. Brazilian soils had a total of 36.4 ± 3.4 Pg C in the 0 cm to 30 cm layer, of which, 22.7 Pg C, i.e., 62% of the total, was in the Legal Amazon. The use of soil for agricultural purposes resulted in a decrease of the total stock, to 34.4 Pg C in 1995 (BERNOUX et al., 2002). Under the Agrogases network, Fidalgo et al. (2007) estimated the soil C stocks in Brazil from the SigSolos soil database (CHAGAS et al., 2004). As suggested by the IPCC (2006), the study considered the Brazilian Soil Classification System (SBCS) [Sistema Brasileiro de Classificação de Solos] (SISTEMA..., 1999) which is in harmony with the FAO (1998) system and the national biomes. Out of 8,441 horizons, 2,257 profiles (Figure 2) containing C values, also determined by wet combustion, only 1,542 horizons contained soil density information. The following pedotransfer function, developed by Benites et al. (2007) was used to estimate the soil density of the remaining horizons: SD = 1.56 − (0.0005 * clay) − (0.01 * C) + (0.0075 * SB) where SD is the soil density in g cm−3, clay is the clay content in g kg−1, C is the content of organic C in g kg−1, and SB is the sum of bases (Ca2+ + Mg2+ + K+ + Na+).

CHAP 2 - FIGURE 2 Perfil de Solo

Soil Profile

Bioma

Biome

Amazónia

Amazon

Mata Atlântica

Atlantic Forest

Figure 2. Map of Brazil with the location of the soil profiles used in this study to estimate the soil carbon stock.

Afterward, for each soil profile, the total content of C was calculated, considering the set of horizons with depths to 30 cm. For this, it was necessary to establish a set of rules to standardize the estimate to that depth, i.e., profiles with final depth less than 20 cm were excluded, horizons with initial depth exceeding 25 cm or with final depth exceeding 40 cm were disregarded. Total C in the initial 30 cm of the soil was obtained for each profile by adding the estimate of C for each horizon in the profile with a maximum depth of 30 cm. Two different datasets, here called groupings, were considered: 1) average C by soil class; and 2) average C in soils under the same type of use in each biome of the Brazilian territory. Due to the lack of representative data on all types of soils under various land uses in each biome, these scenarios were created to analyze the results of the soil C amount estimates, considering some of these factors in isolation and thus getting samples with a larger number of profiles.

To estimate the area for each soil class, land use and biome, the following sources were used: the soil map of Brazil (IBGE, 2001), the vegetation map of Brazil (IBGE, 2004a) and the map of biomes (IBGE, 2004b), at scale 1:5,000,000. The ArcGIS software, version 9.1, from ESRI was used for the calculation of the area and integration of spatial information. Average C estimates were obtained directly from SigSolos. To estimate the average per soil class, soils present in this database were classified to the second category level according to the Brazilian Soil Classification System (SBCS) [Sistema Brasileiro de Classificação de Solos] (SISTEMA..., 1999). In this case, the information in the database regarding the soil attributes and classification of profiles according to previous taxonomic systems was used. Averages by land use were obtained from the use categorization defined based on its description in each profile. Averages by biome were obtained by identifying, for each profile, the biome in which it occurred. In this case, the following information from the profiles obtained in SigSolos was used: geographical coordinates of the profiles or (when this information was lacking), location of the seat of the municipality where the profile was collected. The C stock estimate in each grouping was obtained by summing the product of the estimated C average and each dataset’s corresponding area. This refers to soil classes or to various land uses for each biome. ????? where C = soil carbon stock, ????? = average of C of set j, Aj = area of set j. Considering the various soil types, data from 1,712 profiles under all land uses was used, resulting in an estimate of 36.59 Pg of C for the 0 cm to 30 cm layer of Brazilian soils (Table 1). Table 1. C stock from 0 cm to 30 cm, number of profiles (n), standard deviation (s) and product of the standard deviation by the area of each soil class in Brazil. Soil class

C (kg m−2)

n

s

Area (km2)

C (Pg)

s × area (Pg)

Alisol

8.48

2

1.75

245,404

2.08

0.43

Argisol [US: Ultisol]

3.79

383

1.91

2,028,094

7.68

3.86

Cambisol [US: Inceptisol]

5.50

125

2.61

446,548

2.46

1.16

Chernosol

4.70

3

1.39

37,206

0.17

0.05

Spodosol

4.12

6

3.75

167,506

0.69

0.63

Gleysol

6.60

75

3.67

398,901

2.63

1.46

Latosol [US: Oxisol]

4.18

608

1.59

2,678,308

11.21

4.27

Luvisol

5.39

56

2.72

241,910

1.30

0.66

3.65

76

1.70

25,268

0.09

0.04

5.62

80

3.36

615,748

3.46

2.07

3.03

46

2.88

477,915

1.45

1.38

Psamment]

1.26

3

0.31

18,842

0.02

0.01

Nitisol

5.50

118

2.07

96,533

0.53

0.20

Planosol

2.27

50

1.34

170,560

0.39

0.23

Hydromorphic Planosol

3.22

2

0.32

56,002

0.18

0.02

Plinthosol

3.67

64

2.05

588,449

2.16

1.21

Vertisol

5.00

15

2.68

17,631

0.09

0.05

8,310,828

36.59

17.72

Fluvic Neosol [US: Fluvent] Litholic Neosol [US: Lithic Orthent / Lithic Psamment] Quartzarenic Neosol [US: Quartzipsamment] Regolithic Neosol [US:

Totals

1,712

Source: Adapted from Fidalgo et al. (2007).

The distribution of estimated C classes on the map of Brazil based on soil types is presented in Figure 3. Spaces with zero C values correspond mainly to areas with water, rocky outcrop and dunes (2.0% of the national territory). Spaces in blue correspond to areas for which there is no profile information for estimating C – the area of Organosols [US: Histosols] – totaling 0.026% of the national territory.

CHAP 1 - FIGURE 3

Estoque de carbono

Carbon stock

Sem informação

No information

Figure 3. Estimate of C stock considering soil types in Brazil. Source: Fidalgo et al. (2007).

Making a correction for the national territory’s total soil area and including the area for which there is no information (Organosols [US: Histosols]), an estimate of 36.60 Pg of C at 0 cm – 30 cm was obtained. This correction was made by associating the average amount of carbon obtained for the rest of the national territory to areas without information. The estimated national stock was very close to the one calculated by Bernoux et al. (2002), although they considered only profiles under native vegetation, i.e., representing the initial state before the colonization of Brazil, in 1500. Considering the various soil types in each biome, 1,700 profiles were used and the C estimate was somewhat less, equal to 36.30 Pg of C. A possible explanation for the small difference between the two estimates is that most C stocks of the Brazilian soils are concentrated in the Amazon biome, which is still predominantly under native vegetation. It is important to point out the high variability of the estimate in the present study (s = 17.7 Pg C), which reflects the high diversity of Brazilian soils. Figure 4 shows the map of Brazil with distribution of soil C classes estimated for each biome and use. Water, sand dunes and rocky outcrop areas appear with a C value of zero. On the map, the lack of data for the Pantanal and Pampa biomes can be noted.

CHAP 1 - FIGURE 4 Estoque de carbono

Carbon stock

Sem informação

No information

Figure 4. Estimate of C stock considering biome types and soil uses in Brazil. Source: Fidalgo et al. (2007).

For analyses considering soils in use (anthropic area) and under original vegetation, combined with the various biomes, due to the lack of standardization and detail in the information on land use in the database (e.g., timeframe for particular type of use), there was a need to establish a general classification to standardize the available samples. Table 2 presents the estimates of C by use and biome. The total number of profiles (770) and biomes (Amazon, Caatinga, Cerrado and Atlantic Forest) analyzed was lower, resulting in lower estimates of C (31.2 Pg C at 0 cm – 30 cm). Table 2. C Estimates at 0 cm to 30 cm, number of samples (n), standard deviation (s) for each biome and land use. Biome

Land

C

n

s

Area

C

use / cover

(kg −2 m )

(km²)

(Pg)

Amazon

Anthropic area

4.62

67

2.16

672,631.21 3.11

Amazon

Original vegetation 3.47

40

1.49

3,408,607.63 11.83

Caatinga

Anthropic area

184

1.88

320,915.04 1.01

Caatinga


2

1.72

497,774.19 1.95

Cerrado

Anthropic area

3.97

173

2.03

1,167,933.83 4.63

Cerrado


14

2.70

859,528.86 3.77

4.40

267

2.29

960,491.65 4.23

Atlantic Forest Original vegetation 5.29

22

2.33

1,245,25.93 0.66

1 0.07

14,187.63 0.02

Atlantic Forest Anthropic area

Pantanal

Anthropic area

Total

3.14

1.10

770

8,026,595.98 31.20

Source: Data from Fidalgo et al. (2007).

The land use / biome combinations in the map in Figure 4 for which there was no C estimate occupy a total area of 1.5% of the national territory. Making a correction of the estimate for the national territory’s total area, including the area for which there is no information, an estimate of 32.32 Pg of C at 0 cm – 30 cm was obtained.

Final considerations Compiling a national soil database and adding new profiles, primarily for regions where there is a lack of information, is necessary for a more accurate assessment of Brazilian carbon stocks. The estimate of carbon stocks in Brazilian soils in this study reveals several sources of errors, caused mainly by gaps in the soil database, in terms of the amount of representative profiles by soil type and biome, and the quality of the information on the type of land use. In this study, due to information gaps, areas in which there was forest extraction were also classified as anthropic soil; i.e., areas under heavy use and areas under systems where there was no soil disturbance by plowing or harrowing, such as chestnut collecting, were included in the same category. Accurate information on land use, and on the timeframe when a particular type of use has been adopted up when the profile sample was collected, would provide a more robust estimate of carbon in soils under the influence of various types of use.

References

BATJES, N. H. Organic carbon stocks in the soils of Brazil. Soil Use and Management, Oxford, v. 21, p. 22-24, 2005. BENITES, V. M.; MACHADO, P. L. O. A.; FIDALGO, E. C. C.; COELHO, M. R.; MADARI, B. E. Pedotransfer functions for estimating bulk density of Brazilian soils. Geoderma, Amsterdam, NL, v. 139, p. 90-97, 2007. BENITES, V. M.; MACHADO, P. O. A.; FIDALGO, E. C. C.; COELHO, M. R.; MADARI, B. E.; LIMA, C. X. Funções de pedotransferência para estimativa da densidade dos solos brasileiros. Rio de Janeiro: Embrapa Solos, 2006. 31 p. (Embrapa Solos. Boletim de pesquisa e desenvolvimento, 104). BERNOUX, M.; ARROUAYS, D.; CERRI, C.; VOLKOFF, B.; JOLIVET, C. Bulk densities of Brazilian Amazon soils related to other soil properties. Soil Science Society of America Journal, Madison, v. 62, p. 743-749, 1998. BERNOUX, M.; CARVALHO, M. C. S.; VOLKOFF, B.; CERRI, C. C. Brazil’s soil carbon stocks. Soil Science Society of America Journal, Madison, v. 66, p. 888-896, 2002. CHAGAS, C. S.; CARVALHO JUNIOR, W.; BHERING, S. B.; TANAKA, A. K.; BACA, J. F. M. Estrutura e organização do sistema de informações georreferenciadas de solos do Brasil (SIGSOLOS – Versão 1.0). Revista Brasileira de Ciência do Solo, Viçosa, v. 28, p. 865-876, 2004. DE VOS, B.; MEIRVENNE, M. van; QUATAERT, P.; DECKERS, J.; MUYS, B. Predictive quality of pedotransfer functions for estimating bulk density of forest soils. Soil Science Society of America Journal, Madison, v. 69, p. 500-510, 2005. ESWARAN, H.; BERG, E. van den; REICH, P. Organic carbon in soils of the world. Soil Science Society of America Journal, Madison, v. 57, p. 192-194, 1993. FAO. Food and Agriculture Organization. World reference base for soil resources. Rome, IT: FAO: ISSS: ISRIC, 1998. 88 p. (FAO. World soil resources reports, 84). Available at: . Accessed on: 17 Sept. 2010. FIDALGO, E. C. C.; BENITES, V. M.; MACHADO, P. L. O. A.; MADARI, B. E.; COELHO, M. R.; MOURA, I. B.; LIMA, C. X. Estoque de carbono nos solos do Brasil. Rio de Janeiro: Embrapa Solos, 2007. 26 p. (Embrapa Solos. Boletim de pesquisa e desenvolvimento, 121). HOWARD, P. J. A.; LOVELAND, P. J.; BRADLEY, R. I.; DRY, F. T.; HOWARD, D. M.; HOWARD, D. C. The carbon content of soil and its geographical distribution in Great Britain. Soil Use and Management, Oxford, v. 11, p. 9-15, 1995. IBGE. Instituto Brasileiro de Geografia e Estatística. Mapa de biomas do Brasil. Rio de Janeiro, 2004b. 1 mapa, color. Escala 1:5.000.000. Available at: . Accessed on: 17 Nov. 2006. IBGE. Instituto Brasileiro de Geografia e Estatística. Mapa de solos do Brasil. Rio de Janeiro, 2001. 1 mapa, color. Escala 1:5.000.000. Available at: . Accessed on: 17 Nov. 2006. IBGE. Instituto Brasileiro de Geografia e Estatística. Mapa de vegetação do Brasil. Rio de Janeiro, 2004a. 1 mapa, color. Escala 1:5.000.000. Available at: . Accessed on: 17 Sept. 2010. IPCC. Intergovernmental Panel on Climate Change; UNEP. United Nations Environment Programme; OECD. Organization for Economic Co-Operation and Development; IEA. International Energy Agency. Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories. Paris, FR, 1997. v. 3. IPCC. Intergovernmental Panel on Climate Change. Climate change 2001: synthesis report: third assessment report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press, 2001. IPCC. Intergovernmental Panel on Climate Change. Guidelines for National Greenhouse Gas Inventories Programme. Hayama: IGES, 2006. MANRIQUE, L. A.; JONES, C. A. Bulk density of soils in relation to soil physical and chemical properties. Soil Science Society of America Journal, Madison, v. 55, p. 476-481, 1991. SCHROEDER, P. E.; WINJUM, J. K. Assessing Brazil’s carbon budget: 1. Biotic carbon pools. Forest Ecology and Management, Amsterdam, NL, v. 75, p. 77-86, 1995. SISTEMA brasileiro de classificação de solos. Brasília, DF: Embrapa Produção de Informação; Rio de Janeiro: Embrapa Solos, 1999. 412 p.

SISTI, C. P. J.; SANTOS, H. P.; KOHANN, R.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M. Change in carbon and nitrogen stocks in soil under 13 years of conventional or zero tillage in southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 76, p. 39-58, 2004. SMITH, P.; POWLSON, D. S.; GLENDINNING, M. J.; SMITH, J. U. Preliminary estimates of the potential for carbon mitigation in European soils through no-till farming. Global Change Biology, Oxford, v. 4, p. 679-685, 1998. STATSOFT. STATISTICA for Windows. Version 5.5. Tulsa, 1999. Software. TOMASELLA, J.; HODNETT, M. G. Estimating soil water retention characteristics from limited data in Brazilian Amazonia. Soil Science, Baltimore, v. 163, p. 190-202, 1998.

Chapter 2

Carbon stocks in Brazilian soils: Quantity and mechanisms for accumulation and preservation Robert M. Boddey, Bruno J. R. Alves, Segundo Urquiaga, Cláudia P. Jantalia, Ladislau Martin-Neto, Beata E. Madari, Débora M. B. P. Milori, Pedro L. O. de A. Machado

Abstract: The increase of carbon stocks in soil under agricultural use represents an important strategy to mitigate greenhouse gas emissions, especially in the case of Brazil, where this sector is responsible for over 20% of the country’s total emissions. The results of research studies conducted within Embrapa’s Agrogases network have confirmed the benefits of the no-tillage technique to accumulate C in soil, but this accumulation is dependent on a positive N balance in the system; i.e., N inputs must exceed quantities exported in harvested products and losses caused by various soil processes. This conclusion also extends to grazing pastures. Introducing N2-fixing leguminous plants in a green manure system in crop rotation or intercropping in pastures makes it possible to balance the system with N, fostering primary production and accumulation of C in the soil. In the case of grazing pastures, controlling the animal stocking rate has an effect on losses and, if performed well, can contribute to maintaining plant production and soil C gains. The magnitude with which the C increase occurs in deeper tropical soils is best assessed when soil samples are taken at a depth up to 1 m, contrasting with global estimates, which are limited to shallower soil layers (0 cm – 20 cm and 0 cm – 30 cm). Studies conducted in sugarcane crops revealed only tendencies for the elimination of burning to increase soil C stocks. Results are presented on ways and mechanisms to protect soil C, including chemical and physical fractions, determined in the first case by spectroscopic techniques, and in the second case by density and particle size. Keywords: carbon, soil organic matter, grazing pastures, tropical soils, Brazil.

Introduction Agricultural activities in tropical and subtropical regions of Brazil have proven to be highly dependent on organic matter for their sustainability, due to interrelationships between the soil’s physical, chemical and biological characteristics and soil C. In this context, C sequestration, a natural process of transferring atmospheric CO2 to soil organic matter (SOM) through photosynthesis, becomes much more important, due to the need to restore degraded soils and ecosystems, improve water quality, and increase biodiversity and agricultural productivity, in order to achieve adequate levels of food security (LAL, 2006). However, production systems currently in use in various types of soil and climate in Brazilian agricultural areas have often contributed less C to soil when compared to natural ecosystems, since tilling the soil with plows and other devices stimulates the degradation of SOM and the release of soil C in the form of CO2. The accumulation of soil C is controlled by the deposition rate of waste added to the soil and its decomposition rate. In soils with high natural fertility, soil C stocks under native vegetation are often near that soil’s maximum C storage capacity (CHUNG et al., 2008). In some ecosystems, however, the soil’s natural fertility is low and primary plant production is limited in such a way that, after applying lime and chemical fertilizers to agricultural systems, pastures or planted forests, their primary production could significantly exceed that of the native vegetation. This phenomenon is

more common in the highly weathered soils of tropical regions; and in Brazil, the most important example is the Cerrado area (CORAZZA et al., 1999; JANTALIA et al., 2007). When there is a loss of soil C after submitting a natural ecosystem to agricultural use, etc., it is due to the influence of several factors: • Change in plant species: perennial species return larger amounts of C to the soil, especially in the form of roots, when compared with annual species. • Change in decomposition rate: tilling the soil increases contact between plant residues and soil, which, when associated with changes in water and temperature regimes and increased use of mineral sources of nitrogen, increases the decomposition rate of the C in the soil. • Soil erosion and structure: the cultivation of the surface layer promotes the breakdown of soil aggregates, favoring the erosion process. When land use is changed, the C stock begins to shift until a new equilibrium is reached. The new state of equilibrium of soil C is reached when the decomposition rate of organic C in the soil is equal to the rate of C input in the form of plant residues. The point of equilibrium is difficult to estimate due to the influence of factors such as the increasing recalcitrance of the remaining undecomposed fraction and the regulation of the quantity and diversity of the soil’s biological component, which is heavily dependent on the substrate supply. Several chemical mechanisms (intrinsic complexity of the chemical compounds of C, complexation with mineral fraction or organo-mineral interaction) and physical mechanisms (degree of soil aggregation) can protect the various SOM pools from the action of microbial decomposers (PICCOLO; MBAGWU, 1999). These mechanisms can determine which molecules of the same composition and complexity constitute the various SOM fractions. In addition to determining how much organic C is in the soil, it is important to identify the degree of stability of the organic matter it is associated to, since this will determine how long the C is sequestered in the soil. Soil aggregation (joining of primary particles to form aggregates) is subject to various processes in the soil. Primary particles are joined by electrostatic interactions mediated by SOM, forming microaggregates. These are then joined by the action of roots and fungal hyphae or organic compounds produced by microbial decomposition of plant residues, forming macroaggregates. Plant residues and organic compounds are incorporated within and between aggregates, acquiring various degrees of protection against decomposition. Humification is a process of stabilization of SOM that requires more time to contribute to C sequestration but, on the other hand, can ensure that it remains longer in the soil. This chapter presents the results of studies conducted on the impact of using various soil management practices on carbon stocks of Brazilian soils. Results of studies are also presented assessing agricultural practices that can potentially promote accumulation of C, such as: no-tillage, crop-livestock integration in no-tillage, productive grazing pastures and agroforestry systems. First, issues will be approached relating to the assessment of C stocks in annual crops under conventional preparation and under conservation systems. Then sugarcane production systems will be discussed. Studies on grazing pasture areas were divided into biomes according to the diversity of use and management practices for this agricultural activity in each region. Lastly, mechanisms are discussed for protecting C in soil aggregates and the characterization of SOM in various soil fractions.

Changes in the soil’s carbon stocks based on management practices Annual crops: no-tillage versus conventional tillage Ten studies were conducted within the Agrogases network in the South region of Brazil, in medium and long term experiments with no-tillage (NT) and conventional tillage (CT), on the accumulation of soil C, in order to assess the impact of various soil management practices. Most soils belonged to the class of Red-Yellow Latosols [Latosols = US: Oxisols; Red-Yellow Latosols = US: Rhodic/Xanthic Haplustox], with free drainage, but with a differentiated texture. The first striking observation is the impact of the composition of crop rotations on soil C accumulation when comparing the NT and CT systems (Table 1). The results suggest that two interdependent factors have a strong influence on the potential of crop rotation systems to accumulate C in the soil: total crop yield and amount of crop residues returning to the system, and presence of a positive N balance in the crop system. A third important factor to determine C stocks is sampling depth. Table 1. Summary of studies on long-term experiments comparing the crops rotations / sequences using no-tillage (NT) and conventional tillage (CT) with respect to changes in soil C stocks. Location

Soil

Crop rotation / sequence

Time (years)

Depth (cm)

8

20

9

30

NT-CT difference(1) (Mg ha−1)

Depth (cm)


Alic Humic 1. Lages (SC)

Cambisol [Cambisol = US:

Maize/soybean, beans, wheat, oats(3)

8.5

Inceptisol] 2. Eldorado do Sul (RS)

Dystrophic Red US: Ultisol] Dystrophic Red

3. Santa Maria (RS)

Oats/maize

Argisol [Argisol =

4.6

Oats + vetch/maize + cowpea

6.4 0.7(3)

Black oats + vetch/maize

1.24(3)

Latosol [Red

Ryegrass + vetch/maize

Latosol = US:

Velvet bean + maize

5.42(3)

Limber caper + maize

0.65(3)

Wheat/soybean

-1.3

Rhodic Haplustox]

Dystrophic Red 4. Passo Fundo

Latosol [Red

(RS)(2)

Latosol = US: Rhodic Haplustox]

8

20

Wheat/soybean – vetch/maize

13

30

Wheat/soybean –

(RS)

US: Ultisol]

2.9(4)

4.3(4)

Oats/maize with N

Argisol [Argisol =

17

17.5

16.9

16.7

1.7(4)

Oats/maize without N 5. Eldorado do Sul

100

9.1

oats/Soybean – vetch/maize

Dystrophic Red

5.4

0.5

7.2(4) 107.5

Lablab + maize without N

13.9(4)

20.3(4)

Lablab + maize with N

18.6(4)

29.4(4)

Location

Soil

Crop rotation / sequence

Time (years)

Depth (cm)

Pigeon pea + maize

(RS)

28.4(4)

Pigeon pea + maize with N

20.5(4)

33.2(4)

Oats/maize without N

6.8

Oats/maize with N

3.2

Ryegrass/maize without N

6.3

Argisol [Argisol =

Ryegrass/maize with N

13

30



7.7

with N Dystrophic Red

Wheat – soybean

Latosol [Red Latosol = US:

5.8 7.2

without N

7. Cruz Alta (RS)


Dystrophic Red US: Ultisol]

(2)

Depth (cm)

13.4(4)

without N

6. Eldorado do Sul


30

-4.1

100

-3.8

17 Wheat/soybean – vetch/maize

3.6

8.8

Rhodic Haplustox] Dystrophic Red 8. Londrina (PR)

(2)

Latosol [Red

Lupin/maize – oats/soybean

Latosol = US:

– vetch/soybean

6

20

0.58

80

1.05

Rhodic Haplustox] Dystrophic Red 9. Londrina (PR)(2)

Wheat/soybean

Latosol [Red

Lupin/maize – oats/soybean-

Latosol = US:

wheat/soybean – wheat/

Rhodic Haplustox]

8.7 15

50 3.4

soybean Oats/maize – oats/soybean – 3.6

wheat/soybean – lupin/maize – oats/soybean Dystrophic Red 10. Londrina (PR)

(2)

Latosol [Red Latosol = US: Rhodic Haplustox]

Oats/soybean – lupin/maize – oats/ soybean –

5

40

5.9

wheat/soybean – lupin/maize Oats/soybean – wheat/soybean – lupin/maize –

-2.0

oats/maize – wheat/maize (1)

Indicates no difference between NT and CT.

(2)

Experiments conducted by researchers in Embrapa’s Agrogases network.

(3)

Crop sequences not provided.

(4)

The reference is fallow/maize under NT.

Source: 1. Bayer and Bertol (1999); 2. Bayer et al. (2000b); 3. Amado et al. (2001); 4. Sisti et al. (2004); 5. Dieckow et al. (2005); 6. Lovato et al. (2004); 7. Jantalia (2005); 8. Zotarelli et al. (2005); 9. Franchini et al. (2006); 10. Franchini et al. (2007).

Accumulation of carbon in the soil and deposition of crop residues The amount of crop residues deposited in the soil depends on crops yields and on the proportion of total dry matter removed from the field in the form of agricultural products. In the case of conventional tillage, residues are buried at the depth of penetration of the plow; in the no-tillage system, these residues remain on the soil surface, so only the first few centimeters of soil are enriched with C derived from that material. The accumulation of C, however, occurs under the no-tillage system mainly due to the soil remaining untilled, which prevents the destruction of aggregates and thus inhibits mineralization of SOM (oxidation of organic compounds with CO2 release). The mechanisms responsible for the “stabilization” of C in soils are discussed in the section “SOM Stabilization and Protection Mechanisms” of this chapter. Thus, C from crop shoot residues will be considered to contribute to C accumulation only in topsoil. C accumulation in deeper layers of the soil is derived from the root system of the crops in the system. The presence of plant residues is of great importance to reduce soil erosion, preserve moisture and to maintain the soil’s temperature regime, reducing the occurrence of high temperatures, an extremely important factor in field conditions in many regions of Brazil. The quality of the plant residues, largely influenced by the C/N ratio may be associated with the positive relationship between the dry waste produced annually in rotation systems and the differences in soil C stocks between NT and CT systems.

The accumulation of carbon in the soil and the N balance N used in agricultural systems comes from two main sources: synthetic fertilizer and biological nitrogen fixation (BNF). N from fertilizer is more prone to immediate losses promoted by processes such as ammonia volatilization, leaching or denitrification than BNF. Fertilizer, however, has an advantage in the fact that the producer can schedule the application in accordance with times of greater crop demand. A crop rotation system’s N balance is determined by the sum of inputs from fertilizer and BNF, minus the N exported in agricultural products and the N released by various processes. Soybean is the main grain legume crop in Brazil, which is why it is compulsory in most crop rotation systems under NT, but the N exported in the beans corresponds to approximately 75% of N in the above-ground part. Results of the quantification of the BNF of soybean in commercial production areas and in several experiments included in Table 1 show that the proportion of N derived from BNF in crops rarely exceeds 80% (ALVES et al., 2003, 2006; HUNGRIA et al., 2006). Therefore, in most cases the N balance of soybean crops tends to be neutral, or not positive enough to suppress fertilization in other crops in the system. So, in crop rotation systems in which soybean is the only legume, there are few cases in which the C stock of soils under NT is larger than under CT (see parts 4, 7 in Table 1). In many studies with crop rotations, when another legume besides soybean is part of the rotation system, such as winter legumes, which efficiently associate with atmospheric N fixing bacteria, the N balance becomes positive. In order for there to be an effective increase in SOM, it was crucial for the following crop to be grown on its residues (typically maize). Another common point in these studies (JANTALIA, 2005; LOVATO et al., 2004; SISTI et al., 2004) (see parts 4, 6 and 7 in Table 1)

was that the C gains under rotations were observed in NT several years (13 or more) after adoption of the system, in relation to CT. Results presented by Dieckow et al. (2005) also demonstrated that the addition of fertilizer N and the inclusion of legumes benefited both crop residue production and increase of soil C in long-term (17 years). The same behavior was not observed in crop rotation systems that did not contain legumes. The difference in soil C stocks between NT and CT in lablab (Lablab purpureum) / maize rotation increased from 13.9 Mg C ha−1 to 18.6 Mg C ha−1 after adding 150 kg N ha−1 year−1 to the maize (location 5). For pigeon pea (Cajanas cajan) / maize rotation, the increase was from 13.4 Mg C ha−1 to 20.5 Mg C ha−1 with the addition of the same amount of fertilizer N. In the rotation system that included oats in winter and maize in summer, and with no addition of N, the stock of C in the soil (0 cm – 100 cm) decreased by 7.2 Mg C ha−1; however, with an annual average application of 150 kg N ha−1 to maize, the decrease was reduced to 4.1 Mg C ha−1. The quality of the plant biomass added to the soil, especially in terms of N content, plays an important role in increasing the soil’s C stock, but in might not be the determining factor. Franchini et al. (2007) observed differences in soil management systems only in rotation with higher proportion of legumes, with C stock in the layer up to 40 cm at 74 Mg ha−1 in NT and at 68 Mg ha−1 in CT. After five years, this difference of 6 Mg ha−1 determined an accumulation rate of 1.2 Mg ha−1 year−1, under NT. Stock differences, however, were more associated with a decrease in C stocks under CT than to an actual increase in C stock under no-tillage systems. Therefore, an increase in the supply of N from leguminous plants may also increase the decomposition rate of C under CT conditions.

Sampling depth In a comprehensive review of 276 long-term experiments comparing no-tillage, reduced tillage and conventional tillage systems, West and Post (2002) observed a mean increase of 570 kg C ha−1 (57 g m−2) year−1 in soil C stocks when comparing systems with low soil disturbance (notillage and reduced tillage) with conventional tillage. Baker et al. (2007), however, analyzing the results of that work, observed that soil sampling did not go deeper than 30 cm in any of the studies, and that in 60% of them, the depth was less than 20 cm. These authors also noted that in studies where soils were sampled to greater depths (especially those conducted in Canada) (VANDENBYGAART et al., 2003), C stocks were higher in CT, suggesting that the apparent preferential accumulation of C in soils under NT was an artifact of sampling depth. Similar results were found in pioneering studies in the Cerrado region (CENTURION et al., 1985; CORAZZA et al., 1999). Research groups from Embrapa’s Agrogases network and others conducted studies in which soils were sampled at depths up to 100 cm or more (BODDEY et al., 2010; DIECKOW et al., 2005; JANTALIA, 2005; JANTALIA et al., 2007; SISTI et al., 2004; ZOTARELLI et al., 2005). These studies showed higher C stocks in NT when compared to CT in soil samples taken at depths up to 17.5 cm, 20 cm or 30 cm (depending on the study), with differences significantly increasing when the soil was sampled at 80 cm or more (see locations 9, 10, 11–16, 24 and 25; Table 1). Boddey et al. (2010) found in 10 NT – CT comparisons, in which a legume was included in intercropping with maize or in the winter before the maize, an increase of 60%, on average, in the difference between C stocks when the soil was sampled at 100 cm, instead of the 30 cm depth recommended by the IPCC (2006). This set of results contradicts the conclusion presented by Baker et al. (2007) that the change from CT to NT, or other reduced tillage system, would not have any benefits in terms of increasing

C stocks in soil (sequestration of atmospheric CO2). A possible reason for these contradictory findings may lie in the different physical characteristics between these regions (North America x southern Brazil). The free drainage Latosols [US: Oxisols] of southern Brazil allow for deeper roots. On the other hand, in the Cerrado region, the concentration of rainfall in the summer and the presence of toxic levels of aluminum, or deficiency of calcium, in the deeper layers of the soil (GOEDERT et al., 1985; RITCHEY et al., 1980) limit the growth of crop roots. This may be a possible cause for the lack of differences in soil C stocks when comparing NT and CT, in samples taken at depths up to 100 cm, as reported by Centurion et al. (1985) and Corazza et al. (1999). Besides the drainage factor, in the South region of Brazil, the agricultural use of the soil for a longer period of time, when compared to the Cerrado, entails the application of larger quantities of lime and a higher cation exchange capacity; this probably provides more favorable conditions for development of deeper roots in crops such as maize and oats. This fact may at least help understand why C stocks in soils under NT are the same as in soils under CT, but it still does not explain why C stocks at greater depths are higher than under CT. The fact that there is a large area in the South region of Brazil dedicated to crop rotations under NT (estimated at over 12 M ha today) (FEBRAPDP, 2010), entails important consequences. In rotation systems which included other legumes besides soybean, and maize or oats, the assessment of the impact of introducing NT on accumulation of soil C must consider soil layers up at least to 80 cm deep. These results also show that the South region of Brazil has the greatest potential, in terms of climate and soil, for C sequestration under the NT system.

Sugarcane In Brazil, more than half of the area used for sugarcane cultivation is for bioethanol production (currently about 27 billion liters per year), to be used in hydrated form (95% ethanol) or mixed with gasoline (24%) in light vehicles (cars and trucks). Currently, approximately 40% of all fuel used in light vehicles is made from ethanol. Recent studies have shown that producing nine units of bioethanol energy requires only one unit of fossil fuel energy, an energy balance of 9:1 (BODDEY et al., 2008; MACEDO, 1998), which is an extremely favorable proportion. With increases in the price of crude oil in international markets, bioethanol made from sugarcane is becoming extremely attractive, which combined with the recent development of bifueled “FlexFuel” engines (gasoline or hydrated ethanol) indicates that the consumption of bioethanol will grow significantly and, consequently, so will the sugarcane cultivation area in Brazil. Assessing the impact of expanding the sugarcane area on C stocks requires more information. Sugarcane crops will probably expand over flat areas currently with low productivity pastures or that are degraded. There are, however, few studies on the impact of such land use changes on C stocks. A study by Campos (2003), in the northeast of the State of Espírito Santo, analyzed the effects of the conversion of an area that was deforested in 1980, kept under pasture for 10 years and subsequently cultivated with sugarcane for 12 years, while another area was kept under pasture (22 years after deforestation). All the areas in the study (forest, pasture and sugarcane) were adjacent and had the same type of soil (Latosol [US: Oxisol], 80% sand). The C stock observed in the first 100 cm depth was 62.0 Mg ha−1 under forest area, 50.9 Mg ha−1 under sugarcane, and 70.8 Mg ha−1 under pasture. The pasture (Brachiaria decumbens) was kept without any fertilization. Although the sugarcane has been fertilized (ratoon received 80 kg N ha−1), the reduction of C stocks

was probably associated with the intensive cultivation used in the crop establishment and in the two subsequent renewals every five years. These results demonstrate that, in sandy soils, cultivation operations entail significant losses in the soil’s C stock. In recent years, a new management practice in which raw (green) sugarcane is harvested in sugarcane areas has been growing. Since the 1940s until recently, sugarcane would be burned just before the harvest to facilitate manual harvesting. In the past decade there has been increasing pressure from environmental agencies, especially in São Paulo, to abandon pre-harvest burning to reduce environmental pollution. To this date, no long-term studies (over 10 years long) have been conducted in São Paulo on the impact of green cane harvesting on C stocks. However, in 1983, the team at Embrapa Agrobiology conducted an experiment in Pernambuco to assess the effects of green cane harvesting versus burnt cane, the fertilization of plant cane and ratoon with N (0 kg N ha−1 and 80 kg N ha−1) and the annual addition of 80 m3 ha−1 of vinasse on sugarcane fields, the accumulation of N by the crop and C stocks in the soil (RESENDE et al., 2006). All treatments were harvested manually. The data indicates that raw sugarcane showed higher productivity than burnt cane, with differences of 24% (12.5 Mg ha−1 year−1) for the first cycle (1983–1992), and 45% (16 Mg ha−1 year−1) for the second cycle (1992–1999). Regarding C stock, the data shows that after 16 years, in the layer up to 20 cm deep, there was an increase of 4.28 Mg ha−1 of C in raw cane when compared to burnt cane, or a mean increase of 270 kg C ha−1 year−1. Another study on the impact of burning in the production of sugarcane was conducted in 1989 in the Lagrisa Plant, in Linhares, Espírito Santo, also with manual harvesting. In 2003, the soil under that experiment was sampled at depths up to 100 cm and the stocks of C and N in the soil were assessed (PINHEIRO et al., 2010). The cane yield did not differ from one treatment to another. After 14 years, however, C stocks were 13 Mg C ha−1 higher where the cane trash was conserved, an average of nearly 1 Mg C ha−1 year−1. In this study, the plantation was not renewed during those 14 years, which could explain why the difference in soil C stocks was higher than in the previous study conducted in Pernambuco (RESENDE et al., 2006). Studies on the short-term (4 years) effects of this practice were conducted in a Red Latosol [US: Rhodic Haplustox], near Ribeirão Preto, São Paulo (SP), in areas with mechanical harvesting (LUCA et al., 2008). An increase in C stock was observed at depths up to 40 cm, in relation to burned areas, with values ranging between 4.8 Mg ha−1 and 7.8 Mg ha−1. Based on this data, the authors estimated that there was an accumulation of 1.2 kg C ha−1 year−1 to −1 −1 1.9 kg C ha year , in a period without replanting. When the cane is replanted, the soil is plowed to a depth of approximately 40 cm, with heavy and powerful machinery, incorporating residues, which accelerates the processes of SOM mineralization. For this reason, the C accumulation rate should be less than 1 Mg C ha−1 year−1 when considering a longer period of time, including replanting operations.

The accumulation of carbon in soils under Brachiaria pastures Amazon Rainforest Biome The deforestation of the Amazon Rainforest is a subject generating great concern among ecologists and other specialists worldwide. A very common situation is deforestation followed by installation of pastures, especially with the Brachiaria spp. species. Currently, in pastures of this region, there is a predominance of two species, B. humidicola and B. brizantha. Surveys conducted by researchers from Embrapa Western Amazon and the National Institute for Amazonian Research

(INPA) [Instituto Nacional de Pesquisas da Amazônia] suggest that there are between 20 and 24 million hectares (Mha) of Brachiaria spp. in this region (FEARNSIDE; BARBOSA, 1998). In general, the main objective of deforestation is to secure land tenure, livestock production being of secondary importance. Therefore, in many cases, pastures, although well established, have low cattle stocking rates. In some regions, at the end of the rainy season, there is a surplus of forage: and in the dry season, pastures are burned (OLIVEIRA et al., 2001), which leads to loss of nutrients in the form of aerosols and by leaching, apart from great losses of N and S in gaseous form. Since the 1980s, several studies are being conducted within the program Large Scale Biosphere-Atmosphere Experiment in Amazonia (LBA) in various regions of the Amazon. This cooperation involves researchers from France and the United States working with Brazilian institutions such as the National Institute for Amazonian Research (INPA), the National Institute for Space Research (INPE) [Instituto Nacional de Pesquisas Espaciais], the Center for Nuclear Energy in Agriculture (CENA) [Centro de Energia Nuclear na Agricultura] and the Luiz de Queiroz College of Agriculture – University of São Paulo (ESALQ – USP) [Escola Superior Luiz de Queiroz – Universidade de São Paulo]. Some of these studies concerned changes in soil C stocks during the process of deforestation followed by the establishment of pastures. As these studies were not conducted by Embrapa’s Agrogases network, only a summary of the most important results are presented in Table 2. Table 2. Estimates(1) of changes in soil carbon stocks after deforestation and establishment of Brachiaria spp. pasture in the Amazon Rainforest biome.

Location

Pasture age (years)

Sampling depth (cm)

2

20

8

Carbon stock in soil (kg C ha−1) Forest

Pasture

30

90

98

2

10

12

12

4

10

12

17

8

10

12

15

11

10

12

14

19

10

12

16

80

10

12

21

5

30

35

46

81

30

33

72

41

30

81

30

1. Manaus (AM)

2. Nova Vida Farm (RO)


47


7 5. Paragominas (PA)

12 17

100

27

50

116

116

116

123

116

126

6. Vitória Paragominas Farm (PA)

Degraded

100

109

93

The data presented leads to the conclusion that keeping pastures under fertility managements and proper animal stocking rates can increase the soil’s C stocks, in relation to the original forest 10 years after conversion. However, if pastures are not fertilized and stocking rates are high, or if fire is used too often to “clean” pastures, this will result in degradation of pastures, with consequent reduction of C stocks in the soil, when compared to the original forest. Atlantic Forest biome There are no reliable estimates on the extent of pastures planted with African grasses in the Atlantic Forest biome. Boddey et al. (2006) reached an estimate of 20 Mha based on the 1995–1996 Agricultural Census (IBGE, 2006). In this biome, there are still few studies on the influence of establishing pastures on soil C stocks. A study was conducted by members of Embrapa’s Agrogases network, which included a medium-term experiment (9 years) (TARRÉ et al., 2001). The C stock at depths of up to 30 cm was monitored at three different times of the experiment: at the beginning (1988), after 6 years and after 9 years since the establishment (1994 and 1997). The experiment was conducted in the far south of Bahia and compared intercropped pastures (Brachiaria humidicola with Desmodium ovalifolium) with B. humidicola in monoculture, subject to continuous grazing with three beef cattle stocking rates: 2, 3 and 4 animals ha−1. The pasture’s C incorporation was determined by analyzing the change in the SOM’s C isotope ratio (13C / 12C). After six years, approximately 30% of C derived from the forest was lost (9.1 Mg C ha−1), tending to stabilize thereafter. From 1988 to 1997, the C derived from Brachiaria (Cdb) increased by 13.9 Mg C ha−1 under B. humidicola pastures, and did not differ significantly between stocking rates. In 1997, samples were collected from all pastures, at depths up to 100 cm, and no significant contribution from Cdb was found in layers deeper than 40 cm. At less than 30 cm depth, there was a greater accumulation of C in the soil under pastures intercropped with D. ovalifolium than under pastures with B. humidicola in monoculture. The authors estimated that, on average, the rate of C accumulation in the soil (0 cm – 100 cm) under intercropped pastures was 1.17 Mg C ha−1 year−1, almost twice the accumulation rate observed in soil under pastures with B. humidicola monoculture (0.66 Mg C ha−1). The study conducted by Campos (2003), cited above, monitored adjacent forest and pasture areas for 22 years, and others, which were converted into sugarcane cultivations after 10 years of pasture. Based on these results, it was concluded that after 22 years of pasture, there was a significant increase in the soil’s (0 cm – 100 cm) C stock (8.8 Mg C ha−1), in relation to the original forest (62.0 Mg C ha−1). In the area that was under pasture for 10 years, followed by 12 years of sugarcane, however, the soil’s C stock was 50.9 Mg C ha−1, 18% less than under the original forest (Figure 1). With the 13C abundance data, it was possible to evaluate the dynamics of C derived from forest, from Brachiaria and from sugarcane in this chronosequence. Under the pasture, after 20 years, C derived from forest decreased by 9%; but in soil where there was intense tilling, under sugarcane crops, the amount of C derived from the original forest decreased by 35%. The decrease in the

SOM’s C stocks under sugarcane, when compared with Brachiaria, occurred by decomposition of the forest’s original SOM. As the areas had the same amounts of crop residues, this increased decomposition was caused by plowing and furrowing. After the establishment of sugarcane, the plantation was renewed twice, with deep plowing and 40 cm deep furrows. This means that the soil in this area has undergone great physical disturbance (tillage).

CHAP 2 - FIGURE 1 Estoques de carbono

Carbon stocks

Mata

Native forest

Pasto

Pasture

Cana-de-açúcar

Sugarcane

Cobertura

Coverage

Figure 1. Carbon stock to a depth up to 1 m and quantities derived from forest vegetation (C-C3) and C4 crops (C-C4) in soil under secondary forest, under a 22 year old Brachiaria brizantha pasture, and under an area that was under a B. decumbens pasture for 10 years, followed by sugarcane for 12 years. Alcon Plant, municipality of Conceição da Barra, Espírito Santo (ES). Source: Campos (2003).

Cerrado Biome In a detailed survey, based on the 1995 Census done by the Brazilian Institute of Geography and Statistics (IBGE) [Instituto Brasileiro de Geografia e Estatística], Sano et al. (2000) estimated that there were 49.5 M ha of pastures (mainly Brachiaria spp.) in the Cerrado biome, which corresponds to 25% of the total area of this biome. However, there are still few studies on changes in soil C stocks under pastures in this region. Corazza et al. (1999) studied soil C stock under Brachiaria decumbens, in a Dark Red Latosol [US: Rhodic Haplustox], at the Embrapa Cerrados experimental station, in 1976. In 1982, the area was prepared with plowing plus two disc harrowings and cultivated with soybean, and then reestablishing the pasture in 1983. Soil samples were taken from the pasture and from an area of native Cerrado vegetation, in soil layers between 0 and 100 cm deep. The apparent density of each layer was determined, and the C stock was calculated, but without correction by the equivalence of the mass in the profile. The soil (0 cm – 100 cm) under pasture had 150 Mg C ha−1, 16.6 Mg C ha−1 higher than the one under the Cerrado vegetation. The accumulation rate was, on average, 0.92 Mg ha−1 year−1 for the 18 years since establishing the pasture. However, these values are probably overestimated, as no correction was made for soil mass contained in the profile. No data was provided on the beef cattle stocking rate, but the authors reported that, at the time of sampling, the pasture was degraded. In another study conducted by Silva et al. (2004), in the same Embrapa Cerrados experimental station a comparison was made between C stocks under Cerrado and in 6 pastures, intercropped with legumes or not. The C stock was calculated as carried out by Corazza et al. (1999), with no information on the date when the pastures were established. Under some pastures

there was an increase in the C stock up to 13 Mg ha−1 in relation to that observed in native Cerrado. Several observations were made by the authors: the pastures B. brizantha and Paspalum atratum (treatments 6 and 7) were not grazed; they were used for seed production. The intercropped pastures (treatments 4 and 5) did not survive until the time of sampling, which makes these pastures similar to the non-intercropped, at the end of the study. The B. decumbens pasture was described as degraded at the time of sampling. The authors pointed out that only the pastures fertilized with P and K and with a low (or zero) stocking rate accumulated significant amounts of C in the soil profile. Because these management practices are rarely adopted in the Cerrado biome, they concluded that in most cases it is unlikely that C stocks under pastures with Brachiaria or others are higher than under native vegetation. In studies with chronosequences (areas under native vegetation next to altered areas), it is essential to verify whether the physical parameters, among others, are similar, in order for the chronosequence to be validated. This means examining soil properties under various vegetation types (native vegetation and pastures) to verify the possibility of soil C stocks being equal or close at the time of the pasture establishment. Many studies conducted in tropical and subtropical regions showed that there is a strong positive relationship between the percentage of silt+clay in the soil and its total C content (FELLER et al., 1991; JONES, 1973; LEPSCH et al., 1982, 1994; TALINEAU et al., 1980). In a study with samples (0 cm – 20 cm) from 65 soils with low-activity clays from Africa, the West Indies, Brazil and Southern India, Feller et al. (1991) reported a high correlation coefficient (r = 0.87***) between the quantity of silt+clay and total C concentration. For this reason, the amount of silt+clay is used as an indicator of the original C content in the soils of the various chronosequences. Recently, a study on C stocks in soil under productive pastures, degraded pastures and native vegetation aimed to validate chronosequences at four sites under Cerrado vegetation (BRAZ et al., 2010). Each site’s location, soil type and land use history are presented in Table 3. Table 3. Location, soil type and land use history of four sites in the Cerrado biome where carbon stocks in chronosequences of productive and degraded pastures were evaluated in comparison with adjacent native vegetation. Location Soil

History

Clay content Ribeirão Farm, Chapadão do Sul (MS) Red Latosol [US: Rhodic Haplustox] 11% Paraíso Farm, Penápolis (SP) Red-Yellow Latosol [Latosols = US: Oxisol; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox]

Area deforested in 1980 to establish B. decumbens pastures. In 1991, part of the area was planted with soybean for four years with subsequent establishing of a B. brizantha pasture (“productive pasture”). The oldest pasture is in a stage of degradation Area deforested in 1985 to establish B. decumbens pasture. In 1996, part of the area was planted with sugarcane for subsequent regeneration of the B. decumbens pasture (“productive pasture”). The oldest pasture is in a stage of degradation

16% Carumbézinho Farm, Itaporã (MS) Dark Red Latosol [US: Rhodic Haplustox] 46%

Area deforested in 1985 to establish B. decumbens pasture. In 1998, lime was applied in part of the area and a B. brizantha pasture (“productive pasture”) was established. The oldest pasture is in a stage of degradation

Paloto Farm, Luz (MG) Dark Red Latosol [US: Rhodic Haplustox] 67%

Coffee and other crops have been cultivated in this area since the 1930s. The area was subsequently abandoned and a natural pasture of Melinis minutiflora [molasses grass; capim gordura] and Hyparrhenia rufa [thatching grass; jaraguá] was formed. In the 1970s, a B. decumbens pasture was established in the area. In 1998, lime and rock phosphate were applied in part of the area and a B. brizantha pasture (“productive pasture”) was established. The oldest pasture is in a stage of degradation

Source: Braz et al. (2010).

As expected, among these sites, soil C stocks were higher where clay content was higher (Paloto Farm) (Table 4). Also for those sites, C stocks up to 40 cm or 100 cm were higher where productive pastures of Brachiaria were established and maintained. C stocks under degraded pastures were lower than under productive pastures, but were still above those under native forest. In several of these areas, productive pasture had a higher soil C stock than under native forest. Table 4. Corrected soil carbon stocks (Mg C ha−1), up to depths of 40 cm and 100 cm, in chronosequences of Chapadão do Sul, Penápolis, Itaporã and Luz, under native forest vegetation, productive pasture and degraded pasture. Stock to depth of 40 cm Location

Ribeirão Farm Chapadão do Sul (MS) Paraíso Farm Penápolis (SP) Carumbézinho Farm Itaporã (MS) Paloto Farm Luz (MG)

Native forest

Productive pasture

Degraded pasture

34.3

38.2

29.7

27.8

33.6

32.4

51.8

56.8

NA

70.7

84.3

77.4

Stock to depth of 100 cm Ribeirão Farm Chapadão do Sul (MS) Paraíso Farm Penápolis (SP) Carumbézinho Farm Itaporã (MS) Paloto Farm Luz (MG)

57.1

62.6

53.1

55.6

62.0

60.5

83.3

95.4

NA

117.0

164.6

138.0

NA: No data available. Source: Braz et al. (2010).

This study also evaluated the C derived from the original forest that was gradually partially decomposed and replaced by C derived from Brachiaria (Figures 2 and 3). Under productive

pastures, C derived from Brachiaria was found up to the maximum sampling depth (100 cm). In soil under degraded pastures, however, C derived from Brachiaria was only found at depths less than 50 cm. The results obtained in this study under Cerrado vegetation led to the following conclusions: 1) Pastures kept in production with low stocking rates and modest additions of fertilizer (P, K and lime) can accumulate much higher quantities of soil C than the original native forest, especially in soils with higher clay; 2) Under these pastures, a considerable quantity of C derived from grass can be stored deep in the soil; 3) When the pasture’s productivity decreases due to lack of maintenance and fertilizer and an excessive stocking rate, soil C derived from grass begins to be released and will no longer contribute to soil C stock in layers deeper than 50 cm.

CHAP 2 - FIGURE 2 Profundidade

Depth

C derivado da vegetação nativa

C derived from native vegetation

C derivado da Brachiaria spp.

C derived from Brachiaria spp.

Conteúdo de C no solo

Soil C content

VN

NV

PP

PP

PD

DP

Figure 2. Carbon derived from native Cerrado vegetation C3 and Brachiaria spp. C4, in native Cerrado vegetation (NV) compared to productive pasture (PP) and degraded pasture B. decumbens (DP). Paloto Farm (Luz, Minas Gerais (MG)). Source: Braz et al. (2010).


Depth

C derivado da vegetação nativa

C derived from native vegetation

C derivado da Brachiaria spp.

C derived from Brachiaria spp.

Conteúdo de C no solo

Soil C content

VN

NV

PP

PP

PD

DP

Figure 3. Carbon derived from native Cerrado vegetation C3 and Brachiaria spp. C4, in native Cerrado vegetation (NV) compared to productive pasture (PP) and degraded pasture B. decumbens (DP). Ribeirão Farm (Chapadão

do Sul, Mato Grosso do Sul (MS)). Note that the scale on the horizontal axis is approximately half of that in Figure 2. Source: Braz et al. (2010).

Rehabilitation of degraded areas using leguminous trees Soils under roads or other constructions, or that are severely degraded, are extremely poor in all essential plant growth nutrients. This makes it necessary to apply large amounts of fertilizers and soil conditioners to recover those areas, which significantly increases costs and makes this an unappealing enterprise. Embrapa Agrobiology’s team for Rehabilitation of Degraded Areas (RAD) [Recuperação de Áreas Degradadas] developed a set of technologies using mycorrhizal legume trees inoculated with N-fixing bacteria, which ensure re-vegetation at more attractive costs (FRANCO et al., 1992). Leguminous trees, such as Acacia spp., Enterolobium contortisiliquum, Gliricidia sepium, Mimosa caesalpiniifolia and Paraserianthes falcataria, are able to obtain large quantities of N via BNF, especially when they are inoculated with more efficient bacterial strains. Interaction with arbuscular mycorrhiza ensures a better absorption of phosphorus. These interactions with diazotrophic bacteria and mycorrhizal fungi cause these plants to be able to develop even under very drastic conditions of soil degradation. These plants have intense biomass production, causing the area to be re-vegetated in a short period of time, by activating biological processes that result in an increase of SOM and they also favor colonization by other tree species and other plants. A hillside area degraded by the removal of the surface layer in the city of Angra dos Reis, Rio de Janeiro (RJ) was revegetated using this technique in 1991 (MACEDO et al., 2008). The species used in this plantation were: Acacia mangium, Acacia auriculiformis, Enterolobium contortisiliquum, Gliricidia sepium, Leucaena leucocephala, Mimosa caesalpiniifolia and Paraserianthes falcataria, all inoculated with selected strains of rhizobia and mycorrhizal fungi spores. In 2004, 13 years after planting, soil samples were taken at up to 60 cm deep in the rehabilitated area, in a non-rehabilitated adjacent area (covered with spontaneous vegetation, mainly Panicum maximum [guinea grass; colonião]) and in an area with primary native forest. The apparent density of the soil was also determined. The C and N stocks were quantified and compared under the same soil mass. In the non-rehabilitated area, the soil carbon stock was only 65.1 Mg C ha−1, against 88.1 Mg C ha−1 in the rehabilitated area, and 107.7 Mg C ha−1 in the area under native forest. In relation to the non-rehabilitated area, the rehabilitated area accumulated an additional 23 Mg C ha−1 in 13 years, demonstrating the great value of this technology, not only in restoring the forest and stabilizing the hillside, but also in sequestering carbon at a mean rate of 1.8 Mg C ha−1 year−1.

Soil carbon stability indicators SOM stabilization and protection mechanisms Light and heavy SOM fractions

The absence of soil tillage in systems such as NT or pastures reduces the decomposition rate of SOM and avoids the breakdown of aggregates, contributing to preserving or increasing C stocks in the soil. SOM is made up of two fractions with different stages of decomposition, which are called light and heavy. The light fraction is composed of organic fractions with an intermediate degree of decomposition, between plant residues and humus, a relatively low density, about 1.0 g cm−3 (ANDERSON; INGRAM, 1993). This fraction, also called macroorganic matter, or particulate organic matter (POM), constitutes one of the smallest pools of the OM and contains around 10% of the total organic C (TOC) (SILVA; RESCK, 1997). The heavy fraction, responsible for the majority of C present in the soil. comprises more processed fractions, in an intimate association with soil minerals, and has an advanced degree of decomposition and higher density (> 2 g cm−3) than the light fraction. In addition to management practices adopted, levels of C in the light fraction are also influenced by the soil type and the climate (JANZEN et al., 1992). In tropical soils, the C loss rate of the light fraction is 2 to 11 times greater than that of the heavy fraction (DALAL; MAYER, 1986). This higher rate of decomposition in the light fraction is due to the labile nature of its constituents and the lack of protection by soil colloids (DALAL; MAYER, 1986). The dynamics of the light fraction are influenced by the type of soil preparation and are dependent on C added to the soil. Thus, rotation systems that prioritize the return of plant residues back to planting areas make for a greater availability of macroorganic matter in the soil (JANZEN et al., 1992). In general, effects of management practices on the light fraction contents are restricted to the plowing layer. In no-tillage areas, fluctuations in the light fraction’s C content are most noticeable in surface layers, usually in the top 5 cm of soil, where the highest deposition of plant residues occurs (JANZEN et al., 1992). In areas with greater tillage and low return of C to the soil, OM decomposition is accelerated. This loss is more dramatic in the SOM’s light fraction (CHRISTENSEN, 1992). Results in Brazilian conditions have also indicated that organic C is present in various soil pools, which have different potentials for emitting C into the atmosphere (FREITAS et al., 2000; FREIXO et al., 2002). Data from Bayer (1996), in long-term experiments in the South region of Brazil, showed the distribution of C in various soil pools and indicated that, after 9 years of cultivation, the light fraction increased significantly under NT. However most of the C was accumulated in fractions considered heavy [coarse silt (53 μm – 20 μm), fine silt (20 μm – 2 μm) and clay (< 2 μm)], as shown in Figure 4. It is noteworthy that the highest accumulation of C was observed in the fine silt fraction (20 μm – 2 μm). These fractions are considered to be the ones with greater stability in the soil and are called the humified fraction of the SOM. The next section will discuss in greater detail the characteristics of SOM added to the soil and the respective humified fractions. Protection of the SOM in soil aggregates The physical fractionation of soil has also indicated that its aggregation is controlled by management practices (SIX et al., 1999, 2000), representing a mechanism to protect and stabilize inter- and intra-aggregate C in Brazilian soils with low-activity clay (DENEF et al., 2007; MADARI et al., 2005; ZOTARELLI et al., 2005, 2007). In this respect, Zotarelli et al. (2005) observed that NT increased the number of macroaggregates and kept higher levels of organic matter than CT. However, no differences were

observed in C content in various classes of aggregates, i.e., there was no clear hierarchy between aggregates in terms of accumulation of C, as suggested by Tisdall and Oades (1982). The authors suggest that, while accumulation of organic matter was favored by the absence of soil tillage, organic matter played a secondary role in formation of aggregates in these low-activity clay soils. Therefore, although increased aggregate stability in response to NT is common in this type of soil, the relationship between aggregate stability and organic matter content was not confirmed.

CHAP 2 - FIGURE 4 Tamanho da partícula

Particle size

Fração leve

Light fraction

Fração humificada

Humified fraction

COT

TOC

PD

NT

PC

CT

Figure 4. Distribution of C in the light and humified fractions of organic matter in an Argisol [US: Ultisol], in Eldorado do Sul, Rio Grande do Sul (RS), after 5 and 9 years of cultivation. (CT): conventional tillage; (NT): notillage. Source: Bayer (1996).

A complementary study characterizing free light and intra-aggregate light fractions was conducted by Zotarelli et al. (2007). Again, fractionation of aggregates showed that NT promotes conditions for forming aggregates, mainly in the surface layer, this effect being associated with higher C accumulation in NT in relation to CT. Furthermore, the concept of forming macroaggregates from microaggregates would explain the distribution of particulate C in various classes of aggregates. The results support the conceptual model of C recycling in macroaggregates determining the stabilization of organic matter such as fine particulate C in microaggregates proposed by Six et al. (1998, 1999) for temperate soils. The validation of this model seems to be conditioned to the NT practice timeframe, since it was confirmed in the area of Passo Fundo, which had 16 years of NT, and not in the Londrina area, where NT was only used for 5 years. Subsequently, Denef et al. (2007) confirmed the fraction of microaggregates occluded within macroaggregates as the location in which stabilization of organic matter occurs in NT. In the model proposed by the authors, stabilization of organic matter in microaggregates would be a reflection of the greater stability of these aggregates, determining a lower recycling of C. Thus, C associated with microaggregates occluded within macroaggregates could be used as an indicator of C sequestering potential in the conversion from CT to NT. Madari et al. (2005) observed a direct relationship between aggregate size and C content. The main difference between the works of Denef et al. (2007) and Zotarelli et al. (2005, 2007) in relation to Madari et al. (2005) is the methodology of obtaining aggregates in the field by wet

fractionation. In the studies by Denef et al. (2007) and Zotarelli et al. (2005, 2007) the aggregates selected had a maximum diameter of 8 mm, whereas Madari et al. (2005) assessed aggregates with a maximum diameter of 19 mm. The use of a 19 mm sieve for homogenization of the sampling and the use of an 8 mm sieve for the wet sieving procedure allowed for identification of a larger class of macroaggregates (8 mm-19 mm). The advantage of this method was that the capacity of NT to form macroaggregates was not underestimated, allowing for an association between aggregate size and C accumulation in NT. Thus, it appears that the process of protecting C in macroaggregates can be better understood if there is a minimum disturbance of the soil’s natural structure at the time of sampling.

Qualitative characterization of the SOM Determining the degree of humification of humic substances and organo-mineral fractions in areas with potential for accumulating C in the soil The concentration of stable organic free radicals, detected by Electron Paramagnetic Resonance (EPR), is related to the degree of humification of SOM (JERZYKIEWICZ et al., 2002; MARTIN-NETO et al., 1998; RIFFALDI; SCHNITZER, 1972; SAAB; MARTIN-NETO, 2004; SCHNITZER; LEVESQUE, 1979; SENESI, 1990a). Complex aromatic structures are associated with stabilization of semiquinone-type free radicals in humus (SENESI, 1990a; STEVENSON, 1994). There is a high positive correlation between concentration of semiquinone-type free radicals (spin) and C/H molar ratio (which increases with aromaticity) as well as other aromaticity measurements (MARTIN-NETO et al., 1994; OLK et al., 2000; SENESI, 1990a, 1990b; WIKANDER; NORDEN, 1988). With regard to C accumulation in the soil, knowledge on humification becomes important because, according to Lal (1997), one of the strategies to increase soil C accumulation is to create mechanisms to increase humification both on the surface and in the soil profile, from residues derived from biomass. In fact, determining the degree of humification from the concentration of semiquinone-type free radicals has been widely used to assess the effects of various soil management practices and crop rotations, allowing for a comparison of areas under forest (without cultivation), CT, reduced tillage and NT, areas with addition of sewage sludge, among others. Bayer et al. (2000a, 2000b, 2002) have demonstrated that areas under NT, with higher amounts of residues on the soil surface, in subtropical regions of Brazil, show a decrease in the degree of humification of humic acids (HA) and organo-mineral aggregates (fractions 53 μm – 20 μm and 20 μm – 2 μm), when compared with soils under CT. Figure 5 presents the data from comparison of the degree of humification of an Argisol [US: Ultisol] in the experimental field of Eldorado do Sul, Rio Grande do Sul (RS), detected by the concentration of semiquinone-type free radicals, evaluated in the fifth and ninth years of the experiments with soil management practices. Two concurrent aspects were observed: an increase in degree of humification of humic acid samples in areas under CT, and a decrease in degree of humification in areas under NT, when comparing the fifth and ninth year. It is worth mentioning that, in areas under NT, when compared to CT, there is a significant increase in the soil’s C content (BAYER et al. 2000a), and that the material incorporated into the soil in areas under NT is less humified. This indicates that keeping residues on the soil surface in NT causes less processing, as compared to CT, where residues are incorporated and intensively processed by microorganisms (BAYER et al., 2002; MILORI et al., 2002).

CHAP 2 - FIGURE 5 PD

NT

PC

CT

anos

years

A/M

O/M

A+V/M+C

O+V/M+CP

Manejo do solo e rotação de culturas

Soil management and crop rotation −1

Figure 5. Concentration of semiquinone-type free radicals, given in spins g , in humic acids (HAs) extracted from Argisol [US: Ultisol] soils in Eldorado do Sul, Rio Grande do Sul (RS), as a function of soil management and crop rotation. CT: conventional tillage, NT: no tillage; O/M: oats and maize; O+V/M+CP: oats+vetch/maize+cowpea. Source: Bayer et al. (2002).

Analysis of organo-mineral fractions extracted from this same soil has shown a decreasing degree of humification in the 53 μm – 20 μm to 20 μm – 2 μm fractions (which are considered to be humified); meanwhile, in the < 2 μm fraction, there were no changes after nine years of soil management practices (Figure 6).

CHAP 2 - FIGURE 6 Semiquinona (… de C)

Semiquinone (… of C)

PD

NT

PC

CT

Complexo organo-mineral

Organo-mineral fractions

−1 Figure 6. Concentration of semiquinone-type free radicals, in spins (g C) , for organo-mineral fractions (53 µm – 20 µm; 20 µm – 2 µm and < 2 µm) of an Acrisol, for areas under CT and NT.

Source: Bayer et al. (2002).

These results confirm the observations made with humic acid (HA) samples and demonstrate that the most stable SOM (humic acid or organo-mineral aggregates smaller than 53 μm) may undergo changes in relatively short periods of time. Also in areas under NT, there is a decrease in the degree of humification, indicating that, although the fractions being measured are considered to be the more stable ones of the SOM, they exhibit different levels of humification and, therefore, different levels of stability in the soil. Besides its importance for agricultural aspects, from the point of view of HA structure, the fact that there is a change in degree of humification suggests that the most probable molecular structure for HAs is an association of molecules of relatively low molecular weight, held together by

weak forces, such as hydrophobic interactions and hydrogen bonds, which is consistent with recent results published in the international literature (SUTTON; SPOSITO, 2005). The interaction between hydrophobic compounds, such as lignin and fatty acids, derived from plant residues and having microbial activity, increases the resistance and the permanence of humic substances in the soil (PICCOLO et al., 1996; PICCOLO; MBAGWU, 1999). Studies show that hydrophobic substances lead to the formation of stable aggregates in the soil, through hydrophobic protection mechanisms, capable of reducing the decomposition rate of that fraction of the SOM (BASTOS et al. 2005; PICCOLO; MBAGWU, 1999). These substances are mostly aliphatic saturated and it is commonly accepted that they are derived from biopolyesters, such as suberin and cutin, from esterified and/or complexed organic residues present in the macromolecular matrix of soils, and also in the form of iron and aluminum organic complexes (HANSEL et al., 2007). The origin of hydrophobicity can be explained by the association of the amphiphilic substances’ polar terminal with active sites of the clay and soil, which results in the forming of a hydrophobic film on the surface of these minerals. Hydrophobic protection can be a useful tool to limit decomposition of organic matter recently introduced into the soil and thus reduce CO2 emissions in agricultural soils on a global scale (PICCOLO; MBAGWU, 1999). Hydrophobicity of SOM can be considered an important criterion to evaluate its potential for accumulating stable C in the soil. Samples collected at various depths in a Dystrophic Litholic Neosol [Litholic Neosol = US: Lithic Orthent / Lithic Psamment], medium texture, native stage, soft undulating topographic relief, under a plantation of pines (Pinus taeda) showed a decrease of hydrophobicity in the surface up to a depth of 30 cm at 16 years, in the municipality of Piraí do Sul, Paraná (PR) (Table 5). Spectroscopic analysis (UV-Vis, Fluorescence, DRIFTS, X-Ray and EPR) of HAs extracted at various depths showed that, in deeper layers of the soil, there was an increase in conjugated organic structures, greater intensity of aromatic groups and organic free radicals, an increase in the degree of humification and a decrease in aliphatic groups in the HA fraction. These results suggest a surface with higher hydrophobicity, which is in accordance with the water repellency tests conducted on the same samples. Fluorescence of humic substances (HS) is already a widely used technique in environmental sampling (CARVALHO et al., 2004; ROSA et al., 2005; SENESI, 1990c). Techniques have been published in recent literature to determine the degree of humification in humic substances (HS) in solutions, based on photoluminescence. Zsolnay et al. (1999) proposed the ratio of areas (upper quarter more towards red, and lower quarter more towards blue) of an emission spectrum with ultraviolet (240 nm) excitation to assess humification of dissolved OM. Kalbitz et al. (1999) used the ratio between peaks observed in the synchronous scan spectrum (higher peak intensity for red / higher peak intensity for blue) to define the degree of humification of fulvic acids. Milori et al. (2002) demonstrated a good correlation between intensity of the fluorescence emission with excitation at blue (465 nm) and concentration of semiquinone-type free radicals, determined by EPR, for samples of humic acid extracted from soils. Table 5. Soil water repellency class, number of spins per gram and degree of humification of humic acids in a Litholic Neosol [US: Lithic Orthent / Lithic Psamment] at various depths. Depth (cm)

Repellency class WDPT(1) (s)

Number of spins g−1

Humification index(2) Fluorescence – excitation 465 nm

0–5

Strongly repellent – 430

2.96E+16

5–10

Strongly repellent – 66

8.11E+16

22,500

f10–30

Hydrophilic – 1

8.57E+16

29,000

(1)

Water drop penetration time (WDPT).

(2)

According to Milori et al. (2002).

17,000

Figure 7 shows a typical result of analyses performed on Brazilian Cerrado soils. This high correlation is due to the fact that the excitation in this wavelength is resonant with structures derived from the oxidation of phenolic compounds and quinones. Therefore, the fluorescence intensity for excitation at blue is proportional to the degree of humification of OM.

CHAP 2 - FIGURE 7 Intensidade

Intensity

Floresta Nativa

Native forest

Plantio Direto

No tillage

Plantio convencional

Conventional tillage

Emissão de fluorescência

Fluorescence emission

Excitação de varredura síncrona

Synchronous scan excitation

Excitação de fluorescência

Fluorescence excitation

Emissão de fluorescência

Fluorescence emission

Figure 7. Typical examples of UV-Vis light fluorescence spectra of humic acid from tropical soils under various management systems, with native area (without soil management), NT and CT. Samples were prepared in aqueous solutions (20 mg L−1, pH = 8). A) fluorescence emission (λexc = 240 nm); B) synchronous scan excitation (∆λ = 55 nm), C); fluorescence excitation (λem = 517 nm); D) fluorescence emission (λexc = 465 nm). Source: Martin-Neto et al. (2009) and Milori et al. (2002).

Figure 8 A, B and C present graphs corresponding to data obtained from these three methodologies in pastures and Cerradão reference areas, in São Carlos, São Paulo (SP).

CHAP 2 - FIGURE 8 Tratamentos

Treatments

Cerradão

Cerradão

Arado

Plowed soil

Figure 8. Humification degrees of HA from the pasture experiment, obtained from fluorescence studies proposed by: (A) Zsolnay et al. (1999) [A4/A1]; (B) Kalbitz et al. (1999) [I465/I399]; and (C) Milori et al. (2002) [A465]; for treatments: Cerradão area, plowed soil, T00 (pasture without N or lime); t0 (zero lime on surface, receiving NK);

t2m (2 t ha−1 of lime on the surface, with NK fertilization, and annual addition of 1 t ha−1 of lime); and t4sa (4 t ha−1 of lime on the surface in plots without NK) at depths of 0 cm – 10 cm and 10 cm – 20 cm. Source: Segnini (2007) and Segnini et al. (2011).

Considering the degree of humification only of the more humified material in the soil, such as HAs, it is observed that the degree of humification is higher in treatments that included addition of lime (t2m and t4sa). Due to the increase in pH, there was a stimulation of microbial decomposing activity, which led to an increase in SOM humification degree (SEGNINI, 2007; SEGNINI et al., 2011). Soil carbon analysis by infrared spectroscopy The implementation of the Kyoto Protocol, in 2008, has been encouraging projects that promote accumulation of carbon (C) in terrestrial ecosystems (MACHADO, 2002). Future policies to promote C sequestration in agriculture will inevitably result in the need to measure C (total or in different fractions) in the soil, at least at the beginning and end of a project, and in several sampling locations to verify the occurrence of additional accumulation of C and, if any, how much has been added in the period being considered (SMITH, 2004). Considering that the area of soil under economic use in Brazil (e.g., crops, pastures and forests) is 236.1 million hectares (MANZATTO et al., 2002), the estimated quantity of soil samples to be analyzed requires a method of determining soil C that is fast, inexpensive and at the same time highly accurate and precise. Conventional methods such as dry combustion and oxidation with potassium dichromate (NELSON; SOMMERS, 1996; WATSON et al., 2000) are either expensive, slow or both. Although the accuracy of dry combustion is considered satisfactory, the equipment for this analysis is expensive and requires costly maintenance. Oxidation with potassium dichromate determines only part of the organic C, it is inaccurate due to additional correction factors and it produces toxic laboratory waste containing copper. The muffle method (loss-on-ignition, LOI) is relatively fast, but also entails accuracy problems due to the decomposition of some mineral fractions together with organic matter (NELSON; SOMMERS, 1996; WATSON et al., 2000). Spectroscopic techniques such as medium infrared spectroscopy, diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and near infrared spectroscopy (NIRS), if previously calibrated, can simultaneously determine any number of variables in a single spectrum, and without costly additional reagents. Madari et al. (2006a), after testing the methods on 1,135 samples of various soil types collected in the various Brazilian biomes, verified that NIRS and DRIFTS are suitable techniques to quantify soil carbon. Madari et al. (2006b) also confirmed the possibility of using infrared spectroscopy to quantify the total N, clay and silt in the soil. Lastly, infrared spectroscopy also allows for evaluation of soil structure aspects by quantifying aggregate stability indexes (MADARI et al., 2006b), identified in this chapter as the main factors responsible for protecting soil organic matter (SOM) against rapid mineralization and consequent emission of greenhouse gases into the atmosphere.

Final considerations The issue of organic matter (OM) is sparking increasing interest due to its numerous functions in various environments, and especially because of its direct connection to the carbon

cycle, which is in turn associated with global climate change. Soil is the third largest global carbon reservoir (estimated at 2,300 Pg), exceeded only by the ocean (estimated at 38,000 Pg) and fossil (> 6,000 Pg) reservoirs (FIELD; RAUPAUCH, 2004). Considering that the ocean reservoir is not easily managed and that the fossil reservoir has substantially contributed to the emissions of carbon dioxide, the soil reservoir becomes very relevant, given its potential to sequester carbon, and thus contribute to mitigation of climate changes, especially in countries less dependent on energy from fossil fuels, such as Brazil, when compared with other industrialized or developing countries (TEIXEIRA et al., 2006). However, data and detailed studies on potential for carbon accumulation of soils from various biomes and under various management systems, as well as on the mechanisms involved and the stability of the sequestered carbon, the balance between carbon accumulation and greenhouse gas flow, among numerous associated issues, indicates that, despite the already available results and current efforts, there are several needs for further research, focusing on the following activities: 1) Assessing and monitoring carbon accumulation and greenhouse gas flow in various production systems, expanding and seeking to consolidate a database on a continental country like Brazil, with various biomes and numerous local and regional specificities. Studies would be carried out in areas with potential to sequester carbon, such as areas under no-tillage, crop-livestock-forestry integration, pastures, sugarcane, planted forests, recovering degraded areas, and areas with addition of organic waste, such as sewage sludge, etc. This work requires a broad cooperative effort by Embrapa Units and partner institutions throughout the country, with multidisciplinary and interinstitutional teams. The expectation is to ensure a representative and reliable database and to quantify the degree of sustainability of Brazilian agriculture, in terms of greenhouse gas balancing (sequestration in soil and emissions into the atmosphere), and generate contributions to public policies that encourage adoption of sustainable management practices by producers, including activities eligible for the Clean Development Mechanism (CDM) and which mitigate increased in the greenhouse effect. 2) Developing cheaper and more efficient procedures to identify the best functional combination of soil organisms that contribute to growth of plants and, consequently, greater production, (e.g., symbiotic microorganisms) and to the carbon stock in soil (e.g., activity of organisms for greater aggregation and stability of soil aggregates). 3) Developing and adapting methods and equipment, mainly portable ones, enabling faster and less expensive measurement of soil parameters, and, if possible, under real-time conditions. This group should emphasize the importance of measuring soil carbon, both regarding quantity (FERREIRA et al., 2008; SILVA et al., 2008), stability (MILORI et al., 2006, 2011), and systems for evaluating texture and density. Although the pedotransfer function is already used to estimate soil density in Brazilian biomes (BENITES et al., 2007), more agile and less costly measuring techniques would make it easier to quantify a soil’s carbon content and its various pools with their respective rates of CO2 emissions into the atmosphere. With versatile equipment and lower-cost analyses, it would be possible to quantify carbon stocks and assess their stability in broader areas, and to monitor actual changes in the landscape. This would significantly reduce the risks caused by the extrapolation of values obtained occasionally (e.g., experimental plots) to larger scales of the same soil type.

4) Developing qualitative studies on OM in soils, with physical fractionation (based on particle size and density) and chemical fractionation (obtaining humic substances), and physical methods of analysis (chromatography and spectroscopy), generating consistent information on the stability and reactivity of carbon compounds, to contribute to parameterizing the data to be used in carbon balance models in Brazil.

References ALVES, B. J. R.; BODDEY, R. M.; URQUIAGA, S. The success of BNF in soybean in Brazil. Plant and Soil, The Hague, v. 252, p. 1-9, 2003. ALVES, B. J. R.; ZOTARELLI, L.; FERNANDES, F. M.; HECKLER, J. C.; MACEDO, R. A. T. de; BODDEY, R. M.; JANTALIA, C. P.; URQUIAGA, S. Fixação biológica de nitrogênio e fertilizantes nitrogenados no balanço de N em soja, milho e algodão. Pesquisa Agropecuária Brasileira, Brasília, DF, v. 41, p. 449-456, 2006. AMADO, T. J. C.; BAYER, C.; ELTZ, F. L. F.; BRUM, A. C. Potencial de culturas de cobertura em acumular carbono e nitrogênio no solo no plantio direto e a melhoria da qualidade ambiental. Revista Brasileira de Ciência do Solo, Campinas, v. 25, p. 189-197, 2001. ANDERSON, J. M.; INGRAM, J. S. I. Tropical soil biology and fertility: a handbook of methods. 2nd ed. Wallingford: CAB International, 1993. 117 p. BAKER, J. M.; OCHSNER, T. E.; VENTEREA, R. T.; GRIFFIS, T. J. Tillage and carbon sequestration: What do we really know? Agriculture, Ecosystems and Environment, Amsterdam, NL, v. 118, p. 1-5, 2007. BASTOS, R. S.; MENDONÇA, E. S.; ALVAREZ, V. H. V.; CORRÊA, M. M. Formação e estabilização de agregados do solo decorrentes da adição de compostos orgânicos com diferentes características hidrofóbicas. Revista Brasileira de Ciência do Solo, Campinas, v. 29, p. 11-20, 2005. BAYER, C. Dinâmica da matéria orgânica em sistemas de manejo de solos. 1996. 240 f. Tese (Doutorado em Ciência do Solo) - Universidade Federal do Rio Grande do Sul, Porto Alegre. BAYER, C.; BERTOL, I. Características químicas de um Cambissolo húmico afetadas por sistemas de preparo, com ênfase à matéria orgânica. Revista Brasileira de Ciência do Solo, Campinas, v. 23, p. 687-694, 1999. BAYER, C.; MARTIN-NETO, L.; MIELNICZUK, J.; CERETTA, C. A. Effect of no-till cropping systems on soil organic matter in a sandy clay loam Acrisol from Southern Brazil monitored by electron spin resonance and nuclear magnetic resonance. Soil and Tillage Research, Amsterdam, NL, v. 53, p. 95-104, 2000a. BAYER, C.; MIELNICZUK, J.; AMADO, T. J. C.; MARTIN-NETO, L.; FERNANDES, S. V. Organic matter storage in a sandy clay loam Acrisol affected by tillage and cropping systems in Southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 54, p. 101-109, 2000b. BAYER, C.; MIELNICZUK, J.; MARTIN-NETO, L.; ERNANI, P. R. Stocks and humification degree of organic matter fractions as affected by no-tillage on a subtropical soil. Plant and Soil, The Hague, v. 238, p. 133-140, 2002. BENITES, V. M.; MACHADO, P. L. O. A.; FIDALGO, E. C. C.; COELHO, M. R.; MADARI, B. E. Pedotransfer functions for estimating bulk density of Brazilian soils. Geoderma, Amsterdam, NL, v. 139, p. 90-97, 2007. BODDEY, R. M.; JANTALIA, C. P.; CONCEIÇÃO, P. C.; ZANATTA, J. A.; BAYER, C.; MIELNICZUK, J.; DIECKOW, J.; SANTOS, H. P. dos; DENARDIN, J. E.; AITA, C.; GIACOMINI, S. J.; ALVES, B. J. R.; URQUIAGA, S. Carbon accumulation at depth in Ferralsols under zero-till subtropical agriculture. Global Change Biology, Oxford, v. 16, p. 784-795, 2010. BODDEY, R. M.; JANTALIA, C. P.; MACEDO, M. O.; OLIVEIRA, O. C. de; RESENDE, A. S.; ALVES, B. J. R.; URQUIAGA, S. Potential of carbon sequestration in soils of the Atlantic Forest region of Brazil. In: LAL, R.; CERRI, C. C.; BERNOUX, M.; ETCHEVERS, J.; CERRI, C. E. P. (Ed.). Carbon sequestration in soils of Latin America. New York: Howarth, 2006. p. 305-347. BODDEY, R. M.; SOARES, L. H. B.; ALVES, B. J. R.; URQUIAGA, S. Bio-ethanol production in Brazil. In: PIMENTEL, D. (Org.). Biofuels, solar and wind as renewable energy systems. Berkeley: Springer, 2008. p. 321-355.

BRAZ, S.; URQUIAGA, S.; ALVES, B. J. R.; JANTALIA, C. P.; GUIMARÃES, A. P.; BODDEY, R. M. Mudanças nos estoques do carbono do solo dos cerrados cultivados com pastagens de braquiária. In: REUNIÃO BRASILEIRA DE MANEJO E CONSERVAÇÃO DO SOLO E DA ÁGUA, 18., 2010, Teresina. Anais... Teresina: Sociedade Brasileira de Ciência do Solo, 2010. 1 CD-ROM. CAMPOS, D. V. B. de. Uso da técnica de 13C e fracionamento físico da matéria orgânica em solos sob cobertura de pastagens e cana-de-açúcar na região de mata atlântica. 2003. 220 f. Tese (Doutorado em Ciência do Solo) Universidade Federal Rural do Rio de Janeiro, Seropédica. CARVALHO, E. R.; MARTIN-NETO, L.; MILORI, D. M. B. P.; ROCHA, J. C.; ROSA, A. H. Interactions of chlorine with tropical aquatic fulvic acids and formation of intermediates observed by fluorescence spectroscopy. Journal Brazilian Chemistry Society, São Paulo, v. 15, p. 421-426, 2004. CENTURION, J. F.; DEMATTÊ, J. L. I.; FERNANDES, F. M. Efeitos de sistemas de preparo nas propriedades químicas de um solo sob cerrado cultivado com soja. Revista Brasileira de Ciência do Solo, Campinas, v. 9, p. 267-270, 1985. CHONÉ, T.; ANDREUX, F.; CORREA, J. C.; VOLKOFF, B.; CERRI, C. C. Changes in organic matter in an oxisol from the central Amazonian forest during eight years as pasture, determined by 13C isotopic composition. In: BERTHELIN,J. (Ed.). Diversity of environmental biogeochemistry. Amsterdam, NL: Elsevier, 1991. p. 397-405. CHRISTENSEN, B. T. Physical fractionation of soil and organic matter in primary particle size and density separates. Advances in Soil Science, New York, v. 20, p. 1-90, 1992. CHUNG, H.; GROVE, J. H.; SIX, J. Indications for soil carbon saturation in a temperate agroecosystem. Soil Science Society of America Journal, Madison, v. 72, p. 1132-1139, 2008. CORAZZA, E. J.; SILVA, J. E. D.; RESCK, D. V. S.; GOMES, A. C. Comportamento de diferentes sistemas de manejo como fonte ou depósito de carbono em relação à vegetação de Cerrado. Revista Brasileira de Ciência do Solo, Campinas, v. 23, p. 425-432, 1999. DALAL, R. C.; MAYER, R. J. Long-term trends in fertility of soils under continuous cultivation and cereal cropping in southern Queensland: II Total organic carbon and its rate of loss from the soil profile. Australian Journal of Soil Research, Victoria, v. 24, p. 281-292, 1986. DENEF, K.; ZOTARELLI, L.; BODDEY, R. M.; SIX, J. Microaggregate-associated carbon as a diagnostic fraction for management-induced changes in soil organic carbon in two Oxisols. Soil Biology and Biochemistry, Oxford, v. 39, p. 1165-1172, 2007. DIECKOW, J.; MIELNICZUK, J.; KNICKER, H.; BAYER, C.; DICK, D. P.; KOGEL-KNABNER, I. Carbon and nitrogen stocks in physical fractions of a subtropical Acrisol as influenced by long-term no-till cropping systems and N fertilisation. Plant and Soil, The Hague, v. 268, p. 319-328, 2005. FEARNSIDE, P. M.; BARBOSA, R. I. Soil carbon changes from conversion of forest to pasture in Brazilian Amazonia. Forest Ecology and Management, Amsterdam, NL, v. 108, p. 147-166, 1998. FEBRAPDP. Available at: . Accessed on: 10 Oct. 2010. FEIGL, B. J.; MELILLO, J.; CERRI, C. C. Changes in the origin and quality of soil organic matter after pasture introduction in Rondônia (Brazil). Plant and Soil, The Hague, v. 175, p. 21-29, 1995. FELLER, C.; FRITSCH, E.; POSS, R.; VALENTIN, C. Effets de la texture sur le stockage et la dynamique des matières organiques dans quelques sols ferrugineux et ferrallitiques (Afrique de l’Ouest, en particulier). Cahier ORSTOM: série Pédologie, Paris, FR, v. 26, p. 25-36, 1991. FERREIRA, E. C. ; MILORI, D. M. B. P. ; FERREIRA, E. J. ; SILVA, R. M. da; MARTIN-NETO, L. Artificial neural network for Cu quantitative determination in soil using a portable laser induced breakdown spectroscopy system. Spectrochimica Acta Part B, Oxford, v. 63, p. 1216-1220, 2008. FIELD, C. B.; RAUPAUCH, M. R. The global carbon cycle: integrating humans, climate, and the natural world. Washington, DC: Island Press: Scope, 2004. 527 p.

FRANCHINI, J. C.; CRISPINO, C. C.; SOUZA, R. A.; TORRES, E.; HUNGRIA, M. Microbiological parameters as indicators of soil quality under various tillage and crop-rotation systems in southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 92, p. 18-29, 2007. FRANCHINI, J. C.; MILORI, D. M. B. P.; TORRES, E.; MARTIN-NETO, L.; URQUIAGA, S. Grau de humificação e estoques de matéria orgânica em sistemas de manejo do solo e rotação de culturas. In: REUNIÃO BRASILEIRA DE MANEJO E CONSERVAÇÃO DO SOLO E DA ÁGUA, 16., 2006, Aracaju. Resumos e palestras... Aracaju: SBCS, 2006. 1 CD-ROM. FRANCO, A. A.; CAMPELLO, E. F. C.; SILVA, E. M. R.; FARIA, S. M. de. Revegetação dos solos degradados. Rio de Janeiro: EMBRAPA-CNPBS, 1992. 9 p. (Comunicado Técnico, 9). FREITAS, P. L. D.; BLANCANEAUX, P.; GAVINELLI, E.; LARRÉ-LARROY, M. C.; FELLER, C. Nível e natureza do estoque orgânico de Latossolos sob diferentes sistemas de uso e manejo. Pesquisa Agropecuária Brasileira, Brasília, DF, v. 35, p. 157-170, 2000. FREIXO, A. A.; MACHADO, P. L. O. D. A.; SANTOS, H. P. D.; SILVA, C. A.; FADIGAS, F. D. S. Soil organic carbon and fractions of a Rhodic Ferralsol under the influence of tillage and crop rotation systems in southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 64, p. 221-230, 2002. GOEDERT, W. J.; RITCHEY, K. D.; SANZONOWICZ, C. Desenvolvimento radicular do capim-andropogon e sua relação com o teor de cálcio no perfil do solo. Revista Brasileira de Ciência do Solo, Campinas, v. 9, p. 89-91, 1985. HANSEL, F. A.; AOKI, C. T.; MAIA, C. M. B. F.; DEDECEK, R. A.; FUKAMAKI, C. R. B. Estudo preliminar da hidrólise alcalina e a sua relação com a repelência à água em solos. In: CONGRESSO BRASILEIRO DE CIÊNCIA DO SOLO, 31., 2007, Gramado. Anais... Gramado: SBCS, 2007. 1 CD-ROM. HUNGRIA, M.; FRANCHINI, J. C.; CAMPO, R. J.; CRISPINO, C. C.; MORAES, F. L.; SIBALDELLI, R. N. R.; MENDES, I. C.; ARIHARA, J. Nitrogen nutrition of soybean in Brazil: contributions of biological N2 fixation and of N fertilizer to grain yield. Canadian Journal of Plant Science, Ottawa, CA, v. 86, p. 927-939, 2006. IBGE. Instituto Brasileiro de Geografia e Estatística. Mapa de biomas do Brasil. Rio de Janeiro. Mapa colorido. Escala 1:5.000.000. Available at: . Accessed on: 17 Nov. 2006. IPCC. Intergovernmental Panel Climate Change. Guidelines for National Greenhouse Gas Inventories Programme. Hayama: IGES, 2006. JANTALIA, C. P. Estudo de sistemas de uso do solo e rotações de culturas em sistemas agrícolas brasileiros: dinâmica de nitrogênio e carbono no sistema solo-planta-atmosfera. 2005. 302 f. Tese (Doutorado em Fitotecnia) – Universidade Federal Rural do Rio de Janeiro, Seropédica. JANTALIA, C. P.; RESCK, D. V. S.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M. Effect of tillage intensity on carbon stocks under a soybean based crop rotation in the Brazilian Cerrado. Soil and Tillage Research, Amsterdam, NL, v. 85, p. 97-109, 2007. JANZEN, H. H.; CAMPBELL, C. A.; BRANDT, S. A.; LAFOND, G. P.; TOWNLEY-SMITH, L. Light-fraction organic matter in soils from long-term crop rotations. Soil Science Society of America Journal, Madison, v. 56, p. 1799-1806, 1992. JERZYKIEWICZ, M.; JEZIERSKI, A.; CZECHOWSKI, F.; DROZD, J. Influence of metal ions binding on free radical concentration in humic acids: a quantitative electron paramagnetic resonance study. Organic Geochemistry, Amsterdam, NL, v. 33, p. 265-268, 2002. JONES, M. J. The organic matter content of the savanna soils of West Africa. Journal of Soil Science, Oxford, v. 24, p. 4253, 1973. KALBITZ, K.; GEYER, W.; GEYER, S. Spectroscopic properties of dissolved humic substances: a reflection of land use history in a fen area. Biogeochemistry, Amsterdam, NL, v. 47, p. 219-238, 1999. KOUTIKA, L. S.; BARTOLI, F.; ANDREUX, F.; CERRI, C. C.; BURTIN, G.; CHONÉ, T.; PHILIPPY, R. Organic matter dynamics and aggregation in soils under rain forest and pastures of increasing age in the eastern Amazon Basin. Geoderma, Amsterdam, NL, v. 76, p. 87-112, 1997. LAL, R. Residue management, conservation tillage and soil restoration for mitigating greenhouse effect by CO2enrichment. Soil and Tillage Research, Amsterdam, NL, v. 43, p. 81-107, 1997.

LAL, R. Soil carbon sequestration in Latin América. In: LAL, R.; CERRI, C. C.; BERNOUX, M.; ETCHEVERS, J.; CERRI, C. E. P. (Ed.). Carbon sequestration in soils of Latin America. New York: Howarth, 2006. p. 49-64. LEPSCH, I. F.; MENK, J. R. F.; OLIVEIRA, J. B. Carbon storage and other properties of soils under agriculture and natural vegetation in São Paulo State, Brazil. Soil Use and Management, Oxford, v. 10, p. 34-42, 1994. LEPSCH, I. F.; SILVA, N. M. da; ESPIRONELO, A. Relação entre matéria orgânica e textura de solos sob cultivo de algodão e cana-de-açúcar, no estado de São Paulo. Bragantia, Campinas, v. 41, p. 231-236, 1982. LOVATO, T.; MIELNICZUK, J.; BAYER, C. & VEZZANI, F. Adição de carbono e nitrogênio e sua relação com os estoques no solo e com o rendimento do milho em sistemas de manejo. Revista Brasileira de Ciência do Solo, Campinas, v. 28, p. 175-187, 2004. LUCA, E. F.; FELLER, C.; CERRI, C. C.; BARTHÈS, B.; CHAPLOT, V.; CAMPOS, D. C.; MANECHINI, C. Avaliação de atributos físicos e estoques de carbono e nitrogênio em solos com queima e sem queima de canavial. Revista Brasileira de Ciência do Solo, Campinas, v. 32, p. 789-800, 2008. MACEDO, I. C. Greenhouse gas emissions and energy balances in bio-ethanol production and utilization in Brazil. Biomass and Bioenergy, Oxford, v. 14, p. 77-81, 1998. MACEDO, O. M.; RESENDE, A. S.; GARCIA, P. C.; BODDEY, R. M.; JANTALIA, C. P.; URQUIAGA, S.; CAMPELLO, E. F. C.; FRANCO, A. A. Changes in soil C and N stocks and nutrient dynamics 13 years after recovery of degraded land using leguminous nitrogen-fixing trees. Forest Ecology and Management, Amsterdam, NL, v. 255, p. 1516-1524, 2008. MACHADO, P. L. O. de A. Mecanismo de desenvolvimento limpo (MDL): funcionamento, pontos críticos e possibilidades para alguns sistemas agrícolas no Brasil. Rio de Janeiro: Embrapa Solos, 2002. 28 p. (Embrapa Solos. Documentos, 41). MADARI, B. E.; REEVES, J. B.; COELHO, M. R.; MACHADO, P. L. O. A.; DE-POLLI, H.; COELHO, R. M.; BENITES, V. M.; SOUZA, L. F.; MCCARTY, G. W. Espectroscopia infravermelha para a determinação de carbono do solo: perspectiva de um método economicamente viável e ambientalmente seguro. Santo Antonio de Goiás: Embrapa Arroz e Feijão, 2006a. 6 p. (Embrapa Arroz e Feijão. Comunicado Técnico, 126). MADARI, B. E.; REEVES, J. B.; COELHO, M. R.; MACHADO, P. L. O. A.; GUIMARÃES, C. M.; TORRES, E.; MCCARTY, G. W. Mid- and near-infrared spectroscopic assessment of soil compositional parameters and structural indices in two Ferralsols. Geoderma, Amsterdam, NL, v. 136, p. 245-259, 2006b. MADARI, B.; MACHADO, P. L. O. A.; TORRES, E.; ANDRADE, A. G. D. E.; VALENCIA, L. I. O. No-tillage and crop rotation effects on soil aggregation and organic carbon in a Rhodic Ferralsol from southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 80, p. 185-200, 2005. MANZATTO, C. V.; RAMALHO FILHO, A.; COSTA, T. C. C. da; SANTOS, M. de L. M.; COELHO, M. R.; SILVA, E. F. da; OLIVEIRA, R. P. de. Potencial de uso e uso atual das terras. In: MANZATTO, C. V.; FREITAS JUNIOR, E.; PERES, J. R. R. (Ed.). Uso agrícola dos solos brasileiros. Rio de Janeiro: Embrapa Solos, 2002. p. 13-21. MARTIN-NETO, L.; ANDRIULO, A. E.; TRAGHETTA, D. G. Effects of cultivation on ESR spectrum of organic matter from soil size fractions of a Mollisol. Soil Science, Baltimore, v. 157, p. 365-372, 1994. MARTIN-NETO, L.; MILORI, D. M. B. P.; SILVA, W. T. L. da; SIMÕES, M. L. EPR, FTIR, Raman, UV-visible light absorption and fluorescence spectroscopies in studies of humic substances. In: SENESI, N.; XING, B.; HUANG, P. M. (Ed.). Biophysicochemical processes involving natural organic matter in environmental systems. Hoboken: Wiley IUPAC Series, 2009. p. 651-728. MARTIN-NETO, L.; ROSSEL, R.; SPOSITO, G. Correlation of spectroscopic indicators of humification with mean annual rainfall along a temperate grassland climosequence. Geoderma, Amsterdam, NL, v. 81, p. 305-311, 1998. MILORI, D. M. B. P.; MARTIN-NETO, L.; BAYER, C.; MIELNICZUK, J.; BAGNATO, V. S. Humification degree of soil humic acids determined by fluorescence spectroscopy. Soil Science, Baltimore, v. 167, p. 739-749, 2002. MILORI, D. M. B. P.; SEGNINI, A.; SILVA, W. T. L. da; POSADAS, A.; MARES, V.; QUIROZ, R.; MARTIN-NETO, L. Emerging techniques for soil carbon measurements. Copenhagen, DK,: CCAFS, 2011. 28 p. (CCAFS Working Paper, 2). Available

. Accessed on: 22 Nov. 2011. MILORI, D. M. P. B.; GALETI, H. V. A.; MARTIN-NETO, L.; DIECKOW, J.; GONZALEZ-PEREZ, M.; BAYER, C.; SALTON, J. Organic matter study of whole soil samples using laser-induced fluorescence spectroscopy. Soil Science Society of America Journal, Madison, v. 70, p. 57-63, 2006. MORAES, J.F.L.D.; VOLKOFF, B.; CERRI, C.C; BERNOUX, M. Soil properties under Amazon forest and changes due to pasture installation in Rondônia, Brazil. Geoderma, Amsterdam, NL, v. 70, p. 63-81, 1996. NEILL, C.; MELILLO, J. M.; STEUDLER, P. A.; CERRI, C. C.; MORAES, J. F. L. D.; PICCOLO, M. C.; BRITO, M. Soil carbon and nitrogen stocks following forest clearing for pasture in the southwestern Brazilian Amazon. Ecological Applications, Tempe, v. 7, p. 1216-1225, 1997. NELSON, D. W.; SOMMERS, L. E. Total carbon, organic carbon, and organic matter. In: SPARKS, D. L.; PAGE, A. L.; HELMIKE, P. A.; LOEPPERT, R. H.; SOFTANPOUR, P. N.; TABATABAI, M. A.; JOHNSTON, C. T.; SUMMER, M. E. (Ed.). Methods of soil analysis. Madison: Soil Science Society of America, 1996. p. 997-983. NEPSTAD, D. C.; CARVALHO, C. R. D.; DAVIDSON, E. A.; JIPP, P. H.; LEFEBVRE, P. A.; NEGREIROS, G. H.; SILVA, E. D. D.; STONE, T. A.; TRUMBORE, S. E.; VIEIRA, S. The role of deep roots in the hydrological and carbon cycles of Amazonian forests and pastures. Nature, London, GB, v. 372, p. 666-669, 1994. OLIVEIRA, O. C. de; OLIVEIRA, I. P. de; FERREIRA, E.; ALVES, B. J. R.; MIRANDA, C. H. B.; VILELA, L.; URQUIAGA, S.; BODDEY, R. M. Response of degraded pastures in the Brazilian Cerrado to chemical fertilization. Pasturas Tropicales, Cali, v. 23, p. 1418, 2001. OLK, D. C.; BRUNETTI, G.; SENESI, N. Decrease in humification of organic matter with intensified lowland rice cropping: A wet chemical and spectroscopic investigation. Soil Science Society of America Journal, Madison, v. 64, p. 1337-1347, 2000. PICCOLO, A.; MBAGWU, J. S. C. Role of hydrophobic components of soil organic matter in soil aggregate stability. Soil Science Society of America Journal, Madison, v. 63, p. 1081-1810, 1999. PICCOLO, M. C.; NEILL, C.; MELILLO, J. M.; CERRI, C. C.; STEUDLER, P. A. N-15 natural abundance in forest and pasture soils of the Brazilian Amazon Basin. Plant and Soil, The Hague, v. 182, p. 249-258, 1996. PINHEIRO, É. F. M.; LIMA, E.; CEDDIA, M. B.; URQUIAGA, S.; ALVES, B. J. R.; BODDEY, R. M. Impact of pre-harvest burning versus trash conservation on soil carbon and nitrogen stocks on a sugarcane plantation in the Brazilian Atlantic forest region. Plant and Soil, The Hague, v. 333, p. 71-80, 2010. RESENDE, A. S.; XAVIER, R. P.; OLIVEIRA, O. C. D.; URQUIAGA, S.; ALVES, B. J. R.; BODDEY, R. M. Long-term effects of preharvest burning and nitrogen and vinasse applications on yield of sugar cane and soil carbon and nitrogen stocks on a plantation in Pernambuco, N.E. Brazil. Plant and Soil, The Hague, v. 281, p. 339-351, 2006. RIFFALDI, R.; SCHNITZE, M. Electron-spin resonance spectrometry of humic substances. Soil Science Society of America Proceedings, Madison, v. 36, p. 301-305, 1972. RITCHEY, K. D.; SOUSA, D. M. G.; LOBATO, E.; CORREA, O. Calcium leaching to increase rooting depth in a Brazilian savannah Oxisol. Agronomy Journal, Madison, v. 72, p. 40-44, 1980. ROSA, A. H.; SIMÕES, M. L.; OLIVEIRA, L. C. de; ROCHA, J. C.; MARTIN-NETO, L.; MILORI, D. M. B. P. Multimethod study of degree humification of humic substances extracted from different tropic soils profiles in Brazil’s Amazonian region. Geoderma, Amsterdam, NL, v. 127, p. 1-10, 2005. SAAB, S. C.; MARTIN-NETO, L. Studies of semiquinone free radicals by ESR in the whole soil, HA, FA and humin substances. Journal Brazilian Chemical Society, São Paulo, v. 15, p. 34-37, 2004. SANO, E. E.; BARCELLOS, A. O.; BEZERRA, H. S. Assessing the spatial distribution of cultivated pastures in the Brazilian Savanna. Pasturas Tropicales, Cali, v. 22, p. 2-15, 2000. SCHNITZER, M.; LEVESQUE, M. Electron-spin resonance as a guide to the degree of humification of peats. Soil Science, Baltimore, v. 127, p. 140-145, 1979.

SEGNINI, A. Estrutura e reatividade da matéria orgânica em áreas com potencial para seqüestro de carbono no solo: estudos com métodos espectroscópicos. 2007. 131 f. Tese (Doutorado em Ciências) – Universidade de São Paulo, São Carlos. SEGNINI, A.; PRIMAVESI, O.; MILORI, D. M. B. P.; SILVA, W. T. L. da; SIMÕES, M. L.; MARTIN-NETO, L. Assessment of organic matter stability in Brazilian pastureland soils using spectroscopic methods. Soil Science Society of America Journal, Madison, 2011. No prelo. SENESI, N. Application of electron spin resonance (ESR) spectroscopy in soil chemistry. Advance Soil Science, New York, v. 14, p. 77-130, 1990a. SENESI, N. Molecular and quantitative aspects of the chemistry of fulvic-acid and its interactions with metal-ions and organic-chemicals: 1. the electron-spin-resonance approach. Analytica Chimica Acta, Amsterdam, NL, v. 232, p. 51-75, 1990b. SENESI, N. Molecular and quantitative aspects of the chemistry of fulvic acids and its interactions with metal ions and organic chemicals. Part II. The fluorescence spectroscopy approach. Analytica Chimica Acta, Amsterdam, NL, v. 232, p. 77-106, 1990c. SILVA, J. E, da; RESCK, D. V. S.; CORAZZA, E. J.; VIVALDI, L. Carbon storage in clayey Oxisol cultivated pastures in the “Cerrado” region, Brazil. Agriculture, Ecosystems and Environment, Amsterdam, NL, v. 103, p. 357-363, 2004. SILVA, J. E.; RESCK, D. V. S. Matéria orgânica do solo. In: VARGAS, M. A. T.; HUNGRIA, M. (Ed.). Biologia dos solos do cerrado. Planaltina: EMBRAPA-CPAC, 1997. p. 465-524. SILVA, R. M. da; MILORI, D. M. B. P.; FERREIRA, E. C.; FERREIRA, E. J.; KRUG, F. J.; MARTIN-NETO, L. Total carbon measurement in whole tropical soil sample. Spectrochimica Acta Part B, Oxford, v. 63, p. 1221-1224, 2008. SISTI, C. P. J.; SANTOS, H. P. D.; KOCHHANN, R. A.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M. Change in carbon and nitrogen stocks in soil under 13 years of conventional or zero tillage in southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 76, p. 39-58, 2004. SIX, J.; ELLIOTT, E. T.; PAUSTIAN, K. Aggregate and soil organic matter dynamics under conventional and no-tillage systems. Soil Science Society of America Journal, Madison, v. 63, p. 1350-1358, 1999. SIX, J.; ELLIOTT, E. T.; PAUSTIAN, K.; DORAN, J. W. Aggregation and soil organic matter accumulation in cultivated and native grassland soils. Soil Science Society of America Journal, Madison, v. 62, p. 1367-1377, 1998. SIX, J.; PAUSTIAN, K.; ELLIOTT, E. T.; COMBRINK, C. Soil structure and organic matter: I. distribution of aggregate-size classes and aggregate-associated carbon. Soil Science Society of America Journal, Madison, v. 64, p. 681-689, 2000. SMITH, P. Monitoring and verification of soil carbon changes under Article 3.4 of the Kyoto Protocol. Soil Use and Management, Oxford, v. 20, p. 264-270, 2004. STEVENSON, J. F. Humus chemistry: genesis, composition, reactions. 2nd ed. New York: John Wiley & Sons, 1994. 496 p. SUTTON, R.; SPOSITO, G. Molecular structure in soil humic substances: the new view. Environmental Science and Technology, Washington, DC, v. 39, p. 9009-9015, 2005. TALINEAU, J. C.; BONZON, B.; FILLONNEAU, C. ; HAINNAUX, G. Contribution à l’étude d’un agrosystème prairial dane le milieu tropical humide de la Côte d’Ivoire. 2. Analyse des données relatives à l’état de la matière organique. Cahier ORSTOM : série Pédologie, Paris, FR, v. 18, p. 29-48, 1980. TARRÉ, R. M.; MACEDO, R.; CANTARUTTI, R. B.; REZENDE, C. D. P.; PEREIRA, J. M.; FERREIRA, E.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M. The effect of the presence of a forage legume on nitrogen and carbon levels in soils under Brachiaria pastures in the Atlantic forest region of the South of Bahia, Brazil. Plant and Soil, The Hague, v. 234, p. 15-26, 2001. TEIXEIRA, M. A.; MURRAY, M. L.; CARVALHO, M. G. Assessment of land use and land use change and forestry (LULUCF) as CDM projects in Brazil. Ecological Economics, New York, v. 60, p. 260-270, 2006. TISDALL, J. M.; OADES, J. M. Organic matter and water stable aggregation in soils. Journal of Soil Science, Oxford, v. 33, p. 141-163, 1982.

VANDENBYGAART, A. J.; GREGORICH, E. G.; ANGERS, D. A. Influence of agricultural management on soil organic carbon: a compendium and assessment of Canadian studies. Canadian Journal of Soil Science, Ottawa, CA, v. 83, p. 363-380, 2003. WATSON, R. T.; NOBLE, I. R.; BOLIN, B.; RAVINDRANATH, N. H.; VERARDO, D. J.; DOKKEN, D. J. Land use, land-use change, and forestry: a special report of the IPCC. Cambridge: Cambridge University Press, 2000. p. 53-126. WEST, T. O.; POST, W. M. Soil organic carbon sequestration rates by tillage and crop rotation: a global data analysis. Soil Science Society of America Journal, Madison, v. 66, p. 1930-1946, 2002. WIKANDER, G.; NORDEN, B. Investigations of paramagnetic species in peat. Soil Science, Baltimore, v. 145, p. 289-297, 1988. ZOTARELLI, L.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M.; SIX, J. Impact of tillage and crop rotation on light fraction and intra-aggregate soil organic matter in two Oxisols. Soil and Tillage Research, Amsterdam, NL, v. 95, p. 196-206, 2007. ZOTARELLI, L.; ALVES, B. J. R.; URQUIAGA, S.; TORRES, E.; SANTOS, H. P. D.; PAUSTIAN, K.; BODDEY, R. M.; SIX, J. Impact of tillage and crop rotation on aggregate-associated carbon in two Oxisols. Soil Science Society of America Journal, Madison, v. 69, p. 482-491, 2005. ZSOLNAY, A.; BAIGAR, E.; JIMENEZ, M.; STEINWEG, B.; SACCOMANDI, F. Differentiating with fluorescence spectroscopy the sources of dissolved organic matter in soils subjected to drying. Chemosphere, Oxford, v. 38, p. 45-50, 1999.

Chapter 3

Carbon dynamics in a humid area of the Cerrado Maria Lucia Meirelles, Augusto Cesar Franco, Eloisa Aparecida Belleza Ferreira

Abstract: Carbon dynamics in a humid area of the Cerrado (humid grasslands) was studied, taking into account carbon stocks in phytomass and in soil, CO2 fluxes within the soil-plant-atmosphere system, the influence of groundwater level on these fluxes and the set of environmental factors affecting photosynthesis in the dominant herbaceous species in the area studied. From January 2005 to December 2006, data was obtained on the dry weight of above-ground phytomass and leaf area index (LAI), soil carbon stocks at seven depths up to 60 cm, on variables from a weather station installed in the area, on groundwater height, atmospheric CO2 fluxes through eddy covariance, soil CO2 fluxes using a portable system comprising a soil chamber coupled with an infrared analyzer, and photosynthetic standards using a portable system for measurements on leaves. It was observed that a large quantity of carbon was accumulated mainly in the soil, that the quantity of CO2 taken from the atmosphere during the day decreased and the soil’s CO2 flux increased as the groundwater level decreased in relation to the soil surface, and that there are differences among herbaceous species regarding the influence of groundwater height on the photosynthetic process, with the species covering the largest area (Axonopus comans) proving to be poorly sensitive to this influence. It was found that observed annual variations in rainfall altered atmospheric CO2 emission and uptake patterns, and that seasonal variation in groundwater height is a basic attribute for understanding and predicting patterns in carbon dynamics in the humid area studied. Keywords: carbon stock, photosynthesis, CO2 flux, humid grasslands.

Introduction Wetlands are natural ecosystems with a floodable substrate, a determining factor in the origin and development of the soil and of plant and animal communities. They show high primary production, and dead matter decomposes slowly by anaerobiosis due to soil flooding, which leads to a substantial accumulation of organic matter (YAVITT, 1994). Even though they only occupy approximately 2% of the world’s area, it is estimated that wetlands contain from 10% to 14% of the world’s total accumulated organic carbon (ARMENTANO, 1980). In the Cerrado biome, various wetland phytophysiognomies occur, which are, in general, transition ecosystems with great input and output of matter and energy in relation to adjacent systems, such as watercourses and savanna and forest formations, and which function as water storage facilities for the region’s waterways. Humid grassland is one of these phytophysiognomies and usually occurs in relatively large, flat areas, adjacent to waterways and periodically flooded, with a continuous herbaceous stratum without any trees or shrubs (RIBEIRO, WALTER, 2008). Due to the Cerrado’s rainy and dry seasons, in the humid grassland there is a variation in the height of the groundwater sheet in relation to the surface, with periods of anaerobiosis in the soil, which slow down the decomposition of organic matter accumulated in the soil. These soils, when drained for agricultural use, undergo significant and continuous changes that, through gradual oxidation, cause the gradual loss of organic matter (MIRANDA, 1990). Vast areas of humid grassland in the Cerrado have already been drained, or have suffered other anthropic disturbances, and as they are

Permanent Protection Areas, they need to be rehabilitated. This rehabilitation requires knowledge of the original structure and functioning of this ecosystem. Periodic flooding events hamper establishment of arbustive and arboreal species in the humid grasslands. When anthropic disturbances cause a significant lowering of the groundwater, in these areas there is a colonization by native tree and shrub species, such as Trembleya parviflora and Vochysia pyramidales, which begin to cast shade over the herbaceous stratum, thus initiating a reduction in the frequency and coverage of herbaceous species adapted to a higher degree of flooding (MEIRELLES et al. 2004). Soon after, local disturbances (grazing, drainage, deforestation) and land use change in terms of Hydrographic Basin level, causing a reduction in groundwater depth, lead to the destruction of the Cerrado humid grassland, entailing loss of biodiversity and changes in the carbon cycle. To extend the knowledge of its ecological processes, the carbon dynamics of a humid area of the Cerrado (humid grasslands) will be presented here, taking into account carbon stocks in phytomass and in soil, CO2 fluxes within the soil-plant-atmosphere system, influence of the groundwater level on these fluxes and the set of environmental factors affecting photosynthesis in the dominant herbaceous species (Figure 1). Data was obtained simultaneously, through the use of non-destructive methodologies, with measurements performed in situ. Density of the atmospheric CO2 flux was quantified using the micrometeorological eddy covariance method, and soil carbon was analyzed using a soil chamber connected to an infrared gas analyzer (IRGA). Photosynthetic processes were studied using a portable device. The environmental impact caused by data collection in the area was minimal. This was necessary because the humid grassland is an easily degraded, low-resilience native ecosystem. The information obtained clarified important processes in the carbon dynamics of a wetland that accumulates a large amount of carbon in the soil and that includes native species adapted to significant variations in the groundwater height.

CHAP 3 - FIGURE 1 Fluxo de Co2 Vegetação-Atmosfera

Vegetation-Atmosphere Co2 flux

Biomassa aérea

Above-ground biomass

Fatores reguladores da fotossíntese

Photosynthesis regulating factors

Altura da lâmina do lençol freático

Groundwater level

Fluxo de Co2 do solo

Co2 flux from soil

Matéria orgânica

Organic matter

Figure 1. Variables considered in a study on carbon dynamics in a humid area of the Cerrado (Brasília, Distrito Federal (DF)). Illustration: Wellington Cavalcanti Source: Adapted from Meirelles et al. (2006).

Area studied

The wetland studied has approximately 16 ha and is located at an altitude of 1,060 m, between 15° 55' 31.3" and 15° 55' 45.5" S, and 47° 54' 23.3" and 47° 54' 17.1" W, on the Água Limpa Farm (FAL) [Fazenda Água Limpa], an experimental area at the University of Brasília. The climate, according to Köppen’s classification, is of type AW (tropical climate with rains in summer and drought in winter), generally with dry season from May to September and rainy season from October to April. The predominant soil type is Organosol [US: Histosol], with pH in H2O around 6.1, density about 0.5 g cm−3 and saturated hydraulic conductivity of 13.1 cm h−1. Data was collected from January 2005 to December 2006. The last fire in the area had occurred in August 1999. The phytophysiognomy corresponds to Cerrado humid grassland, occurring in poorly drained land, with no trees or shrubs, continuous herbaceous stratum, crown approximately 70 cm high and with a predominance of species of the Poaceae family (Table 1). The area has a 3% declivity from the Cerrado strictly speaking toward the riparian forest, the two phytophysiognomies bordering the humid grasslands (Figure 2). A rainfall sensor (TR525I, Texas Electronics Inc., Dallas, Texas, USA) was installed, connected to a data logger (CR23X, Campbell Scientific Inc., Logan, Utah, USA). Figure 3a shows the monthly rainfall data for the years studied (2005 and 2006) and the historical average values (1965 through 2005) from the Institute of Meteorology’s Brasília Weather Station. Total rainfall was 1,391 mm in 2005 and 1,829 mm in 2006, and the historical average was 1,308 mm. Table 1. Species with greater relative coverage (RC) in a humid grassland area (Brasília, Distrito Federal (DF)). Species Axonopus comans (Trin.) Henrad

Family Poaceae

RC (%) 42.11

Andropogon lateralis Nees subsp. cryptopus

Poaceae

9.29

Andropogon bicornis L.

Poaceae

5.68

Andropogon virgatus Desv.

Poaceae

5.13

Paspalum polyphyllum Nees

Poaceae

3.18

Lagenocarpus rigidus Nees

Cyperaceae

2.66

Eupatorium vindex DC.

Asteraceae

2.59

Hyptis carpinifolia Benth.

Lamiaceae

2.59

Ctenium cf. brachystachyum (Nees) Kunth

Poaceae

2.50

Echinolaena inflexa (Poir.) Chase

Poaceae

2.37

Source: Munhoz (2003).

CHAP 3 - FIGURE 2 (sensu stricto)

(strictly speaking)

Mata de galeria

Riparian forest

Foto

Photo

Cerrado limpo húmido

Cerrado humid grassland

Figure 2. Humid grassland bordered by riparian forest and Cerrado strictly speaking in the Água Limpa Farm, of the University of Brasília (Brasília, Distrito Federal (DF)).

A 90 m transect was delimited in the humid grassland area using a nylon line to monitor groundwater height. Stakes were placed every 10 meters, thus marking 9 points in the transect, with point 1 near the edge of the Cerrado strictly speaking and point 9 on the edge of the riparian forest. A 2 m PVC tube was buried at each point and the distance between the soil surface and the groundwater sheet was monitored every month. Negative values indicate how far the groundwater sheet was below the surface, and positive values indicate the height of the sheet of water above the soil surface (Figure 3b).

CHAP 3 - FIGURE 3 Precipitação

Rainfall

Lençol freático

Groundwater sheet

Média

Average

Mês

Month

Figure 3. a) Total monthly rainfall (mm) in a Cerrado humid grassland, Brasília, Distrito Federal (DF) (2005 and 2006) and historical average (1965 through 2005) of the Brasília weather station (Institute of Meteorology); b) Monthly mean distance between groundwater sheet and soil surface (cm) in a humid grassland area (Brasília, Distrito Federal (DF)) in 2005 and 2006. Source: Meirelles et al. (2010).

Carbon stocks Phytomass On August 30, 2005 and November 28, 2006 leaves on 1 m2 plots were cut at ground level, with 8 replicates, next to monitoring points 3, 4, 5 and 6. After collection, living and dead biomass were separated in each sample. The leaf area of the living biomass was obtained using a planimeter to calculate the leaf area index (LAI), which expresses the amount of leaf area (m2) per unit of ground surface area (m2). After drying in an oven at 80 °C for 48 hours, the dead and living biomass were weighed to obtain dry weight. The mean values of living and dead above-ground biomass (Table 2) were not significantly different for the samples collected in August 2005 and November 2006. The estimated living biomass leaf area index was approximately 1.3. −2 Table 2. Dry weight (mean and standard deviation) of above-ground living and dead plant biomass (g m ) and leaf area index (LAI) obtained in August 2005 and November 2006 in a Cerrado humid grassland (Brasília, Distrito Federal (DF)).

Aug. 2005

Dead phytomass (g m−2) 594 ± 206

Living phytomass (g m−2) 262 ± 37

1.30

Nov. 2006

504 ± 37

310 ± 53

1.34

LAI

Carbon in soil 9 points were determined for soil sampling, close to the places where the PVC tubes used for reading the height of the groundwater sheet were installed. Each sample point was considered a replicate. Each soil sample was composed of 10 simple samples per replicate and per depth, collected within a radius of 1.5 meters around each sample point. Deformed soil samples were collected with a Dutch-type auger, and non-deformed samples were collected with volumetric ring, in August 2005, at 7 different depths (0 cm to 5 cm; 5 cm to 10 cm; 10 cm to 20 cm; 20 cm to 30 cm; 30 cm to 40 cm; 40 cm to 50 cm and 50 cm to 60 cm). The soil’s organic carbon content was evaluated using the wet oxidation method (WALKLEY; BLACK, 1934). The soil’s organic carbon content decreased with depth (Figure 4) and ranged from 17.22 dag kg−1 in the 0 cm to 5 cm layer to 1.25 dag kg−1 in the 50 cm to 60 cm layer. For the average of all points, the linear model showed adequate fit for the decrease in carbon content at greater depths (y = −1.4575x + 11.998; R2 = 0.95). Considering a mean soil density of approximately 0.5, the estimated organic carbon stock to a depth of 60 cm was 241 t ha−1 (Table 3).

CHAP 3 - FIGURE 4 Carbono orgânico

Organic carbon

Profundidade

Depth

Figure 4. Organic carbon in soil at various depths along the transect established in a humid grassland of the Cerrado (Brasília, Distrito Federal (DF)). Source: Adapted from Meirelles et al. (2006).

Table 3. Estimated organic carbon stock of soil at depths up to 60 cm in a humid grassland of the Cerrado (Brasília, Distrito Federal (DF)). Organic C stock Depth

Organic C (kg ha−1)

0 cm – 5 cm

36,596

5 cm – 10 cm

30,570

10 cm – 20 cm

56,600

20 cm – 30 cm

54,461

30 cm – 40 cm

30,633

40 cm – 50 cm

20,857

50 cm – 60 cm

11,539

Total (0 cm – 60 cm)

241,257

Source: Meirelles et al. (2006).

CO2 flux Atmospheric CO2 flux

The energy balance equation for a given surface primarily relates the following flux densities (OKE, 1987): Rn = LE + H + G + Fc where Rn is net radiation, LE is latent heat, H is sensible heat in the air, G is soil heat and Fc is photochemical energy. An automated micrometeorological station and a system to use the eddy covariance technique (SWINBANK, 1951) were installed in the humid grassland to obtain LE, H and Fc. The automated micrometeorological station included the following sensors, with respective model and brand: wind speed (014A – Met One) and direction (024A – Met One); net radiation (Q7 – Rebs); photosynthetic active radiation (LI190 – LI-COR); incoming and reflected solar radiation (CM3 – Kipp & Zonen); air temperature and humidity (HMP45C – Vaisala); soil heat (HFT3 – Rebs); soil temperature (Tcav – Campbell) and humidity (CS615 – Campbell). These sensors were connected to a data acquisition system (CR23X – Campbell) containing the management software, with data being collected every minute, then collected weekly using a storage module (SM4M – Campbell) and storing 30 minute averages. Rn was obtained using the Q7, and the values were corrected based on wind speed. G was calculated from the average of the 2 soil heat plates buried 8 cm deep plus the energy stored in the soil (S) obtained by: ????? in which Cv is the specific heat of humid soil, and ΔT / Δt is the mean variation in soil temperature, from sensors installed at depths of 2 cm and 6 cm. The water contents used for the calculation of Cv were those obtained at 3 cm depth. The eddy covariance system (Figure 5) had a three-dimensional sonic anemometer (CSAT3, Campbell Scientific Inc., Logan, Utah, USA) and a fast response IRGA (LI-7500, LI-COR, Lincoln, Nebraska, USA), measuring water vapor and CO2 concentrations. The devices were connected to a data logger (CR5000 – Campbell), which stored the data obtained at a frequency of 20 Hz. To obtain the turbulent flux densities, the Eddy3 software program was used, developed by the Alterra micrometeorology team (Netherlands), which, for each 30 minutes, calculates H, LE and Fc from the average of values obtained by multiplying fluctuations in wind speed with temperature, concentration of water vapor and CO2, respectively, making all the necessary corrections in the data.

CHAP 3 - FIGURE 5 Figure 5. Eddy covariance system in a humid grassland of the Cerrado (Brasília, Distrito Federal (DF)), including a three-dimensional sonic anemometer (CSAT3, Campbell Scientific Inc., Logan, UT, USA) and a fast response analyzer to measure CO2 and water vapor concentrations (LI-7500, LI-COR, Lincoln, Nebraska, USA).

Atmospheric CO2 exchanges of a wetland are known to be affected by environmental variations in the ecosystem (BONNEVILLE et al., 2008). In the daytime (7h to 18h), the total carbon removed from the atmosphere through the CO2 flux in humid grassland varied over the course of the months in 2005 and 2006 (Figure 6). The total annual value was 653 g C m−2 for 2005 and 541 g C m−2 for 2006 (MEIRELLES et al., 2010). A significant correlation was observed between the

monthly amount of carbon removed from the atmosphere during the day and the monthly mean distance between groundwater sheet and soil surface in 2005 (r = 0.71, p ≤ 0.005) and 2006 (r = 0.51, p ≤ 0.05). Therefore, as the sheet of water moves further below the soil surface, there is a decrease in the humid grassland’s capacity for removing CO2 from the atmosphere during daytime, when plants are performing photosynthesis.

CHAP 3 - FIGURE 6 Fluxo de CO2

CO2 flux

Distância

Distance

Catm

AtmC

Lençol

Groundwater sheet

Fev.

Feb.

Abr.

Apr.

Maio

May

Ago.

Aug.

Set.

Sep.

Out.

Oct.

Dez.

Dec.

Mês

Month

−2 Figure 6. Monthly total carbon (g C m ) of atmospheric CO2 flux during the day for a humid grassland in the Cerrado (Brasília, Distrito Federal (DF)) and monthly distance between the groundwater sheet and the soil surface (cm) in 2005 and 2006.

Source: Adapted from Meirelles et al. (2010).

Soil CO2 flux The soil’s respiration reveals all its biological activity, including plant roots, macroorganisms (such as earthworms, nematodes and insects) and microorganisms. This microbial respiration is characterized by the production of CO2 as a result of its metabolism, and it is dependent not only on density of microorganisms, but also on their metabolic condition, which, in turn, depends on the physical and chemical conditions of the soil, such as temperature, porosity, water content, pH, etc. The level of the groundwater sheet has important effects on CO2 emissions in humid areas, since soil saturation limits the diffusion of atmospheric oxygen, reducing microbial activity and lowering the decomposition rate (CHIMNER; COOPER, 2003). Considering that aerobic respiration is more efficient in producing CO2 than anaerobic respiration, a decline in the groundwater sheet level in relation to the surface increases the diffusion of oxygen in the soil, allowing for aerobic decomposition, which increases CO2 emissions (SILVOLA et al. 1996). Roots are another major source of oxidizable carbon and they are mainly present in upper peat layers (CHIMNER, 2000 cited by CHIMNER; COOPER, 2003). Roots produce exudates and rootlets which can be quickly decomposed by microorganisms when

exposed to aerobic conditions (THOMAS et al., 1996), in addition to releasing CO2 in their own respiration (SILVOLA et al., 1996; VERVILLE et al., 1998). Quantification of the density of the CO2 flux in the soil surface was obtained using an infrared gas analyzer (Figure 7), which calculates soil respiration from rates of increase of surface CO2 concentration and provides fast and reliable estimates (HAYNES; GOWER, 1995). In this method, the chamber has little influence on microclimate variables, as flux measurements during relatively short periods (2–3 minutes) minimize the effects of air and soil temperature and the increase in the water content and CO2 concentration inside the chamber. A portable device was used (Model LI6400, LI-COR, Lincoln, Nebraska, USA), equipped with a respiration chamber in which the flux of CO2 is retained, and an automatic system triggers the infrared gas analyzer, connected to a digital data storage system (JENSEN; MUELLER, 1996; NORMAN et al., 1997).

CHAP 3 - FIGURE 7 Fotos

Photos

Câmara de respiração do solo

Soil chamber

Sistema digital de armazenamento de dados

Digital data storage system

Figure 7. a) Data collection on a humid grassland in the Cerrado (Brasília, Distrito Federal (DF)) using the infrared gas analyzer (LI-6400, LI-COR, Lincoln, Nebraska, USA), coupled to a soil chamber; b) LI-6400 components used to measure densities of the soil’s CO2 flux.

To determine the CO2 flux, the respiration chamber is carefully placed in a PVC ring (10.5 cm inner diameter × 5 cm height) permanently inserted 2 cm into the soil. The PVC ring was used to avoid any mechanical disturbance at the time of installation. The CO2 concentration inside the chamber is then adjusted by means of a suction pump to levels below the concentration in the ambient atmosphere. The system is shut off and the flux of accumulated CO2 is measured for 2 to 3 minutes, 3 times. The CO2 flux was calculated by interpolating the values of the 10 mg CO2 L−1 increase within the chamber and the measurement time, taking into account the internal volume of the respiration chamber and the previously measured environmental CO2 concentration. 9 PVC rings were installed in April 2005, adjacent to the 9 points where the level of the groundwater sheet was being monitored. CO2 flux measurements were performed between May 2005 and December 2006, at intervals of approximately one week. The annual quantity of carbon released in the soil’s CO2 flux in the humid grassland in the Cerrado was estimated. Mean monthly flux (Fm) and standard deviation were calculated for the 4 monthly measurements, then summing the values for the months of 2005 and 2006. The total annual flux estimated for the period of May to December of 2005 was 2.69 t C-CO2 ha−1 and for the total year of 2006, 2.38 t C-CO2 ha−1. The mean monthly soil surface CO2 flux for the 9 points studied (Fig. 8) varied from 0.26 μmol C-CO2 m−2 s−1 in April 2006 (c.v. = 194%) to 5.21 μmol C-CO2 m−2 s−1 in November 2005 (c.v. = 39%); the mean annual flux was approximately 1.46 μmol C-CO2 m−2 s−1 with high variability (c.v. = 114%). There was a strong seasonal trend, with the highest rates of CO2 emission by soil occurring at the end of the 2005 dry period (Figure 8), when the groundwater level was below the soil surface

(Figure 3b), indicating a significant correlation between the soil’s monthly CO2 flux and the distance between the groundwater sheet and the soil surface, which explained 89% of variations for 2005, and 88% for 2006 (p = 0.005).

Influence of environmental factors on photosynthesis Any plant’s performance and its response to environmental variations largely depends on the structure and functioning of the photosynthetic apparatus and its plasticity. In tropical savanna regions, wetlands are exposed to strong seasonality in water availability. The soil’s water saturation in the rainy season results in anoxic or hypoxic conditions due to a high consumption of oxygen by respiration of roots and microorganisms, and insufficient diffusion of oxygen in the water and submerged tissues (CHIMNER; COOPER, 2003; CRAWFORD, 1992), and accumulation of a variety of potentially toxic compounds (FERREIRA et al., 2009). On the other hand, herbaceous plants may be exposed to hydric deficiency during the dry season, when they are generally exposed to high levels of solar radiation and a large range of variation in light intensity over a diurnal cycle. Under these conditions, performance of the photosynthetic apparatus depends strongly on its ability to adjust to a variety of potentially stressful conditions. To examine the photosynthetic function’s capacity to adjust to environmental seasonality, the response in CO2 assimilation to variations in light intensity throughout the year was determined, in two grasses that are representative of the humid grassland studied, Axonopus comans and Andropogon virgatus, which together account for almost 50% of coverage on the area (Table 1).

CHAP 3 - FIGURE 8 Média mensal do fluxo de CO2 do solo

Mean monthly soil CO2 flux density

Fev.

Feb.

Abr.

Apr.

Maio

May

Ago.

Aug.

Set.

Sep.

Out.

Oct.

Dez.

Dec.

Figure 8. Mean monthly soil CO2 flux density (µmol C-CO2 m−2 s−1) from May 2005 to December 2006 in a humid grassland of the Cerrado (Brasília, Distrito Federal (DF)).

A portable system was used to measure photosynthesis and transpiration – LCpro, from ADC BioScientific Ltd. (Hoddesdon, England) (Figure 9). Variations in light intensity were obtained with the device’s lighting system. It was attempted to maintain a relatively constant air temperature in the chamber of the device, ranging from 25 °C to 27 °C. Between 10 to 40 leaves were placed inside the device chamber, making sure that the entire area of the chamber was filled. At each luminous intensity, CO2 assimilation values were monitored for 5 to 10 minutes until stabilization. To

determine respiration values, the light source was turned off, the chamber of the device was covered with tin foil, and CO2 release values were monitored until they stabilized.

CHAP 3 - FIGURE 9 Foto

Photo

Figure 9. Data collection on photosynthesis and transpiration parameters in a humid grassland of the Cerrado (Brasília, Brazil), using the LCpro portable system (ADC, Hoddesdon, England).

In both species studied, CO2 assimilation showed no saturation, even at maximum photon flux density values in the photosynthetically active range (Figures 10 and 11). This lack of saturation indicates that both grasses are C4, which was confirmed by values of leaf carbon isotope composition, which were -13.54‰ in A. comans and -12.56‰ in A. virgatus. Mean leaf respiration values were 0.34 μmol m−2 s−1 for A. virgatus and -0.57 μmol m−2 s−1 for A. comans. The two species showed contrasting behaviors in response to seasonality in soil water conditions. Seasonal drought and flooding of the soil’s surface did not have a significant effect on carbon assimilation capacity of A. comans (Figure 11), the species with higher coverage (42%) in the area studied (Table 1). Conversely, smaller CO2 assimilation values were measured in A. virgatus in March 2005 and April 2006 (Figure 10), at the end of the region’s rainy season, in which the soil surface was covered by the sheet of water (Figure 3b).


CO2 flux

8 de março de 2005

March 8, 2005

14 de junho 2005

June 14, 2005

16 de agosto 2005

August 16, 2005

20 de Outubro de 2005

October 20, 2005

4 de abril de 2006

April 4, 2006

Densidade de fluxo de fótons

Photon flux density

Figure 10. CO2 assimilation in response to variations in photon flux density in Andropogon virgatus. Data obtained in a humid grassland area at the University of Brasília experimental farm Água Limpa Farm (FAL) [Fazenda Água Limpa], at various times of the year. Source: Meirelles et al. (2010).


CO2 flux

8 de março de 2005

March 8, 2005

14 de junho 2005

June 14, 2005

16 de agosto 2005

August 16, 2005

20 de Outubro de 2005

October 20, 2005

4 de abril de 2006

April 4, 2006

Densidade de fluxo de fótons

Photon flux density

Figure 11. CO2 assimilation in response to variations in photon flux density in Axonopus comans. Data obtained in a humid grassland area at the University of Brasília experimental farm Água Limpa Farm (FAL) [Fazenda Água Limpa], at various times of the year . Source: Meirelles et al. (2010).

Final considerations The humid grasslands of the Cerrado have characteristics of acting as water storage area for the region’s waterways, showing low resilience to environmental impacts, currently presenting vast degraded areas, acting as a major carbon collector and showing vegetation adapted to a large range of variations in the height of the groundwater sheet. Regarding carbon dynamics in a humid grassland of the Cerrado, it was observed that a large quantity of carbon is accumulated mainly in the soil, that the quantity of CO2 taken from the atmosphere during the day decreased and the soil’s CO2 flux increased as the level of the groundwater sheet decreased in relation to the soil surface, and that there are differences among herbaceous species regarding the influence of groundwater height on the photosynthetic process, with the species covering the largest area (Axonopus comans) proving to be poorly sensitive to this influence. It was found that seasonal variations in groundwater height are a basic attribute for understanding and predicting patterns in the carbon cycle and that annual variations in rainfall, such as those of 2005 and 2006, alter atmospheric CO2 emission and sink patterns. Therefore, it is believed that the information presented here might justify activities of conservation, restoration and compensation for environmental services provided by the Cerrado humid grasslands.

References ARMENTANO, T. V. Drainage of organic soils as a factor in the world carbon cycle. BioScience, Washington, DC, v. 30, 825830, 1980. BONNEVILLE, M.-C.; STRACHAN, I. B.; HUMPHREYS, E. R.; ROULET, N. T. Net ecosystem CO2 exchange in a temperate cattail marsh in relation to biophysical properties. Agricultural and Forest Meteorology, Amsterdam, NL, v. 148, p. 69-81, 2008. CHIMNER, R. A.; COOPER, D. J. Influence of water table levels on CO2 emissions in a Colorado subalpine fen: an in situ microcosm study. Soil Biology and Biochemistry, Oxford, v. 35, p. 345–351, 2003. CRAWFORD, R. M. M. Oxygen availability as an ecological limit to plant distribution. In: BERGON, M.; FITTER, A. H. (Ed.). Advances in ecological research. London, GB: Academic, 1992. p. 93–185. FERREIRA, C. S; PIEDADE, M. T. F.; FRANCO, A. C.; GONÇALVES, J. F. C.; JUNK, W. J. Adaptive strategies to tolerate prolonged flooding in seedlings of floodplain and upland populations of Himatanthus sucuuba, a Central Amazon tree. Aquatic Botany, Amsterdam, NL, v. 90, p. 246–252, 2009. HAYNES, B. E.; GOWER, S. T. Belowground carbon allocation in unfertilized and fertilized red pine plantations in northern Wisconsin. Tree Physiology, Victoria, v. 15, p. 317-325, 1995.

JENSEN, L. S.; MUELLER, T.; TATE, K. R.; ROSS, D. J.; MAGID, J.; NIELSEN, N. E. Soil surface CO2 flux as an index of soil respiration in situ: a comparison of two chamber methods. Soil Biology and Biochemistry, Oxford, v. 29, p. 1297-1306, 1996. MEIRELLES, M. L.; FRANCO, A. C.; FERREIRA, E. A. Dinâmica sazonal do carbono em Campo Úmido do Cerrado. Planaltina: Embrapa Cerrados, 2006. 32 p. (Embrapa Cerrados. Documentos, 164). MEIRELLES, M. L.; FRANCO, A. C.; FERREIRA, E. A.; RANDOW, C. van. Fluxo diurno de CO2 solo-planta-atmosfera em um Campo Úmido do Cerrado. Planaltina: Embrapa Cerrados, 2010. 22 p. (Embrapa Cerrados. Boletim de Pesquisa e Desenvolvimento, 274). MEIRELLES, M. L.; GUIMARÃES, A. J. M.; OLIVEIRA, R. C. de; GLEIN, M. A.; RIBEIRO, J. F. Impactos sobre o estrato herbáceo de Áreas Úmidas do Cerrado. In: AGUIAR, L. M. de; CAMARGO, A. J. A. de. (Ed.). Cerrado: ecologia e caracterização. Planaltina: Embrapa Cerrados, 2004. p. 41-68. MIRANDA, L. N. Prioridades e metodologias de pesquisa em várzeas na área de fertilidade do solo. Planaltina: EMBRAPA-CPAC, 1990. 17 p. (EMBRAPA-CPAC. Documentos, 33). MUNHOZ, C. B. R. Padrões de distribuição sazonal e espacial das espécies do estrato herbáceo-subarbustivo em comunidade de Campo Limpo Úmido e de Campo Sujo. 2003. 273 f. Tese (Doutorado em Ecologia) - Universidade de Brasília, Brasília, DF. NORMAN, J. M.; KUCHARIK, C. J.; GOWER, S. T.; BALDOCCHI, D. D.; CRILL, P. M.; RAYMENT, M.; SAVAGE, K.; STRIEGL, R. G. A comparison of six methods for measuring soil-surface carbon dioxide fluxes. Journal of Geophysical Research, Washington, DC, v. 102, p. 28771-28777, 1997. OKE, T. R. Boundary layer climates. 2nd ed. London, GB: Routledge, 1987. 435 p. RIBEIRO, J. F.; WALTER, B. M. T. As principais fitofisionomias do bioma Cerrado. In: SANO, S. M.; ALMEIDA, S. P. de.; RIBEIRO, J. F. (Ed.). Cerrado: ecologia e flora. Brasília, DF: Embrapa Informação Tecnológica; Planaltina: Embrapa Cerrados, 2008. p. 151-212. SILVOLA, J.; ALM, J.; AHLHOLM, U.; NYKANEN, H.; MARTIKAINEN, P. J. CO2 fluxes from peat in boreal mires under varying temperature and moisture conditions. Journal of Ecology, Oxford, v. 84, p. 219–228, 1996. SWINBANK, W. C. The measurement of vertical transfer of heat and water vapor by eddies in the lower atmosphere. Journal of Meteorology, Boston, v. 8, p. 135-145, 1951. THOMAS, K. L.; BENSTEAD, J.; DAVIES, K. L.; LLOYD, D. Role of wetland plants in the diurnal control of CH4 and CO2 fluxes in peat. Soil Biology and Biochemistry, Oxford, v. 28, p. 17–23, 1996. VERVILLE, J. H.; HOBBIE, S. E.; CHAMPIN III, F. S.; HOOPER, D. U. Response of tundra CH4 and CO2 flux to manipulation of temperature and vegetation. Biogeochemistry, Amsterdam, NL, v. 41, p. 215–235, 1998. WALKLEY, A.; BLACK, I. A. An examination of the Degtyareff method for determining soil organic matter and proposed modification of the chromic titration method. Soil Science, Baltimore, v. 37, p. 29-38, 1934. YAVITT, J. B. Carbon dynamics in Appalachian peatlands of west Virginia and western Maryland. Water, Air and Soil Pollution, Dordrecht, v. 77, p. 271-290, 1994.

Chapter 4

Biomass stock in planted forests, agroforestry systems, secondary forests and Caatinga Rosana Clara Victoria Higa, Haron Abrahim Magalhães Xaud, Luciano Jose de Oliveira Accioly, Roberval Monteiro Bezerra de Lima, Steel Silva Vasconcelos, Vanda Gorete Souza Rodrigues, Claudio José Reis de Carvalho, Cintia Rodrigues de Souza, Francisco das Chagas Leonidas, Helio Tonini, João Baptista Silva Ferraz, Maristela Ramalho Xaud, Moisés C. Mourão de Oliveira Junior, Rogério Sebastião Correa da Costa

Abstract: This chapter presents the main results and discusses biomass stock in various types of land use and in the Caatinga biome. Due to the diversity of environments and agroecosystems studied, the methodology used to determinate and estimate biomass was adapted to each case. Several techniques were used, including destructive plant material collection techniques and indirect ones, based on use of allometric equations and remote sensing. This additional knowledge on accumulation of biomass in the Caatinga and in tree species in various planting systems can contribute to the advancement of carbon balance estimates in Brazil and to the formulation of measures to mitigate and compensate greenhouse gas emissions, in order to combat global warming and maintain sustainability. Keywords: carbon, phytomass, land use, forestry.

Introduction Forests are known to be important carbon sinks (MCCARTHY et al., 2001; SOLOMON et al., 2007). Globally, it is estimated that forests store about 283 Gt of C in biomass alone.2 Recent data on deforestation and on maintenance and establishment of new forest plantation areas in the world shows that carbon stocks in forest biomass has decreased in Africa, Asia and South America from 1990 to 2005, but has increased in other parts of the world (FAO, 2005). The carbon stock is dependent on the forest’s total biomass production, which, in turn, is the result of the difference between production through photosynthesis and consumption through respiration and harvest processes (BROWN, 1997). Therefore, forest biomass is relevant in terms of the global climate change issue, because it is the main indicator of carbon stocks. Biomass is also directly related to the quantity of carbon that can potentially be emitted when a forest is burned down or logged / cut down (BROWN, 1997; MARTINELLI et al., 1994). Most data available on forest biomass concerns ecological studies, which, in most cases, are done to characterize regionalized forest structures, with reduced samples, which are insufficient for analyses on a global scale (BROWN, 1997; SALAZAR, 1999). These studies are typically conducted in natural forests, where species diversity and heterogeneity of ages, shapes and sizes of tree individuals pose major difficulties for destructive sampling work.

2

The term biomass has various conceptualizations in the scientific literature. In this study, we present the concept of biomass used by Brown (1997), which corresponds to dry biomass contained in the above-ground part of plant individuals in a given area. Generally, in forest areas, the term only includes tree individuals, whereas in formations like the Caatinga, savanna, grasslands or fields, arbustive and arboreal components may be included, in accordance with suitable characterization of the vegetation.

In forest plantations, especially where the degree of improvement is more advanced, the assessment of biomass, apart from having a lesser degree of difficulty, is more accurate due to less variability between individuals. In the last decade, Brazil has developed a large number of studies on biomass in planted forests, mainly for the Eucalyptus and Pinus genera (REIS et al., 1994). There are still, however, many information gaps, such as planted area, species, age, biomass carbon contents, etc. It is necessary to advance in studies of forest biomass, but it is known that direct measurements of biomass, involving destructive sampling, are laborious and expensive. Thus, there is a need to develop methodologies having a low degree of difficulty in execution and low cost (BOOTH, 2003; BROWN, 2002). In this respect, methodologies based on remote sensing, with potential application in large continuous areas, provided that they generate estimates with an adequate degree of certainty, will facilitate a better approximation to current calculated biomass values, proving useful for refining calculations aimed at integrating models of greenhouse gas emissions (BROWN, 2002). The total biomass accumulation on the above-ground parts, as well as their components (trunk, branches and leaves), can vary greatly according to species and soil and climate conditions of the location (REIS et al., 1994). Comparative studies analyzing variations in tree growth under various treatments (e.g., spacing, fertilization, thinning) in various locations are essential to improve the understanding of biological growth potential, development strategies and responses to management actions (ADEGBIDI et al., 2002, 2004). Martin and Jokela (2004) report that a large number of studies on processes have been conducted to understand the production ecology and ecophysiology of Pinus taeda. The impacts of intensive forestry on processes that affect productivity, however, have yet to be fully understood for this species, due to the lack of adequate testing throughout the entire cycle of the crop. The same authors, analyzing the production dynamics of Pinus taeda under various environmental conditions and various forestry practices, concluded that decreased growth rate is caused by the following potential factors: changes in carbon allocation from above-ground parts to roots (especially fine root production); increase in respiration rate in relation to photosynthesis in older trees, which reduces the carbon available for growth; limited expansion of the root system due to physical constraints; and decrease in the carbon accumulation rate in crowns, due to shading of lower parts by upper parts. Such studies are essential for future mitigation measures in the forest sector. Despite the many research efforts for quantification of carbon stored in Brazilian ecosystems and agroecosystems and the considerable amount of data available, there are still research-related problems in the calculation of greenhouse gas emissions: 1) Problems of a scientific nature: a) low precision in estimates of each ecosystem’s original area; b) low precision in estimates of total area for the most significant types of agroecosystems; c) wide variation in methodologies used for estimates, ranging from forest inventories to sampling for direct determination of phytomass; d) sparse network of sampling points in continental extension areas such as the Amazon; 2) Problems of an organizational/political/structural nature: a) insufficient interest in long-term research; b) low number of researchers in Brazil dedicated to studies on the carbon cycle or dynamics. The result of these problems is gaps still being filled in to refine estimates of Brazilian carbon stocks and dynamics. In this context, the Agrogases network, by disseminating the research results included in this chapter, aims to contribute to deepening the knowledge on carbon stocks in certain types of

natural vegetation formations (Caatinga), secondary forests (in the Amazon), planted forests (exotic and native species) and agroforestry systems.

Methodology To quantify and estimate the above-ground biomass in forest ecosystems and agroecosystems, both direct and indirect methods were used. The methodology is described first in general terms. Then, at the beginning of each specific topic (Caatinga, planted forests, agroforestry systems and secondary forests) there is a description of the relevant methodological details.

Direct method Direct or destructive methods involve quantification of biomass by cutting, collecting and weighing various tree components (usually trunks, branches and leaves). In experimental plots varying in number and size, forest inventories were made to measure dendrometric parameters: diameter at 1.30 m above the ground (diameter at breast height – DBH) and height. Afterward, trees were selected that were representative of the population, which were cut down for new collection of dendrometric data (DBH, total height and green weight) and separated into various components, where subsamples were taken and subsequently weighed, identified and sent to the a laboratory for drying.

Indirect method Allometric equations Indirect methods involve applying allometric equations that generally relate biomass with measurements such as: DBH, height and wood density (BROWN et al., 1989; HIGUCHI, 1994; HIGUCHI et al., 1998), as well as using remote sensing tools (BROWN, 2002). Allometric equations can be developed for species, groups of species, diameter classes, various regions, etc. Some of them have been widely used to estimate the biomass of tropical forest formations (primary and secondary) in the Amazon, such as the equations proposed by Brown et al. (1989), Ducey et al. (2009) and Uhl et al. (1988). Remote sensing Remote sensing has been extensively used to assess the biomass of various types of plant cover. This is due not only to its unique role in covering vast areas, many of which inaccessible, but mainly due to the establishment of empirical relationships between remotely sensed data, enhanced by various techniques (vegetation indexes, spectral mixture analysis, etc.), and the biophysical parameters of the vegetation, such as leaf area index (LAI), plant area index (PAI), above-ground phytomass, etc. In Brazilian vegetation surveys, most studies involving the use of remote sensing have been performed in humid or transition regions, where plant cover is predominantly dense, with low natural exposure of the soil component (AMARAL et al., 1996; SANTOS et al., 1998).

In estimating the biomass of forests in semiarid areas, such as the Caatinga, due to the low percentage of plant cover, the use of remote sensing must consider the influences of soil and herbaceous vegetation. Although the influence of soil has been reduced in some vegetation indexes used most recently, such as, for example, the soil adjusted vegetation index (SAVI), the normalized difference vegetation index (NDVI), which is influenced by the spectral response of that component, has been used as the standard due to the historical data available. Another factor complicating the use of vegetation indexes in semiarid areas is the vegetation’s deciduous nature. In the dry season, when the probability of cloud cover is lower, much of the vegetation loses its foliage, reducing the chances for the widespread use of vegetation indexes to estimate biomass. Nevertheless, whenever good quality images are available, vegetation indexes, such as NDVI, are less subject to saturation with the increase of Caatinga green biomass than when they are used to study the biomass of humid tropical rainforests.

Results Estimate of biomass in the Caatinga The Caatinga region represents about 10% of the national territory. Although, in terms of territory, it is much smaller than other national ecosystems (e.g., the Amazon Forest and the Cerrado), there are several reasons why this ecosystem is considered relevant to the issue of global climate changes induced by greenhouse gas emissions. These reasons include: 1) The Caatinga is one of the ecosystems most frequently abused by man. The reason for this abuse is the anthropic pressure caused by a relatively high population density (about 25% of Brazilian population lives in the Northeast region of the country) on a naturally fragile ecosystem. 2) Due to poverty in terms of natural resources, economically and/or socially viable agroecosystems (irrigated agriculture and family farms) are already at the limit of their exploitation, i.e., there are few alternatives to incorporate new areas for agricultural activities. Therefore, changes in regional climate caused by changes in carbon balance will have far greater social and economic effects than those currently observed with the phenomenon of drought, and they will have devastating consequences on irrigated agriculture (the region’s main source of income). 3) It is estimated that approximately 25% of the Caatinga ecosystem is in a process of desertification. Estimating how much carbon is no longer being fixed due to the Caatinga’s degradation is important to justify the high investments necessary to rehabilitate these areas. In addition to these aspects, due to the lack of planted forests, from a practical point of view, the Caatinga has been the main source of energy in some areas, not only for domestic consumption, but also to power various industries, especially baking, pottery and calcination of several minerals, including gypsum, to obtain plaster. Despite the issues raised above, there are few records of actions that include the Caatinga in any issues relating to CO2 fixation, and even those that exist disregard the Caatinga biome as a

potential instrument for capturing CO2 from the atmosphere. Studies mainly emphasize concern with the release of carbon caused by land use changes (CERRI et al., 2006). Forest inventories conducted in the Caatinga are scarce and report only on specific areas or specific states in the Northeast region of Brazil (IBAMA, 1992; SILVA, 1998). So, between the years 1992 and 1994, the project PNUD/FAO/IBAMA/BRA-87/007 conducted a survey of the plant cover and forest inventory of the Caatinga areas in the States of Ceará, Rio Grande do Norte, Paraíba and Pernambuco (VIRGILIO; PAREYN, 2002). Scarcer still are studies involving quantification of the Caatinga’s biomass. Costa et al. (2002) estimated the above-ground biomass of various typologies in the Seridó region of the Caatinga (Rio Grande do Norte), based on destructive and nondestructive methods, and established the relationship between the above-ground biomass and the normalized difference vegetation index (NDVI). The remnants of Caatinga were recently mapped at a 1:250,000 scale, based on images from the Landsat 7 Enhanced Thematic Mapper Plus (ETM+) satellite, obtained in 2002 (BRASIL, 2007). The results obtained for 7 scenes of this sensor (orbit 216, points 64 to 66 and orbit 215, points 64 to 67), covering an area of approximately 220,000 km2 and including parts of the States of Rio Grande do Norte, Paraíba, Pernambuco, Alagoas and Sergipe, revealed the presence of Caatinga vegetation in about 56% of this area. According to the classification adopted in that study (IBGE, 1992), out of the 56% of areas mapped as Caatinga, about 52% (114,500 km2) were arboreal steppe savannah (a class of Caatinga in which trees are smaller and sparser), and only 4% (8,200 km2) represented forest steppe savanna. The forest steppe savanna corresponds to larger-tree Caatinga, with trees reaching an average height of 5 m and with touching crowns. Costa et al. (2002) estimated the above-ground phytomass of various typologies of Caatinga present in a pilot area which was representative of the Seridó region (Rio Grande do Norte and Paraíba). This study also included an estimate of the above-ground phytomass of Caatinga in desert areas. Costa et al. (2002) conducted one of the few studies, if not the only study, to calibrate the data from the Landsat 7 ETM+ sensor, in order to estimate the above-ground phytomass of the Caatinga. The methodology they used was composed of three stages: 1) assessing above-ground phytomass by destructive method; 2) establishing the relationship between the above-ground total dry weight of the phytomass (TDW = dry weight of stem and branches + dry weight of leaves) and the PAI (plant area index) obtained with the LAI-2000 equipment; 3) establishing the relationship between PAI and NDVI. The equations obtained by Costa et al. (2002) were: TDW = −980.47 + 11,851.25 * PAI, R2 = 76.48 2

PAI = 0.6401 * exp(2.6929 * NDVI), R = 78.02

(1) (2)

Considering the representativeness of this pilot area and the peculiar characteristics of the Caatinga in the Seridó region, those equations were used in this study to estimate the aboveground phytomass of Caatinga vegetation present in two microregions in the Seridó region (Rio Grande do Norte (RN) and Paraíba (PB)). Area studied The area studied comprises the microregions of Seridó Oriental (Rio Grande do Norte (RN)) and Seridó Ocidental Paraibano (Paraíba (PB)), totaling 5,516 Km2, limited by geographical coordinates 6° 06' 14" and 7° 06' 32" S and 36° 15' 32" and 37° 13' 41" W. This region comprises part

of the core of desertification of the Seridó region. The two microregions are part of the northeastern semiarid area, with dry, very warm climate, with a mean annual rainfall of 500 mm and with the rainy season occurring in the months from January to May. According to Köppen’s classification, the climate is of type BsWh. The soils of this area were entirely mapped by Brasil (1971, 1972) at a 1:500,000 scale, and partly by Silva et al. (2002) at a 1:100,000 scale. Those authors found the following soil types: Chromic Luvisols, Litholic Neosols [US: Lithic Orthents / Lithic Psamments], Natric Planosols and Regolithic Neosols [US: Psamments]. In the more detailed mapping made by Silva et al. (2002), highly eroded and truncated stages were identified for some of the Chromic Luvisols located in areas undergoing desertification. Even in areas where there was no detailed mapping of the soil, its degradation is visible in several locations, where it is possible to notice the removal of entire horizons by erosion (Figure 1).

CHAP 4 - FIGURE 1 Fotos

Photos

Figure 1.a) Appearance of a degraded area of Caatinga; b) appearance of preserved Caatinga in an area of the Seridó sertão [semiarid backland], near the seat of the municipality of Parelhas.

Several activities have contributed to the advanced degradation in much of this area, including improper management practices in cotton crops (a major activity in the region until the arrival of Anthonomus grandis [boll weevil; bicudo-do-algodoeiro] in the late 1980s), overgrazing, the many pottery industries that use firewood as their main source of energy, and mining activities (kaolin and other minerals used in construction). The Caatinga, known as Seridó Caatinga, is exclusive to this area and has been described by several authors (AMORIM et al., 2005; CAMACHO; BAPTISTA, 2005; COSTA et al., 2002; MIRANDA et al., 2000; SANTANA; SOUTO, 2006). It is characterized as open hyperxerophytic vegetation, consisting of sparse and stunted plants of arbustive and/or arboreal species (ANDRADE-LIMA, 1981). Material and method The scenes used were 215-65 and 215-64, from the Landsat ETM+ sensor with flyover dates of May 20, 2000 and April 5, 2001, respectively. To correct the effects of differences in weather conditions and solar position at the two times, light and dark pseudo-invariant targets were identified in the area where the two scenes overlap. Based on the spectral response of these targets, the linear regression equation for the transformation of gray levels was established using scene 215-65 as reference. The images were then mosaiced and enhanced using the NDVI. Equations 1 and 2 were used to estimate the above-ground phytomass of the two microregions. Caatinga areas were separated based on supervised classification of the mosaic of these scenes using the method of maximum likelihood, according to the following classes: 1) Caatinga, 2) pasture, 3) crops planted in the ebb zone and riparian vegetation, 4) exposed soil, 5) urban area, 6) water. The reduced presence of clouds and shadows was removed using a mask.

The image of the above-ground phytomass of areas with Caatinga vegetation was divided into five classes, according to the following yields: 0.1 Mg ha−1 to 5 Mg ha−1, 5 Mg ha−1 to 10 Mg ha−1, 10 Mg ha−1 to 15 Mg ha−1, 15 Mg ha−1 to 20 Mg ha−1 and over 20 Mg ha−1 (Figure 2).

CHAP 4 - FIGURE 2 Legenda

Legend

(…) a (…)

(…) to (…)

Maior do que (…)

Greater than

Outras cidades

Other municipalities

Divisão estadual

State border

Figure 2. Map of the above-ground phytomass classes of Caatinga vegetation present in the microregions of the Seridó Oriental (Rio Grande do Norte (RN)) and the Seridó Ocidental Paraibano (Paraíba (PB)).

Due to spectral similarity and resulting confusion with exposed soil areas, the biomass yield estimate for areas degraded by desertification was obtained by identifying and mapping some of these areas in the field. Caatinga coverage in the area studied is 457,496 ha, i.e., about 83% of the total area. Pastures cover most of the remaining 17% of the area. Total above-ground phytomass was estimated to be 4.51 × 107 Mg with an average of approximately 9 Mg ha−1. This average is extremely low when compared to the values found by Kauffman et al. (1993) (74 Mg ha−1) for Caatinga vegetation in Serra Talhada, Pernambuco (PE) and less than half of those observed by Amorim et al. (2005), 25.1 Mg ha−1, in studies conducted at the Seridó Ecological Station. However, they may be considered compatible with the values revised by Amorim et al. (2005) for the Caatinga in the Açu region (10 Mg ha−1). The distribution of that phytomass by yield range is shown in Table 1. Table 1. Production of above-ground phytomass in the Seridó region (Rio Grande do Norte (RN) and Paraíba (PB)). Class of above-ground phytomass (Mg ha−1) 0.1 to 5

Area (ha)

Area studied (%)

Average (Mg ha−1)

Total (Mg)

65,817

11.9

4.29

3.14 x 106

5 to 10

258,425

46.8

7.09

2.04 x 107

10 to 15

92,804

16.8

12.12

1.25 x 107

15 to 20

31,993

5.8

16.95

6.03 x 106

> 20

8,457

1.5

32.85

3.09 x 106

Mg = megagram = ton.

It can be observed that approximately 60% of the area showed above-ground phytomass yield values of less than 10 Mg ha−1. Moreover, it can also be observed that about 12% of the area has a mean above-ground phytomass yield under 5 Mg ha−1. Field observations showed that these areas were associated with

severe degradation problems (Figure 1a), which could, in part, explain their influence on the decrease in the overall mean values found for this area. As reported by the region’s farmers, much of the degraded areas have show signs of desertification for at least 60 years. This period would have been sufficient for recovery of native vegetation, had the area not been so severely degraded. Thus, one can infer that a considerable part of the nearly 66,000 ha that showed an average aboveground phytomass lower than 5 Mg ha−1 consists of areas extremely degraded by human action. This value is, at least, four times lower than the average yield of above-ground phytomass of areas with more preserved Caatinga, which are in the class of above-ground phytomass exceeding 20 Mg ha−1 (Figure 1b).

Pinus spp. biomass in the South region of Brazil Pinus taeda L. The results presented are part of a broader study, and only some of the results are presented. The majority of forests planted in the South region of Brazil, about 130,000 ha, which corresponds to 80% of the total, are of Pinus taeda. It is one of the species most commonly grown in the cooler regions of the southern highlands of Brazil, due to its high volume increase and timber with low resin content. Material and method For this study, biomasses were estimated for two experiments, which tested various spacings and management systems, in the municipality of Rio Negrinho, Santa Catarina (SC), at an altitude of 860 m, latitude 26° 14' S and longitude 49° 30' W. The region’s climate is mesothermal humid subtropical (Cfb). Pruning was carried out at 5 years of age, at a height of 3 m. Biomass collection was done in August and September 2004. The biomass collected included leaves, branches, twigs, fruits, trunk and bark from the two experiments. For each selected treatment, 7 trees from various diameter classes were cut down, chosen from the inventory data: 2 small, 3 medium and 2 large. No trees from the two border strips were used. After cutting, the trees were measured: total height, crown height, and the leaves, green branches, twigs and cones were separated and weighed. Then, a biomass sample from each component was weighed (green weight) and sent to the laboratory to obtain dry weight after drying in a forced-air circulation oven, at a temperature of 100 °C until it reached a constant weight. Cross section disks were obtained at every meter of wood, and taken to the laboratory to determine wood density, bark dry weight, wood dry weight and total dry weight of the log (Figure 3).

CHAP 4 - FIGURE 3 Foto

Photo

Figure 3. Pinus taeda disks used to calculate trunk biomass.

Results In all treatments observed, at 14 years of age, the highest biomass allocation was in the trunk, in agreement with other studies on the species (JOKELA; MARTIN, 2000; VALERI, 1988; WATZLAWICK, 2003). The component with the second highest biomass allocation was the crown, followed by living branches, dead branches, leaves and fruits. The same biomass allocation pattern was observed in various management regimes (Figure 4). As mentioned by Larson et al. (2001), planting density controls the growth of the crown. This can be observed in the different biomass allocation percentages at various spacings (Figure 5). CHAP 4 - FIGURE 4 Alocação de Biomassa

Biomass allocation

Regime de manejo

Management regime

Folhas

Leaves

Galhos mortos

Dead branches

Galhos vivos

Living branches

Frutos

Fruits

Casca

Bark

Tronco

Trunk

Copa

Crown

1 (UA 1): Desbaste seletivo, com remoção 1 (UA 1): Selective cutting, with removal of de 71% das árvores aos 10 anos, somente 71% of trees at 10 years (established June (implantação 6/1985); 1985); 2.3 (UA 7): Idem ao anterior, com remoção 2.3 (UA 7): Same as above, with removal of de 57% das árvores aos 10 anos e de 25% 57% of trees at 10 years and 25% at 14 years (established June 1985); aos 14 anos (implantação 6/1985); 5 (UA 6): Desbaste seletivo, com remoção 5 (UA 6): Selective cutting, with removal of de 37% das árvores aos 10 anos e 44% aos 37% of trees at 10 years and 44% at 15 years (established June 1985); 15 anos (implantação 6/1985); 4.2 (UA 9): Desbaste seletivo, com remoção 4.2 (UA 9): Selective cutting, with removal of de 42% das árvores aos 10 anos e 42% aos 42% of trees at 10 years and 42% at 15 15 anos (implantação 6/1985); years (established June 1985); 6 Desbaste de 56% das árvores, com corte sistemático da 7ª linha e seletivo nas demais, mais 40% aos 15 anos (implantação 6/1985);

6 Cutting of 56% of trees, with systematic cutting of the 7th row and selective cutting of the remaining rows, plus 40% at 15 years (established June 1985);

Figure 4. Biomass allocation (%) of Pinus taeda at 18 years of age in Rio Negrinho, Santa Catarina (SC), according to various management regimes.

CHAP 4 - FIGURE 5 Alocação de Biomassa

Biomass allocation

Espaçamento

Spacing

Folhas

Leaves

Galhos mortos

Dead branches

Galhos vivos

Living branches

Frutos

Fruits

Casca

Bark

Tronco

Trunk

Copa

Crown

Figure 5. Biomass allocation of Pinus taeda at 14 years of age in Rio Negrinho, Santa Catarina (SC).

Wider spacings resulted in greater increases in DBH and cylindrical volume, but less growth in height. The effect of spacing has been extensively discussed, mainly regarding volumetric production (LARSON et al., 2001; OLIVEIRA, 1995; SANQUETTA; BALBINOT, 2004; SANQUETTA et al., 2002). All classes observed showed the same behavior pattern, except for height. A spacing of 6.25 m2 showed higher variations (standard deviation); however, it was not possible to explain such behavior in this study (Figure 5). In narrower spacings, crowns begin to compete much earlier than in wider spacings and, consequently, biomass allocation is shifted to branches and needles, forming smaller crowns. The results observed in this study did not show a significant tendency in this difference. It is likely that the level of competition up until the time of sampling is not yet clear. Other studies on spacing with this species have shown that higher initial planting densities result in higher biomass allocation to the trunk (BURKES et al., 2003). In this study, for the spacings presented, where survival rates remained virtually unchanged, the effect of competition among plants up to 11 years of age can still be considered small. Survival tends to decrease with age at smaller spacings, causing a decrease in growth rate (Figure 6). Probably the biggest drop in this rate could be observed at more advanced ages, but that is the period where thinning is performed in most management regimes.

CHAP 4 - FIGURE 6 Sobrevivência

Survival

Idade (anos)

Age (years)

Figure 6. Pinus taeda survival at various spacings in the region of Rio Negrinho, Santa Catarina (SC).

Similar studies developed by Jokela and Martin (2000) and Martin and Jokela (2004) show a significant drop in the survival rate of Pinus taeda, the authors having attributed this fact to limitations in access to nutrients. The results indicate that the biomass increase curve is altered by the effect of competition (Figure 7). The quantities of biomass found in this study differ from other results obtained in the region (VALERI, 1988; WATZLAWICK, 2003), but are similar to those found by Schumacher (2000, cited by WATZLAWICK, 2003). Part of the difference may be attributed to the different methodologies used to estimate biomass. Other aspects, such as management, degree of enhancement of the material used for planting, soil and climate also affect this characteristic. Johnsen et al. (2001) mention the differences found in Pinus taeda biomass studies and recommend further studies at various locations and various ages to enhance the logic in the selection of models.

CHAP 4 - FIGURE 7 Biomassa

Biomass

Idade (anos)

Age (years)

−1 Figure 7. Above-ground Pinus taeda biomass increase (Mg ha ) at various spacings in the region of Rio Negrinho, Santa Catarina (SC).

Pinus elliotii Eng. The area planted with Pinus elliottii in the South region of Brazil is smaller than the area planted with Pinus taeda, and, although it can be used for resin production, it has slower growth. Area studied The study on biomass was conducted in a commercial area in Caçador, Santa Catarina (SC), altitude 1,080 m, latitude 26° 14' S and longitude 49° 30' W. Samples were taken from 10 trees (3 small, 4 medium and 3 large), with 10 years of age, using the same methodology as above to determine dry biomass per component. For the two species, up to 9 years of age, even at smaller spacings such as those tested for Pinus taeda, stabilization of the growth curve was not observed. Although conditions are different, the biomass quantity is much lower for Pinus elliottii (Figures 7 and 8). At similar spacings, the difference is about half of that observed for Pinus taeda. For this reason, plantations of Pinus taeda are more common than Pinus elliotii in the South region of Brazil.

CHAP 4 - FIGURE 8 Biomassa

Biomass

Idade (anos)

Age (years)

−1 Figure 8. Above-ground Pinus elliotii biomass increase (Mg ha ) in the region of Caçador, Santa Catarina (SC).

The study’s goal was not to compare species, but, as in Pinus taeda, more than 60% of dry matter is also allocated in the trunk (Figure 9). However, due to the lower growth rate with less competition between plants, higher uniformity was observed in the various components. Currently, besides being a support for carbon sequestration projects, this knowledge on biomass is also important for using forests for energy. The use of crop residues or forest biomass for energy co-generation has become an increasingly used practice and, therefore, there is an increasing necessity for basic information to support production potential and to quantify the components to be used for each purpose or even to be left in place for nutrient cycling. CHAP 4 - FIGURE 9 Alocação da biomassa

Biomass allocation

Parte da planta

Plant component

I

L

S

H

GS

DT

GV

GB

Pont

Tip

CG

BB

CT

TB

Aciculas

Needles

Figure 9. Biomass allocation (L: low; M: medium and H: high) in Pinus elliottii at 10 years of age in Caçador, Santa Catarina (SC). (DT: dry twigs, [GB: green branches, Tip: tip, CG BB: branch bark, TB: trunk bark).

Formatado: Não Realce Formatado: Não Realce Formatado: Não Realce Formatado: Não Realce Formatado: Não Realce

Amazon

Formatado: Não Realce Formatado: Não Realce Formatado: Não Realce

Evaluation of biomass stock in homogeneous forest plantations in Central Amazon Native species endemic to the Amazon basin are potentially the ones that can better adapt to soil and climate conditions and that respond more effectively to the rehabilitation of areas altered by agriculture and livestock. This section concerns the assessment of stock and compartmentalization of the biomass of two species of importance for the forestry sector: Sclerolobium paniculatum Vogel [tachi-branco] and Carapa guianensis Aubl [andiroba].

Material and method Criterion for choice of species These species are part of the forest species experimental program conducted by Embrapa Western Amazon and were selected in accordance with the methodology for selecting species and provenances proposed by Burley and Wood (1979). They have the following characteristics: • Adaptability to the Amazon’s degraded and low fertility soils. • Ability to improve the quality of soils in a state of degradation. • Regional economic interest. • Short to medium term rotation period. • Technologies available for seed production and plantation establishment. • Rapid growth (ongoing increase in height over 0.6 m year−1). • Local demand for timber, fruits, oil and for energy purposes. Characterization of species Sclerolobium paniculatum [tachi-branco] is a species of the family Caesalpiniaceae (Leguminosae: Caesalpinioideae). In homogeneous plantations they show very strong apical dominance, favoring silvicultural management for timber production and energy purposes. It is a species that is likely to add agricultural value due to its capacity of storing nitrogen and organic matter, contributing to the rehabilitation of degraded soils in pure, mixed or agroforestry systems. Carapa guianensis [andiroba] is a multipurpose species native to the Amazon, belonging to the Meliaceae family, which has favorable characteristics for its expansion in the form of plantations, mostly in mixed arrangements with other species. Besides producing excellent quality wood, its seed oil can be extracted for use in perfumery, cosmetics, medicine and energy segments. Location The plantations are located in Embrapa’s experimental plots, 30 km from Manaus, Amazonas (AM), on Highway AM-010, latitude 03° 08' S, longitude 60° 01' W, altitude 50 m. The predominant vegetation is tropical forest and the soil type is Dystrophic Yellow Latosol [Yellow Latosol = US: Xanthic Haplustox] (EMBRAPA, 1990). According to Lima3 (2004), the area’s mean annual temperature is 27.5 °C, the mean annual minimum temperature is 22.3 °C and annual rainfall is 2,880.9 mm. Individual relationships for estimating biomass Collection of samples to estimate biomass was conducted in August and September (andiroba and tachi-branco, respectively), 2007.

3

Data on mean temperatures and rainfall recorded from 1998 to 2002.

The estimate of the above-ground biomass of the species was initially made using the equations presented in Table 2. Those equations are based on the Schumacher model (Schumacher, 1939). Table 2. Single and double input equations to estimate the weight of above-ground biomass of trees. No. 1

Mathematical formulation w = β0 · xβ1

2

w = β0 + β1 · x + β2 · x²

3

w = β0 + β1 · x + β2 · x² · h

4

w = β0 + β1 · x + β2 · x² + β3 · x² · h

5

w = β0 + β1 · x³ + β2 · x² · h

6

w = β0 + β1 · x + β2 · h

7

w = β0 + xβ1 + hβ2

Note: X = diameter at 1.30 m above the ground or individual cross-sectional area. h = total height (m). w = dry or green weight of individual trees (kg). β0, β1, β2, β3 = regression coefficients.

Weight data for the components was obtained in green and dry conditions. The equations were adjusted using the R statistical software package (R DEVELOPMENT CORE TEAM, 2006). • Sclerolobium paniculatum The estimate for the green and dry weight (kg) for tachi-branco was obtained using model 1 (w = β0 ∙ DBHβ1), linearized. The graphical representation of the estimated models and the equations with respective statistics are shown in Figures 10 and 11.

CHAP 4 - FIGURE 10 Biomassa verde total

Total green biomass

pv

gw

ajust

adjust

Curva estimada

Estimated curve

Bandas de confiança

Confidence bands

Valores observados

Values observed

DAP

DBH

Figure 10. Observed data and adjusted exponential model to estimate the total green weight (gw) of the aboveground biomass of Sclerolobium paniculatum, at 9 years of age, with initial and final densities of 1,667 trees ha−1 and 1,088 trees ha−1, respectively.

CHAP 4 - FIGURE 11 Biomassa seca total

Total dry biomass

ps

dw

ajust

adjust

Curva estimada

Estimated curve


Confidence bands

Valores observados

Values observed

DAP

DBH

Figure 11. Observed data and exponential model adjusted to estimate the total dry weight (dw) of the aboveground biomass of S. paniculatum, at 9 years of age, with initial and final densities of 1,667 trees ha−1 and 1,088 trees ha−1, respectively.

• Carapa guianensis The model selected to estimate the dry and green weight of andiroba was model 1 (p = DBHβ1), linearized, without the coefficient β0. The high value for estimated standard error (0.3176) is explained by problems of tree bifurcation and an attack by moth larvae of the genus Hypsipyla. Figures 12 and 13 contain all the observed data, the adjusted model and the adjusted equation of green and dry weight (kg), respectively.

CHAP 4 - FIGURE 12 Biomassa seca total

Total dry biomass

ps

dw

ajust

adjust

Curva estimada

Estimated curve


Confidence bands

Valores observados

Values observed

DAP

DBH

Figure 12. Observed data and exponential model adjusted to estimate the total dry weight (dw) of the aboveground biomass of C. guianensis, at 15 years of age, with initial and final densities of 1,111 trees ha−1 and 741 trees ha−1, respectively.

CHAP 4 - FIGURE 13 Biomassa verde total

Total green biomass

pv

gw

ajust

adjust

Curva estimada

Estimated curve

Formatado: Não Realce


Confidence bands

Valores observados

Values observed

DAP

DBH

Figure 13. Observed data and exponential model adjusted to estimate the total green weight (gw) of the aboveground biomass of C. guianensis, at 15 years of age, with initial and final densities of 1,111 trees ha−1 and 741 trees ha−1, respectively.

Estimate of biomass stock The tachi-branco and andiroba native species present characteristics that qualify them as potentially useful for reforestation programs. In addition to their conventional uses, such as wood, fruit and energy production, during their life cycle, these species can provide environmental services such as erosion control, restoration of soil productive capacity of and sequestering of atmospheric carbon. Table 3 presents the mean values of dry biomass production and the estimated increase for a hectare of Carapa guianensis and Sclerolobium paniculatum. Table 3. Average values of dry above-ground biomass and average annual increase (AAI) of Carapa guianensis [andiroba] and Sclerolobium paniculatum [tachi-branco] in homogeneous plantations in Manaus, Amazonas (AM). Species

Final density (n ha−1)

S. paniculatum(1)

1,088

C. guianensis(2)

741

Biomass (kg tree−1)

AAI (kg tree−1 year−1)

145.14 a

16.13 a

158.0

17.56

(±29.35)

(±3.26)

(±29.35)

(±3.26)

284.60 b

18.97 a

182.7

13.38

(±15.43)

(±1.03)

(±15.43)

(±1.03)

Biomass (Mg ha−1)

AAI (Mg ha−1 year−1)

(1)

Spacing: 3 m × 2 m; age: 9 years; death and thinning: 34.7%. Spacing: 3 m × 3 m; age: 15 years; death and thinning: 33.3%. Note 1: Values followed by different letters in the column are significantly different, with P-value < 5% (unpaired ttest). Note 2: Values in parentheses represent the standard error of the average. (2)

The results presented in Table 3 indicate a production of dry biomass in the following classification order: andiroba (284.6 kg tree−1) > tachi-branco (145.1 kg tree−1). Estimating dry biomass at t ha−1 year−1, there is an inversion and tachi-branco presents a higher value (17.6 kg) in relation to andiroba (12.2 kg). This result is due to lower ages and higher planting density of tachi-branco (1,088 kg tree−1). Studies conducted by Dünish et al. (2003), in the same study area as andiroba, presented the following biomass values (t ha−1): 0.1; 77.7 and 86.0 at 1, 5 and 8 years of age, respectively. Results presented by Xaud et al. (2006), working with the same genetic material and same planting density in the State of Roraima, indicated a dry biomass production of 126.7 t ha−1 for tachi-branco, at 7

Formatado: Não Realce

years of age. This value is consistent with the value obtained in this study, of 158.0 t ha−1, at 9 years of age. Figure 14 graphically presents the estimated biomass production values for 1 ha of andiroba and tachi-branco plantation, in an area originally undergoing degradation in the region of Manaus.

CHAP 4 - FIGURE 14 Peso total

Total weight

Peso verde

Green weight

Peso seco

Dry weight

Taxi

Tachi-branco

anos

years

Figure 14. Estimated total above-ground biomass for 1 ha of plantation with 2 species endemic to the Amazon region.

Although andiroba presents a slower initial growth than tachi-branco, it stabilizes at later ages, with greater increases in biomass (kg tree−1 year−1). At 9 years of age, tachi-branco trees showed stabilization in growth with a planting density of 1,088 ha−1, therefore requiring silvicultural interventions to sustain the initial growth rate of the plantation and increase biomass production. Although difficult to obtain, knowledge about forest biomass storage capacity of various species with potential for cultivation in the Amazon region is very important for planning local consumption and accessing the global carbon market. Furthermore, knowing how much biomass is being stored in various components of the species allows for planning silvicultural interventions designed to favor maintenance and expression of all of the genetic productive potential of the species and make decisions regarding its use. The process of soil degradation taking place in the Amazon, due to intensive crop farming, mining and grazing, can be reversed by planting forest species. Currently, there are over than 5 million hectares of disturbed areas in the Amazon, which can be reincorporated into the production process through use of appropriate reforestation techniques. Although methodologies to estimate biomass involving destructive sampling are expensive and laborious, such studies should continue to be prioritized to generate reliable estimates on actual carbon-stocking capacity of tree species native to the Amazon. With advances of knowledge in this area, it will be possible to adopt specific silvicultural practices to promote the increase of stocks on a local and regional level. Assessment of biomass stock in experimental forest plantations at various growth stages in a forest region of Roraima

In 1997, Embrapa Roraima initiated a study project involving fast-growing tree species in the State of Roraima to conduct a forest assessment, for multiple uses in homogeneous plantations and agroforestry systems. As forestry business opportunities arise required by worldwide implementation of the Kyoto Protocol, there is an increasing need for assessments making it possible to generate added value for planted forests, in the form of compensations for environmental services, emphasizing their use for carbon sequestration. Thus, the goal of the present study was to evaluate forest species in homogeneous experimental plantings in Roraima, in terms of their biomass accumulation potential. Material and method As samples are destructive and the planted crown is small, with a maximum of 49 to 81 useful plants in each plot, arranged in 486 m2 (2 m × 3 m spacing) or 972 m2 (4 m × 3 m spacing), the methodology selected was an adapted version of Schumacher et al. (2002) for collecting the average individual of each species. This was selected from the inventory of all the useful individuals in each plot, both regarding diameter at 1.3 m height (diameter at breast height – DBH), and total height. The following components were sampled: trunk, twigs, green branches and leaves. The trunk was sampled at 3 heights (base: 0.10 m, 1.30 m and apex of commercial stem). Weighing in the field was performed using 100 g, 300 g, 12 kg, 20 kg hanging scales (dynamometers) as well as ropes and raffia and paper bags. The evaluated plantations were deployed with minimum phosphorus fertilization (20 g P2O5 hole−1), receiving no maintenance fertilization. Table 4. Initial and final density and suppression of individuals in homogeneous plantations, in the Confiança Experimental Field, Embrapa Roraima, 2005. Species Eucaliptus camaldulensis

Initial density (n ha−1) 1,666.7

Suppression(1) (%) 5.0%

Age (years) 4

Spacing (m × m) 2×3

6

4×3

833.3

799.3

4.1%

6

4×3

833.3

799.3

4.1%

6

4×3

833.3

544.2

34.7%

6

4×3

833.3

612.2

26.5%

6

4×3

833.3

493.2

40.8%

7

2×3

1,666.7

7

2×3

1,666.7

8

4×3

Final density (n ha−1) 1,583.0

Bertholletia excelsa [Brazil nut tree; castanheirado-Pará] Para centrolobium [pau-rainha] Dinizia excelsa [angelim] Jacaranda copaia [Parapará] Schefflera morototoni [morototó] Eucaliptus urograndis Sclerolobium paniculatum [tachi-branco] Schizolobium amazonicum [paricá] (1)

Suppression by thinning and/or death.

Source: Adapted from Xaud et al. (2006).

833.3

1,530.6

8.2%

802.7

51.8%

527.2

36.7%

CHAP 4 - FIGURE 15 Castanha-do-brasil

Bertholletia excelsa

Parapará

Jacaranda copaia

Morototó

Schefflera morototoni

Angelim

Dinizia excelsa

Paricá

Schizolobium amazonicum

Biomassa seca

Dry biomass

Figure 15. Dry biomass of individuals of the average individual of each species, Confiança Experimental Field, 2005. Source: Xaud et al. (2006).

According to biomass data analyzed by individual (Figure 15), it was observed that, among species aged 6 years, planted at a 4 m × 3 m spacing, Schefflera morototoni [morototó] and Jacaranda copaia [parapará] stood out with a production of 33.3 kg of dry biomass per individual, whereas Bertholletia excelsa [Brazil nut tree; castanheira-do-Pará] obtained 25.2 kg individual−1. For species aged 7 and 8 years, with 4 m × 3 m spacing, Centrolobium paraense [pau-rainha] yielded only 11.6 kg individual−1, followed by Dinizia excelsa [angelim], with 55.0 kg individual−1. At this spacing, the species presenting the most effective biomass accumulation per individual was Schizolobium amazonicum [paricá], with 109.07 kg individual−1, at 8 years of age. In plantations for production of cellulose and energy, with a spacing of 2 m × 3 m, at 7 years of age, clone 1,270 of the hybrid Eucalyptus urograndis presented 137.0 kg individual−1 and Sclerolobium paniculatum [tachi-branco] presented 157.8 kg individual−1 (Figure 13). Eucalyptus camaldulensis (4 years), although still at a juvenile stage, did not show good biomass accumulation capacity per individual, with only 13.2 kg individual−1. The analysis of individual biomass by species provides a dimension of the capability of each species to accumulate biomass, regardless of the plantation as a whole. Although ages, spacings and final density were not the same for Sclerolobium paniculatum [tachi-branco], Eucalyptus urograndis [paricá] and Schizolobium amazonicum [paricá] (Table 4), these can unquestionably be grouped as the species that expressed the highest biomass accumulation capacity per individual (Figure 15). However, the success of a forestry plantation in terms of biomass accumulation does not depend solely only on the individual’s potential, but on the overall results of a set of individuals present in a given area. Therefore, the amount of biomass per area unit takes into account not only individual productivity for biomass storage, but mainly the efficiency of the entire plantation. In that case, the spacing pattern has an effect on the total biomass yield. In this situation, the final density of individuals per hectare, which takes into account not only planned density but also recognizes real suppression, makes a clear difference.

In the analysis of the results of biomass ha−1 (Figure 16) at sites aged 7 years, the highest values observed were for Eucalyptus urograndis, with 209.6 t ha−1 and Sclerolobium paniculatum [tachi-branco], with 126.7 t ha−1 of dry matter, both at a 2 m × 3 m spacing.

CHAP 4 - FIGURE 16 Castanha-do-brasil

Bertholletia excelsa

Parapará

Jacaranda copaia

Morototó

Schefflera morototoni

Angelim

Dinizia excelsa

Paricá

Schizolobium amazonicum

Biomassa aérea


Figure 16. Dry biomass per hectare for each species studied, Confiança Experimental Field, 2005. Source: Xaud et al. (2006).

Even Schizolobium amazonicum [paricá], aged 8 years and at a 4 m × 3 m spacing, was well below those values, showing only 57.8 t ha−1. However, when contrasting the results obtained by Schizolobium amazonicum [paricá] at 8 years with the other 7 year plantations, Dinizia excelsa [angelim] and Para Centrolobium [pau-rainha] at equal spacing, S. amazonicum showed a superior performance. Regarding the high performance of Eucalyptus urograndis clone, this is an intensively selected exotic material. Clone 1,270 was selected among 4 others in Roraima as the most productive in terms of diameter, height and volume growth (TONINI et al., 2006). This positive result in biomass accumulation was expected, since this hybrid stood out on all the previously conducted growth evaluations, such as in Arco-Verde et al. (2000, 2002) and Tonini et al. (2006). From these results, hybrid Eucalyptus urograndis (clone 1,270) and the species Sclerolobium paniculatum (Figures 17a and 17b, respectively) can be indicated as being of interest for plantations involved in biomass production and/or accumulation in areas with similar environmental conditions to those of the study. Above-ground biomass in a sequential agroforestry system with fallow enrichment in Igarapé-Açu, Pará Sequential agroforestry systems are characterized by alternating periods of agricultural cultivation and fallow (KATO et al., 2006). In the fallow period, there is an establishment of secondary vegetation consisting of herbaceous, arbustive and woody species, whose growth speed and biomass and nutrient accumulation depends on several factors (previous type of land use, intensity of land use, soil fertility, and climate). The enrichment during the fallow period with fast growing species (usually legumes) aims to accelerate the accumulation of biomass and nutrients. Near the end of the agricultural phase (in the area studied, usually cassava cultivation), seedlings of leguminous trees Racosperma mangium (Willd.) Pedley and Sclerolobium paniculatum

Vog. were planted. Thirty months after planting the arboreal species, when the secondary vegetation was 24 months old, the above-ground biomass was determined destructively, according to the methodology described below.

CHAP 4 - FIGURE 17A Fotos

Photos

Figure 17A. Eucaliptus urograndis, clone 1,270, at 7 years of, Confiança Experimental Field. Source: Xaud et al. (2006).

CHAP 4 - FIGURE 17B Figure 17B. Sclerolobium paniculatum [tachi-branco] at 7 years of age, Confiança Experimental Field. Source: Xaud et al. (2006).

Material and method Samples of leguminous trees used for enrichment and successional species were collected from ten 25 m2 plots. Measurements were taken for total height and for diameter at 1.3 m (diameter at breast height – DBH) (for enrichment legumes) and at the base of the trunk (for successional species). All the plants were divided trunk, branches and leaves and weighed in the field to obtain the total green weight. Subsamples of each component were collected to determine the dry weight after oven drying. Successional species were classified into trees, shrubs, vines, herbaceous, pseudo-stems, palm trees and grass. Trees and shrubs were separated into trunks, branches and leaves, and vines were separated into stem and leaves. The total accumulation of biomass in the sequential agroforestry system with fallow enrichment (Table 5), as well as the difference between enrichment species was consistent with another similar study (BRIENZA JUNIOR, 1999). The biomass accumulation values in secondary vegetation were lower than those obtained in similar studies in the region (BRIENZA JUNIOR, 1999; NUNEZ, 1995), possibly due to competition for water, light and nutrients between the enrichment species and secondary vegetation, as leguminous trees are fast growing species. Table 5. Biomass in sequential agroforestry system with fallow enrichment in Igarapé-Açu, Pará. Coordinates Latitude 0° 55' to 1° 20' S and longitude 47° 20' to 47° 50' W Source: Oliveira (2007).

Components

Spacing (m × m)

Age (months)

Plot size (m2)

Number of samples / replicates

Biomass (Mg ha−1)

Racosperma mangium

2×2

30

25

10

19.6

Sclerolobium paniculatum

2×2

30

25

10

4.7

–

24

25

10

6.3

Secondary vegetation

Biomass accumulation in secondary forests in the Eastern Amazon In the Amazon region, secondary forests are formed after the abandonment of agriculture and livestock areas. Fearnside (1996) estimated that in 1990, about 50% of deforested areas in the Brazilian Amazon were at some stage of succession. In tropical regions, secondary forests often represent an important source of income (timber and non-timber products) for local populations (BROWN; LUGO, 1990) and can promote important environmental services, such as carbon sequestration, reestablishment of the cycle of water and nutrients and maintenance of biodiversity (BROWN; LUGO, 1990; MARKEWITZ et al., 2004; NEPSTAD et al., 2001; SOMMER et al., 2002). At a regional and global level, secondary forests can play an important role in carbon dynamics due to their high rates of biomass accumulation (ZARIN et al., 2001), although frequent cutting of these forests results in reduced net rates of carbon sequestration, when compared to total carbon emissions associated with deforestation of the Amazon (STEININGER, 2004). Observational studies have shown that several factors control the biomass accumulation rate in tropical secondary forests, such as historical land use (including type and intensity), adjacent vegetation, soil fertility and climate (GEHRING et al., 1999, 2005; MORAN et al., 2000; UHL et al., 1988; ZARIN et al., 2001, 2005). Therefore, modeling studies regarding biomass accumulation in secondary forests become quite complex (NEEFF, 2005; ZARIN et al., 2001) due to the diversity of factors that control the growth of these forests. The above-ground biomass of trees with a diameter at breast height (DBH) greater than 1 cm in a secondary forest aged about 19 years in Apeú, Pará, was estimated using allometric equations. The equations used were developed specifically for the study site, based on DBH measures (DUCEY et al., 2009). The above-ground biomass estimates in secondary forests in the Eastern Amazon exhibit great variation (Table 6), which is related to land use history and soil and climate conditions. In Apeú, the above-ground biomass accumulated in the area after 12 years of neglect (51.5 ± 2.6 Mg ha−1) is 13% lower than that projected by a model for accumulation of above-ground biomass in secondary forests in the Amazon (ZARIN et al., 2001) and about 70% less than the value estimated by a recently developed model for secondary forests established after the first slash-and-burn cycle in the Central Amazon (GEHRING et al., 2005). Lower biomass values in relation to the predictions of the model may result from the effect of: a) legacy of successive burning events (ZARIN et al., 2005); b) low soil fertility of the location studied; and c) relatively dry periods in the location studied. Table 6. Above-ground biomass in secondary forests in the Eastern Amazon. Location

Coordinates

Soil type

Annual rainfall (mm)

Successional stage(1)

Plot size (m2)

Number of replicates

Biomass (Mg ha−1)

Source

Apeú, PA

1° 19' S, 47° 57' W

Oxisol Oxisol and Ultisol Oxisol and Ultisol Oxisol and Ultisol Oxisol and Ultisol

2,500

12 years

100

2,650

2m

2,650

4

51.5

1

4

17

2

5m

25

40

2

2,650

10 m

100–200

64

2

2,650

15 m

200–400

90

2

Igarapé-Açu, PA

1° 8' 51.1800" S, 47° 38' 22.7040'' W

Igarapé-Açu, PA

1° 8' 51.1800" S, 47° 38'22.7040" W

Igarapé-Açu, PA

1° 8' 51.1800" S, 47° 38' 22.7040" W

Igarapé-Açu, PA

1° 8' 51.1800" S, 47° 38' 22.7040" W

Igarapé-Açu, PA

0° 45' & 1° 39' S, 46° 16' & 48° 15' W

Oxisol

2,500

5 years

10

25

19.9

3

1° 07' 41" S, 47° 47' 15" W

Oxisol and Ultisol

2,500

2.5 years

80

4

24.0

4

Igarapé-Açu, PA (1)

The successional stage is estimated by the age or height of secondary vegetation.

Sources: 1) Vasconcelos (2006); 2) Puig (2005); 3) Denich (1991); 4) Brienza Junior (1999).

Refining the knowledge about the biomass accumulation potential in tropical secondary forests requires long-term studies in areas with distinct soil and climate characteristics and land use history. There is also a major need to expand research efforts focused on quantification of root biomass. In general, studies on biomass stocks in secondary forests should be linked to additional studies on ecosystem processes related to carbon stocks and fluxes in other pools (e.g., plant litter stock and deposition, greenhouse gas emissions, root production). The data generated in these studies is crucial to refine biogeochemical models and models of biomass accumulation in secondary forests. Characterization of chronosequences to evaluate carbon sequestration in land use systems The importance of tropical forests in global carbon flux has been debated over the past 20 years by the scientific community. The conversion of primary forest into pastures and agricultural crops causes significant changes in carbon flux between atmosphere and ecosystems (COMPONENTE..., 1999). Concentrations of carbon dioxide (CO2), the main greenhouse gas, have increased in the atmosphere, from ranging between 275 ppm and 285 ppm in the pre-industrial era (1000 to 1750 years after Christ) to 379 ppm in 2005 (FORSTER et al., 2007). Brazil is among the ten countries with the largest relative contribution to global emissions, accounting for the emission of about 5.4% of this gas, including emissions from fossil fuels, land use change and other greenhouse gases (BAUMERT et al., 2005). Approximately 80% of the Brazilian contribution results from activities such as change of land use patterns (deforestation, agricultural activities, etc.), and only a smaller portion results from burning of fossil fuels. Studies have shown (DETWILER et al., 1985; DIXON, 1995; HOUGHTON et al., 1993; KOTTOSAME et al., 1997; NOBRE; GASH, 1997; SILVA, 1996) that certain agrosystems could function as carbon stock banks, recovering the CO2 lost in the clearing and burning of forests.

This study aimed to determine which land use systems practiced in Rondônia could contribute to increase carbon sequestration in vegetation and soil. Material and method C stocks were measured in seven land use systems and compared to that of the primary forest, in the municipalities of Machadinho do Oeste, Theobroma, Porto Velho and Ji-Paraná, Rondônia. The soils of these municipalities are of type Yellow Latosol [US: Xanthic Haplustox] (Theobroma and Porto Velho) and Red-Yellow Latosols [Latosols = US: Oxisols; Red-Yellow Latosols = US: Rhodic/Xanthic Haplustox] (Ji-Paraná) with clayey texture (EMBRAPA, 1983). The geographical locations and climate of the regions where the studies were conducted are shown in Table 7. When measuring the carbon stock in vegetation and soil, the reference was primary forest, which was compared with the following land use systems (LUS): • Natural capoeira [scrub regenerating on cleared land] (3 years). • Capoeira [scrub regenerating on cleared land] improved with legumes (Inga edulis and Senna sieames, aged 2 years). • Monoculture coffee (7 years). • Agroforestry systems (coffee × Schizolobium amazonicum [bandarra]; coffee × rubber tree, aged 12 years). • Traditional pasture (where the only management is fire) – 8 years. • Managed pasture, where the producer uses leguminous species (Leucaena leucocephala, Acacia angustissima, Cajanus cajan, C. spectabilis, C. juncea) as a protein bank and rotational grazing – (8 years). Table 7. Geographical positions and major climate components of the municipalities of Theobroma, Ji-Paraná, Machadinho do Oeste and Porto Velho, Rondônia (RO). Municipality Theobroma Ji-Paraná Machadinho do Oeste Porto Velho

Climate (Köpper)

Latitude

Longitude

Altitude (m)

Mean annual temperature (°C)

Mean annual rainfall (mm)

Mean relative air humidity (%)

Am Am

10° 64' S 10° 55' S

62° 11' W 61° 58' W

125 240

25.5 25

2,400 2,300

87 82

Am

9° 26' S

61° 58' W

198

26.2

2,390

85

Am

8° 56' S

63° 55' W

98

26

2,500

88

Source: Wikipedia (2007).

C contained in the phytomass of trees, dead trunks, understory vegetation and plant litter, was calculated assuming that the carbon content in biomass was 45% (BROWN et al., 1989). All material was measured in 5 transects with an area of 5 m × 20 m, randomly distributed in the systems studied. The phytomass of trees with a diameter at breast height (DBH) of at least 5 cm was estimated by the allometric equation of Brown et al. (1989). To determine the biomass in individuals with less than 5 cm at DBH, the destructive method was used, where plants were cut and dried to constant weight. The biomass of fallen and dead trees within the transects was calculated using the formula D

* π * H * s (in which: D = diameter, H = height, and s is the specific density estimated at 0.4 g cm−2) (BROWN et al., 1989). The understory vegetation was cut and collected in two 1 m × 1 m subquadrants within each quadrant, including all plant material, such as herbaceous plants, and plants under 2.5 cm in diameter. For plant litter, two samples were randomly collected within the subquadrants, using a 0.5 m × 0.5 m wooden frame. Both the understory and litter materials were dried to constant weight, to calculate the weight of the dry matter. Two soil samples were collected from each transect at depths of 0 cm – 20 cm and 20 cm – 40 cm to determine the soil’s organic carbon content. Figure 18 shows carbon stocks in the land use systems evaluated in Rondônia. Conversion of primary forest into agricultural production systems represents a significant loss of C in the ecosystem. The study showed primary forest stores an average of 188 t ha−1 of C, and that 148 t ha−1 are present in the above-ground phytomass. The remaining carbon comes from the soil, since no thin or thick roots were collected/weighed.

CHAP 4 - FIGURE 18 Estoque de carbono

Carbon stock

Carbono na fitomassa

Carbon in phytomass

Carbono no solo

Carbon in soil

Floresta

Forest

Café

Coffee

SAF1

AFS1

SAF2

AFS2

Capoeira natural

Natural capoeira

Capoeira melhorada

Improved capoeira

Pastagem tradicional

Traditional pasture

Pastagem manejada

Managed pasture

Idade

Age

Figure 18. Carbon stock in phytomass and soil in various land use systems in Rondônia, Brazil, 2007. Afroforestry system AFS1 = coffee × bandarra Agroforestry system AFS2 = coffee × rubber tree

In agroforestry systems with coffee × Schizolobium amazonicum [bandarra] and coffee × rubber tree, the above-ground C stock was 97.2 t C ha−1 and 64.5 t C ha−1, equivalent to 65.7% and 43.6% of C contained in primary forest. In the coffee monoculture system (7 years), the maximum C stored in above-ground biomass was 16.60 t C ha−1 (11.2% of the C stock in the forest).

Formatado: Não Realce Formatado: Não Realce

For the fallow area with natural capoeira [scrub regenerating on cleared land] (5 years), C stock was 11.23 t C ha−1 (7.6% of the forest), whereas with improved capoeira, in a fallow area with fast growing leguminous species (Inga and Senna – 2 years), it was, on average, 13.03 t C ha−1 (8.8% of the forest C). Figure 18 shows that, in the systems studied, pastures, improved capoeira [scrub regenerating on cleared land] and AFS2 (coffee × rubber tree) had higher amounts of soil carbon than the forest. In improved capoeira, the soil C stock was 21% higher than that of the forests; pastures, both traditional and improved, and AFS2 showed accumulations 18% and 14.6% higher than the primary forest system. Cerri et al. (1996), studying soil carbon dynamics in the Amazon, found that, 8 years after deforestation and establishment of pastures, the soil still contained 50% of the carbon from the forest (KOTTO-SAME et al., 1997). Similar results were obtained by Palm et al. (1999) in Cameroon. According to these authors, organic matter in the soil is physically protected from losses, due to the amount of material that does not undergo complete combustion after cutting and burning the forest. Although some systems present high levels in terms of C stock evaluated at a particular crop stage, consideration must be given to the significance of the values of the annual carbon accumulation rate (Ic, in t C ha−1 year−1) in the systems. This rate was calculated by taking into account the carbon stock at the period studied and the time frame in which each system remains in production or in use. The recovery of the carbon lost as a result of changes in plant cover depends on how much time the systems are in use. Calculating the average time of carbon stock for a coffee plantation with a 7 year establishment stage, where phytomass is the highest, followed by 5 years of production, until cutting and reestablishment (total of 12 years), the plantation can accumulate 19% of the carbon contained in a primary forest system (Table 8). Table 8. Mean values of above-ground carbon (Cabove), accumulation rate of carbon/year (Ic), time for maximum carbon accumulation (Tmax), maximum carbon accumulated in Tmax (Cmax) and comparison of carbon in land use systems in relation to primary forest Cmax/Cpf, Rondônia, Brasil, 1997. Age (years)

Cabove (t ha−1)

Ic (t ha−1 year−1)

Tmax (years)

Cmax (t ha−1)

Cmax/Cpf

100

148

–

–

148

1

5

11.23

2.2 b

5

11.0

0.07

2

13.0

6.5 a

3

19.5

0.13

AFS – Coffee × Rubber tree

12

97.2

8.1 a

15

121.5

0.82

AFS – Coffee × bandarra

12

64.5

5.4 a

15

80.6

0.54

Traditional pasture

8

5.7

0.7 c

10

7.1

0.05

Improved pasture

8

6.0

0.8 c

10

7.6

0.09

Land use system Primary forest Natural capoeira [scrub regenerating on cleared land] Improved capoeira [scrub regenerating on cleared land]

CV = 18%. Means followed by the same letter do not significantly differ among them by Tukey test at 5% significance level.

In agroforestry systems and improved capoeiras [scrub regenerating on cleared land], Ic values did not differ significantly among them. The maximum carbon stock potential of agroforestry systems (AFSs), estimated at a rotation time of 15 years, was 82% and 54%, respectively, for coffee × rubber tree and coffee × Schizolobium amazonicum [bandarra]. Improved capoeiras, aged 3 years, sequestered, on average, 13% of the C contained in the forest. However, if

they had remained in use for the same period estimated for AFSs (15), they might store up to 66% of the C contained in a primary forest. Despite the loss of carbon through phytomass, at the time of cutting and burning of primary forests, it is possible, in space and time, to capture and store significant quantities of carbon in agroforests. Similar values for the carbon storage rate were found by Kotto-Same et al. (1997) in agroforestry systems with cocoa in Cameroon, and by Palm et al. (1999) in Indonesia, in rubber tree plantations aged 25 years. Dixon (1995), evaluating agroforestry systems in over 50 countries with various ecoregions, noted that these systems could reduce greenhouse gas emissions and conserve or capture carbon. According to the author, carbon stock values, including C below and above ground, range from 12 t C ha−1 and 228 t C ha−1, with the humid tropics having the greatest potential for carbon accumulation in biomass. The establishment of agroforestry systems and plantations of fast-growing leguminous species (Inga edulis and Senna seames) in fallow areas accumulates carbon over time and can recover the amounts lost in the cutting and burning of primary forest systems. The agroforestry systems studied can act as a carbon storage bank, recovering between 82% and 54% of the C contained in the forest, over a period of 15 years. Fallow areas with fast-growing leguminous trees can accumulate annual C rates similar to those of agroforests. Pastures established for 8 years accumulate between 5% and 9% of the carbon contained in the phytomass before deforestation. The amount of soil C, however, is higher, when compared to the amount found in the forest ecosystem. Pastures, capoeira [scrub regenerating on cleared land] and coffee × rubber tree agroforestry system (AFS) had higher amounts of soil carbon than the forest. Monoculture coffee and coffee × Schizolobium amazonicum [bandarra] AFS, in natural capoeira [scrub regenerating on cleared land] aged 5 years, showed C stocks below those contained in the forest before cutting and burning. The remaining systems studied showed soil C values above those of the primary forest.

Final considerations Brazil has a vast territorial extension, with wide variations in climate and soil, different vocations regarding land use and a complex forest law. Thus, basic regionalized knowledge is required in order for measures to mitigate greenhouse gases to be efficient and to actually help the fight against global warming and towards maintaining sustainability. The pressure on remaining forests, especially near the larger population concentrations, can be minimized by correctly and adequately employing production systems, which not only function as providers of raw materials but also serve the ecological purpose of sequestering carbon. The results presented here represent a small part, considering the diversity of conditions and production systems in the country, but they unquestionably show the mitigation potential and the need for more studies, which must be systematized and contextualized with global requirements. New research actions must continue and complement these results, emphasizing the need for multiple, long-term observations which can serve as the basis for mitigation measures and support the implementation of environmental services.

References ADEGBIDI, H. G.; JOKELA, E. J.; COMERFORD, N. B.; BARROS, N. F. Biomass development for intensively managed loblolly pine plantation growing on Spodosols in the southeastern U.S.A. of root/shoot biomass. Forest Ecology and Management, Amsterdam, NL, v. 167, p. 91-102, 2002. ADEGBIDI, H. G.; JOKELA, E. J.; COMERFORD, N. B.; BARROS, N. F. Root development of young loblolly pine in Spodosols in Southeast Georgia. Soil Science Society of America Journal, Madison, v. 68, p. 596-604, 2004. AMARAL, S.; SOARES, J. V.; ALVES, D. S.; MELILO, E. M. K.; ALMEIDA, S. A. S.; SILVA, O. F.; SILVEIRA, A. M. Relações entre índice de área foliar (LAI), área basal e índice de vegetação (NDVI) em relação a diferentes estágios de crescimento secundário na Floresta Amazônica em Rondônia. In: SIMPÓSIO BRASILEIRO DE SENSORIAMENTO REMOTO, 8., 1996, Salvador. Anais... Salvador: INPE; SELPER, 1996. p. 485-489. AMORIM, I. L.; SAMPAIO, E. V. S. B.; ARAÚJO, E. L. Flora e estrutura da vegetação arbustivo-arbórea de uma área de caatinga do Seridó, RN, Brasil. Acta Botanica Brasilica, Brasília, DF, v. 19, p. 615-623, 2005. ANDRADE-LIMA, D. The caatinga dominium. Revista Brasileira de Botânica, São Paulo, v. 4, p. 149-153, 1981. ARCO-VERDE, M. F.; SCHWENGBER, D. R.; MOURÃO JÚNIOR, M. Avaliação silvicultural de espécies florestais com potencial para o uso em sistemas agroflorestais no estado de Roraima. In: CONGRESSO BRASILEIRO SOBRE SISTEMAS AGROFLORESTAIS, 4., 2002, Ilhéus. Anais... Ilhéus: Agromídia, 2002. 1 CD-ROM. ARCO-VERDE, M. F.; SCHWENGBER, D. R.; XAUD, H. A. M. Avaliação silvicultural de espécies florestais nativas e introduzidas em região de floresta no estado de Roraima. In: CONGRESSO E EXPOSIÇÃO INTERNACIONAL SOBRE FLORESTAS, 6., 2000, Porto Seguro. Resumos técnicos... Rio de Janeiro: Instituto Ambiental Biosfera, 2000. p. 33-34. BAUMERT, K. A.; HERZOG, T.; PERSHING, J. Navigating the numbers: greenhouse gas data and international climate policy. Washington, DC: World Resources Institute, 2005. 122 p. BOOTH, T. H. International responses to threats of climate change. In: CARBON ACCOUNTING IN FORESTS, 2003, Canberra, AU. Proceedings... Available at: . Accessed on: 29 May 2003. BORN, R. H. O sistema climático e o fenômeno das mudanças climáticas: aspectos técnicos e científicos. [São Paulo]: Vitae Civilis, 1993. 25 p. BRASIL. Ministério da Agricultura. I. Levantamento exploratório - reconhecimento de solos do Estado do Rio Grande do Norte. Recife: Sudene-DRN, 1971. 531 p. (Boletim Técnico, 21; Sudene-DRN. Série Pedologia, 9). BRASIL. Ministério da Agricultura. I. Levantamento exploratório - reconhecimento de solos do Estado da Paraíba: II. Interpretação para uso agrícola dos solos do Estado da Paraíba. Recife: Sudene-DRN, 1972. 683 p. (Boletim Técnico, 15; Sudene-DRN. Série Pedologia, 8). BRASIL. Ministério do Meio Ambiente. Programa de Conservação e Utilização Sustentável da Diversidade Biológica Brasileira – PROBIO: Subprojeto levantamento da cobertura vegetal e do uso do solo do bioma caatinga. Brasília, 2007. Available at: . Accessed on: 29 Nov. 2007. BRIENZA JUNIOR, S. Biomass dynamics of fallow vegetation enriched with leguminous trees in the Eastern Amazon of Brazil. 1999. 129 f. Thesis (PhD) - Agricultural Sciences of the Faculty of Agricultural Sciences George-August, University Goettingen, Göttingen. BROWN, S. A.; GILESPIE, J. R.; LUGO, A. E. Biomass estimation method for tropical forests with applications to forest inventory data. Forest Science, Washington, DC, v. 35, p. 881-902, 1989. BROWN, S. Estimating biomass and biomass change of tropical forests: a primer. Rome, IT: FAO, 1997. 137 p. (FAO. Forestry Paper, 134). BROWN, S. Measuring carbon in forests: current status and future challenges. Environmental Pollution, Essex, v. 116, p. 363-372, 2002. BROWN, S.; LUGO, A. E. Tropical secondary forests. Journal of Tropical Ecology, New York, v. 6, n. 1, p. 1-32, 1990.

BURKES, C. E.; WILL, R. E.; BARRON-GAFFORD, G. A.; TESKEY, R. O.; SHIVER, B. Biomass partitioning and growth efficiency of intensively managed Pinus taeda and Pinus elliottii stands of different planting densities. Forest Science, Washington, DC, v. 49, p. 223-233, 2003. BURLEY, J.; WOOD, P. J. Manual sobre investigaciones de espécies y procedencias con referencia a los trópicos. Oxford: C.F.I., 1979. 233 p. CAMACHO, R. G. V.; BAPTISTA, G. M. M. Análise geográfica computadorizada aplicada à vegetação da caatinga em unidades de conservação do Nordeste: a) Estação Ecológica do Seridó-ESEC/RN/Brasil. In: SIMPÓSIO BRASILEIRO DE SENSORIAMENTO REMOTO, 12., 2005, Goiânia. Anais... Goiânia: INPE: SELPER, 2005. p. 2611-2618. CERRI, C. C.; BERNOUX, M.; CARVALHO, M. C. S.; VOLKOFF, B. Emissões e remoções de dióxido de carbono pelos solos por mudanças de uso da terra e calagem: primeiro inventário brasileiro de emissões antrópicas de gases de efeito estufa: relatórios de referência. Brasília, DF: Ministério da Ciência e Tecnologia, 2006. 49 p. CERRI, C. C.; BERNOUX, M.; VOLKOFF, B.; MORAES, J. L. Dinâmica do carbono nos solos da Amazônia. In: ALVAREZ, V. H.; FONTES, L. E. F.; FONTES, M. P. F. (Ed.). O solo nos grandes domínios morfoclimáticos do Brasil e o desenvolvimento sustentado. Viçosa: SBCS: UFV, 1996. p. 61-69. COMPONENTE LBA-Ecologia. LBA Folha Amazônica: Boletim do Experimento de Grande Escala da BiosferaAtmosfera na Amazônia, Cachoeira Paulista, ano 1, n. 1, p. 1-4, Aug. 1999. COSTA, T. C. C. da; ACCIOLY, L. J. O.; OLIVEIRA, M. A. J.; BURGOS, N.; SILVA, F. H. B. B. Phytomass mapping of the “Seridó caatinga” vegetation by the plant area and the normalized difference vegetation indices. Scientia Agricola, Piracicaba, v. 59, p. 707-715, 2002. DENICH, M. Estudo da importância de uma vegetação secundária nova para o incremento da produtividade do sistema de produção na Amazônia oriental brasileira. Belém: EMBRAPA-CPATU, 1991. 284 p. DETWILER, R. P.; HALL, C. A. S.; BOGDONOFF, P. Land use change and carbon exchange in the tropics: estimates for the entire region. Environmental Management, New York, v. 9, p. 335-344, 1985. DIXON, R. K. Sistemas agroforestales y gases invernadero. Agroforesteria en las Americas, Turrialba, v. 2, p. 22-26, 1995. DUCEY, M. J.; ZARIN, D. J.; VASCONCELOS, S. S.; ARAÚJO, M. M. Biomass equations for forest regrowth in the eastern Amazon using randomized branch sampling. Acta Amazonica, Manaus, v. 39, p. 349-360, 2009. DÜNISH, O.; ERBREICJ, M.; EILERS, T. Water balance and water potentials of a monoculture and a enrichment plantation of Carapa guanensis aubl. In the Central Amazon. Forest Ecology and Management, Amsterdam, NL, v. 172, p. 355-367, 2003. EMBRAPA. Centro de Pesquisa Agroflorestal da Amazônia Ocidental. Boletim agrometeorológico. Manaus: EMBRAPACPAA,1990. 23 p. (Boletim de Pesquisa). EMBRAPA. Serviço Nacional de Levantamento e Conservação do Solo. Levantamento de reconhecimento de média intensidade do solo e avaliação da aptidão agrícola das terras do estado de Rondônia. Rio de Janeiro, 1983. 558 p. (Relatório Final, parte I). FAO. The global forest resources assessment 2005: summary report. Rome, IT, 2005. 166 p. Available at: . Accessed on: 20 Feb. 2006. FEARNSIDE, P. M. Amazonian deforestation and global warming: carbon stocks in vegetation replacing Brazil’s Amazon forest. Forest Ecology and Management, Amsterdam, NL, v. 80, p. 21-34, 1996. FORSTER, P.; RAMASWAMY, V.; ARTAXO, P.; BERNTSEN, T.; BETTS, R.; FAHEY, D. W.; HAYWOOD, J.; LEAN, J.; LOWE, D. C.; MYHRE, G.; NGANGA, J.; PRINN, R.; RAGA, G.; SCHULZ, M.; DORLAND, R. van. Changes in atmospheric constituents and in radiative forcing. In: SOLOMON, S.; QIN, D.; MANNING, M.; CHEN, Z.; MARQUIS, M.; AVERYT, K. B.; TIGNOR, M.; MILLER, H. L. (Ed.). Climate change 2007: the physical science bases. Cambridge: Cambridge University Press, 2007. Available at: . Accessed on: 22 May 2007. GEHRING, C.; DENICH, M.; KANASHIRO, M.; VLEK, P. L. G. Response of secondary vegetation in Eastern Amazonia to relaxed nutrient availability constraints. Biogeochemistry, Amsterdam, NL, v. 45, p. 223-241, 1999.

GEHRING, C.; DENICH, M.; VLEK, P. G. Resilience of secondary forest regrowth after slash-and-burn agriculture in central Amazonia. Journal of Tropical Ecology, New York, v. 21, p. 519-527, 2005. HIGUCHI, N. Utilização e manejo dos recursos madeireiros das florestas tropicais úmidas. Acta Amazonica, Manaus, v. 24, p. 275-288, 1994. HIGUCHI, N.; SANTOS, J. dos; RIBEIRO, J. R.; MINETTE, L.; BIOT, Y. Biomassa da parte aérea da floresta tropical úmida de terra firme da Amazônia Brasileira. Acta Amazonica, Manaus, v. 28, p. 153-166, 1998. HOUGHTON, R. A.; UNRUH, J. D.; LEFEBVRE, P. A. Current land use in the tropics and its potential for sequestering carbon. Global Biogeochemical Cycles, Washington, DC, v. 7, p. 305-320, 1993. IBAMA. Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis. Plano de manejo florestal para a região do Seridó do Rio Grande do Norte. Natal, 1992. v. 1. (Projeto PNUD/FAO/IBAMA). IBGE. Instituto Brasileiro de Geografia e Estatística. Manual técnico da vegetação brasileira. Rio de Janeiro, 1992. 92 p. (Série Manuais Técnicos de Geociências, n. 1). JOHNSEN, K. H.; WEAR, D.; OREN, R.; TESKEY, R. O.; SANCHEZ, F.; WILL, R.; BUTNOR, J.; MARKEWITZ, D.; RICHTER, D.; RIALS, T.; ALLEN, H. L.; SEILER, J.; ELLSWORTH, D.; MAIER, C.; KATUL, G.; DOUGHERTY, P. M. Meeting global policy commitments: carbon sequestration and Southern Pine Forests. Journal of Forestry, Washington, DC, v. 99, p. 14-21, 2001. JOKELA, E. J.; MARTIN, T. A. Effects of ontogeny and soil nutrient supply on production, allocation, and leaf area efficiency in loblolly and slash pine stands. Canadian Journal of Forest Research, Ottawa, CA, v. 30, p. 1511-1524, 2000. KATO, O. R. K.; KATO, M. do S. A.; CARVALHO, C. J. R.; FIGUEIREDO, R. de O.; CAMARÃO, A. P.; SÁ, T. D. de A.; DENICH, M.; VILHAUER, K. Uso de agroflorestas no manejo de florestas secundárias. In: GAMA-RODRIGUES, A. C.; BARROS, N. F.; GAMARODRIGUES, E. F.; FREITAS, M. S. M.; VIANA, A. P.; JASMIN, J. M.; MARCIANO, C. R.; CARNEIRO, J. G. A. (Ed.). Sistemas agroflorestais: bases cientificas para o desenvolvimento sustentável. Campos dos Goytacazes: Universidade Estadual do Norte Fluminense Darcy Ribeiro, 2006. p. 119-138. KOTTO-SAME, J.; WOOMER, P. L.; MOUKAM, A.; ZAPFAK, L. Carbon dynamics in slash and burn agriculture and land use alternatives of the humid forest some in Camaroon. Agriculture Ecosystems & Environment, Amsterdam, NL, v. 65, p. 245-256, 1997. KAUFFMAN, J. B.; SANFORD, R. L.; CUMMINGS, D. L.; SALCEDO, I. H.; SAMPAIO, E. V. S. B. Biomass and nutrient dynamics associated with slash fires in neotropical dry forests. Ecology, Tempe, v. 74, p. 140-151, 1993. LARSON, P. R.; KRETSCHMANN, D. E.; CLARK, A.; ISEBRANDS, J. G. Formation and properties of juvenile wood in southern pines: a synopsis. Madison: U.S. Department of Agriculture–Forest Service, 2001. 42 p. LIMA, R. M. B. de. Predição de crescimento do Sclerolobium paniculatum Vogel em cinco sítios na Amazônia em função de fatores de clima e solo. 2004. 194 f. Tese (Doutorado em Engenharia Florestal) - Centro de Ciências Florestais e da Madeira, Universidade Federal do Paraná, Curitiba. MARKEWITZ, D.; DAVIDSON, E.; MOUTINHO, P.; NEPSTAD, D. Nutrient loss and redistribution after forest clearing on a highly weathered soil in Amazonia. Ecological Applications, Tempe, v. 14, p. 177-199, 2004. MARTIN, T. A.; JOKELA, E. J. Stand development and production dynamics of loblolly pine under a range of cultural treatments in north-central Florida USA. Forest Ecology and Management, Amsterdam, NL, v. 192, p. 39-58, 2004. MARTINELLI, L. M.; MOREIRA, M. Z.; BROWN, I. F.; VICTORIA, R. L. Incertezas associadas às estimativas de biomassa em florestas tropicais. In: SEMINÁRIO EMISSÃO X SEQUESTRO DE CO2: UMA NOVA OPORTUNIDADE DE NEGÓCIO PARA O BRASIL, 1994, Rio de Janeiro. Anais... Rio de Janeiro: Companhia Vale do Rio Doce, 1994. p. 198-221. MCCARTHY, J. J.; CANZIANI, O. F.; LEARY, N. A.; DOKKEN, D. J.; WHITE, K. S. (Ed.). Climate change 2001: impacts, adaptation and vulnerability. Cambridge: Cambridge University Press, 2001. Available at: . Accessed on: 4 Jun. 2001. MIRANDA, I.S.; ACCIOLY, L.J.O.; SILVA, F.H.B. Estrutura da vegetação de duas áreas de caatinga no núcleo de desertificação do Seridó, Rio Grande do Norte. In: CONGRESSO NACIONAL DE BOTÂNICA, 51., Brasília, DF, 2000. Resumos... Brasília, DF: Sociedade de Botânica do Brasil, 2000. p. 264-265.

MORAN, E.F.; BRONDIZIO, E.S.; TUCKER, J.M.; SILVA-FORSBERG, M.C.D.; MCCRACKEN, S.; FALESI, I. Effects of soil fertility and land-use on forest succession in Amazônia. Forest Ecology and Management, Amsterdam, NL, v. 139, p. 93-108, 2000. NEEFF, T. Spatial modeling of primary and secondary forest growth in Amazonia. Forest Ecology and Management, Amsterdam, NL, v. 219, p. 149-168, 2005. NEPSTAD, D.; MOUTINHO, P. R. S.; MARKEWITZ, D. The recovery of biomass, nutrient stocks, and deep soil functions in secondary forests. In: MC CLAIN, M. E.; VICTORIA, R. L.; RICHEY, J. E. (Ed.). The biogeochemistry of the Amazon basin. New York: Oxford University Press, 2001. p. 139-155. NOBRE, C. A.; GASH, J. Desmatamento muda clima da Amazônia. Ciência Hoje, Rio de Janeiro, v. 22, p. 32-41, 1997. NUNEZ, J. B. H. Fitomassa e estoque de bioelementos das diversas fases da vegetação secundária, provenientes de diferentes sistemas de uso da terra no nordeste paraense, Brasil. 1995. 184 f. Dissertação (Mestrado em Ciências Biológicas) – Universidade Federal do Pará, Belém. OLIVEIRA, E. B. Um sistema computadorizado de prognose de crescimento e produção de Pinus taeda L. com critérios quantitativos para a avaliação técnica e econômica de regimes de manejo. 1995. 134 f. Tese (Doutorado em Engenharia Florestal) - Universidade Federal do Paraná, Curitiba. OLIVEIRA, V. C. de. Ecofisiologia de componentes de sistemas agroflorestais sequenciais tradicional e alternativo na Amazônia oriental. 2007. Tese (Doutorado em Ciências Agrárias) - Universidade Federal Rural da Amazônia, Belém. PALM, C. A.; WOOMER, P. L.; ALEGRE, J. C.; AREVALO, L.; CASTILLO, C.; CORDEIRO, D. G.; FEIGL, B.; HAIRIAH, K.; KOTTOSAME, J.; MENDES, A.; MOUKAM, A.; MURDIYARSO, D.; NJOMGANG, R.; PARTON, W. J.; RICSE, A.; RODRIGUEZ, V.; SITOMPUL, S. M.; NOORDWIIJK, M. van. Carbon sequestration and trace gas emissions in slash-and-burn and alternative land uses in the humid tropics. Nairobi, KE: ICRAF: ASB Climate Change Working Group, 1999. 33 p. Final Report Phase II. PUIG, C. J. Carbon sequestration potential of land-cover types in the agricultural landscape of eastern Amazonia, Brazil. Goettingen: Cuvillier Verlag, 2005. 159 p. (Ecology and Development Series, 33). R DEVELOPMENT CORE TEAM. A language and environment for statistical computing. Vienna, AT: R Foundation for Statistical Computing, 2006. Available at: . Accessed on: 22 Nov. 2011. REIS, M. G. F.; REIS, G. G.; VALENTE, O. F.; FERNANDES, H. A. C. Seqüestro e armazenamento de carbono em florestas nativas e plantadas dos estados de Minas Gerais e Espírito Santo. In: SEMINÁRIO EMISSÃO X SEQUESTRO DE CO2: UMA NOVA OPORTUNIDADE DE NEGÓCIO PARA O BRASIL, 1994, Rio de Janeiro. Anais... Rio de Janeiro: Companhia Vale do Rio Doce, 1994. p. 156-195. SALAZAR, R. C. Valuing the environment service of permanent forest stands to the global climate: the case of Costa Rica. 1999. 101 f. Thesis (PhD) - Graduate School of Design, Harvard University, Cambridge. SANQUETTA, C. R.; WATZLAWICK, L. F.; BALBINOT, R.; ZILIOTTO, M. A. B.; GOMES, F. S. (Ed.). As florestas e o carbono. Curitiba: UFPR, 2002. 264 p. SANQUETTA, C. R.; BALBINOT, R. Metodologias para determinação de biomassa florestal. In: SANQUETTA, C. R.; BALBINOT, R.; ZILIOTTO, M. A. B. (Ed.). Fixação de carbono: atualidades, projetos e pesquisas. Curitiba: UFPR: Ecoplan, 2004. p. 7793. SANTANA, J. A. S.; SOUTO, J. S. Diversidade e estrutura fitossociológica da caatinga na Estação Ecológica do Seridó-RN. Revista de Biologia e Ciências da Terra, João Pessoa, v. 6, p. 232-242, 2006. SANTOS, J. R.; LACRUZ, M. S. P.; KEIL, M.; KUX, H. J. H.; XAUD, M. R. Análise da imagem JERS-1 para estimativa da fitomassa aérea de florestas tropicais no sudoeste da Amazônia. In: SIMPÓSIO BRASILEIRO DE SENSORIAMENTO REMOTO, 9., 1998, Santos. Anais... Santos: INPE: SELPER, 1998. p. 1517-1523. SCHUMACHER, M. V.; WITSCHORECK, R.; CALDEIRA, M. V. W. E.; WATZLAWICK, L. F. Estoque de carbono em florestas de Pinus taeda L. e Acacia mearnsii De Wild. plantadas no estado do Rio Grande do Sul – Brasil. In: SANQUETTA, C. R.; WATZLAWICK, L. F.; BALBINOT, R. M.; ZILIOTTO, M. A. B.; GOMES, F. dos S. (Ed.). As florestas e o carbono. Curitiba: UFPR: Ecoplan, 2002. p. 141-152.

SCHUMACHER, F. X. A new growth curve and its application to timber-yield studies. Journal of Forestry, Washington, DC, v. 37, p. 819-820, 1939. SILVA, E. O papel dos plantios florestais no sequestro de gás de carbono atmosférico. In: SIMPÓSIO INTERNACIONAL SOBRE ECOSSISTEMAS FLORESTAIS: FOREST´96, 4., 1996, Belo Horizonte. Anais... Belo Horizonte: [s.n.], 1996. p. 52. SILVA, F. H. B. B.; BURGOS, N.; ACCIOLY, L. J. O.; COSTA, T. C. C.; OLIVEIRA, M. A. J. Caracterização dos recursos naturais de uma área piloto do núcleo de desertificação do Seridó, Estados do Rio Grande do Norte e da Paraíba. Rio de Janeiro: Embrapa Solos, 2002. 52 p. (Embrapa Solos. Boletim de Pesquisa e Desenvolvimento, 4). SILVA, G. C. Relações alométricas de dez espécies vegetais e estimativas de biomassas aéreas da caatinga. 1998. 163 p. Tese (Doutorado em Botânica) - Universidade Federal Rural de Pernambuco, Recife. SOLOMON, S.; QIN, D.; MANNING, M.; CHEN, Z.; MARQUIS, M.; AVERYT, K. B.; TIGNOR, M.; MILLER, H. L. (Ed.). Climate change 2007: the physical science bases. Cambridge: Cambridge University Press, 2007. Available at: . Accessed on: 22 May 2007. SOMMER, R.; SÁ, T. D. D. A.; VIELHAUER, K.; ARAÚJO, A. C. D.; FÖLSTER, H.; VLEK, P. L. G. Transpiration and canopy conductance of secondary vegetation in the eastern Amazon. Agricultural and Forest Meteorology, Amsterdam, NL, v. 112, p. 103-121, 2002. STEININGER, M. K. Net carbon fluxes from forest clearance and regrowth in the Amazon. Ecological Applications, Tempe, v. 14, p. 313-322, 2004. TONINI, H.; ARCO-VERDE, M. F.; SCHWENGBER, D. R.; MOURÃO JUNIOR, M. Avaliação de espécies florestais em área de mata no estado de Roraima. Cerne, Lavras, v. 12, p. 8-18, 2006. UHL, C.; BUSCHBACHER, R.; SERRÃO, E. A. S. Abandoned pastures on Eastern Amazonia: I Patterns of plant succession. Journal of Ecology, Oxford, v. 76, p. 663-681, 1988. VALERI, S. V. Exportação de biomassa e nutrientes de povoamentos de Pinus taeda L. desbastados em diferentes idades. 1988. 164 f. Tese (Doutorado em Engenharia Florestal) - Setor de Ciências Agrárias, Universidade Federal do Paraná, Curitiba. VASCONCELOS, S. S. Moisture and nutrient constraints to ecosystem processes in a forest regrowth stand in Eastern Amazonia, Brazil. 2006. 108 f. Dissertation (PhD) - School of Forest Resources and Conservation, University of Florida, Gainesville. VIRGILIO, J. F.; PAREYN, F. G. C. Situação da cobertura florestal do Nordeste. In: SAMPAIO, E. V. S. B.; GIULIETTI, A. M.; VIRGILIO, J.; GAMARRA-ROJAS, C. F. L. (Ed.). Vegetação & flora da caatinga. Recife: Associação Plantas do Nordeste, 2002. p. 41-48. WATZLAWICK, L. F. Estimativa de biomassa e carbono em floresta ombrofila mista e plantações florestais a partir de dados de imagens de satelite Ikonos II. 2003. 119 f. Tese (Doutorado em Ciências Florestais) – Universidade Federal do Paraná, Curitiba WIKIPÉDIA. Geografia do Brasil. Available at: . Accessed on: 10 Oct. 2007. XAUD, H. A. M.; TONINI, H.; LOUREIRO, E. B.; PAOLINETI, V. W.; XAUD, M. R. Biomassa aérea de espécies arbóreas em plantios experimentais em área de floresta de Roraima. Boa Vista: Embrapa Roraima, 2006. (Embrapa Roraima. Comunicado Técnico, 24). ZARIN, D. J.; DAVIDSON, E. A.; BRONDIZIO, E.; VIEIRA, I. C. G.; SÁ, T.; FELDPAUSCH, T.; SCHUUR, E. A. G.; MESQUITA, R.; MORAN, E.; DELAMONICA, P.; DUCEY, M. J.; HURTT, G. C.; SALIMON, C.; DENICH, M. Legacy of fire slows carbon accumulation in Amazonian forest regrowth. Frontiers in Ecology and the Environment, Washington, DC, v. 3, p. 365369, 2005. ZARIN, D. J.; DUCEY, M. J.; TUCKER, J. M.; SALAS, W. A. Potential biomass accumulation in Amazonian regrowth forests. Ecosystems, New York, v. 4, p. 658-668, 2001.

Chapter 5

Emissions of nitrous oxide and nitric oxide from soils in agricultural systems Bruno José Rodrigues Alves, Arminda Moreira de Carvalho, Cláudia P. Jantalia, Beata E. Madari, Segundo Urquiaga, Julio Cezar Franchini dos Santos, Henrique Pereira dos Santos and Claudio José Reis de Carvalho

Abstract: Agriculture is the main source of nitrous oxide (N2O) and nitric oxide (NO) entering the atmosphere, which originate in the soil mainly due to application of nitrogen fertilizers and decomposition of crop residues. Studies conducted by Embrapa’s Agrogases network have increased the amount of information for agricultural areas, although natural areas have also been studied. Most studies aimed at assessing direct emissions of N2O from the soil. It was found that areas planted with soybean crops did not produce enough quantities of N2O to justify including biological N2 fixation as a direct source of N2O from agricultural soils. In situations where the soil is well aerated, N2O emissions may be very low but NO emissions become relevant. Emissions of N2O are stimulated in wetter conditions when nitrogen fertilizers are applied to the soil. In this case, it was observed that the fraction of N applied to the soil emitted as N2O is below the one indicated by the Intergovernmental Panel on Climate Change (IPCC). Taking into account studies conducted in various environments and crops, it was found that, on average, 0.31% of the N applied is emitted as N2O. Analysis of animal excreta showed that direct emissions of N2O, in areas with cattle dung and urine, are lower than those indicated by the IPCC. It was also shown that adopting agroforestry systems to prevent burning of capoeira [scrub regenerating on cleared land] in the Amazon region leads to increased emissions of N2O and NO, due to higher fertilizer use, but also leads to overall reduction of greenhouse gas by eliminating burning, thus making it beneficial to adopt these systems. Analysis of soil and climate variables measured in several studies suggested that lower emissions of N2O may result from the optimal drainage and low content of labile organic matter in the Latosols [US: Oxisols] where the studies were conducted. Keywords: nitrous oxide, nitric oxide, soils, agriculture, animal excreta, forests, no-tillage.

Introduction Atmospheric concentrations of trace greenhouse gases (GHGs) have increased rapidly due to human activities. It is estimated that deforestation and clearing by burning, along with agricultural and forestry activities account for over 80% of Brazil’s contribution to the planet’s greenhouse effect (TEIXEIRA et al., 2006). Globally, 60% of nitrous oxide (N2O) emissions come from agriculture (SMITH et al., 2007), mainly due to use of nitrogen fertilizers, organic fertilizers, decomposition of crop residues, deposition of animal excreta and mineralization of the soil’s organic matter, which increase concentrations of mineral forms of nitrogen in the soil. Depending on soil moisture, these practices can also lead to emission of nitric oxide (NO), which, although not considered a greenhouse gas, still plays an important role in producing ozone (O3) in the troposphere, while reducing the concentration of O3 in the stratosphere, allowing greater incidence of ultraviolet rays on the surface of the planet (DAVIDSON et al., 2001; PRATHER, 1998). N2O is a trace gas known for its role in the greenhouse effect on the planet, and whose molecule has a high global warming potential (GWP). Basically, the GWP of a gas is a function of its

half-life in the atmosphere and its absorption spectrum of infrared rays emitted by the surface of the planet. A molecule of N2O, with an estimated half-life in the atmosphere of 114 years, has a GWP equal to 298 molecules of carbon dioxide (CO2), although the International Panel on Climate Change (IPCC) still uses a value of 310 in analyses of GHG inventories (FORSTER et al., 2007). Because of these characteristics, it is among the most important greenhouse gases, although its current concentration in the atmosphere is very small, around 317 ppbv, a significant increase relative to the estimate for the beginning of last century (approximately 275 ppbv). Most of this increase has occurred during the last 50 years, at an annual rate of 0.7 ppbv (MOSIER et al., 2004). It is expected that the trend toward increasing concentrations of N2O will continue in the next few decades, mainly due to a tendency towards expanding agricultural areas in developing countries and increasing utilization of nitrogen fertilizers (SMITH et al., 2007; ZHANG; ZHANG, 2007). The projected increase in N2O emissions is accompanied by much uncertainty, since the frontier of agricultural expansion is in tropical and subtropical regions, where the availability of data on this process is still very low, especially in Brazil. Results concerning N2O emissions from agricultural soils in Brazil have started to be published recently, most of them concerning agricultural areas on Latosols [US: Oxisols] in the South and Central-West regions of the country (JANTALIA et al., 2008; METAY et al., 2007), suggesting that emissions of this gas are not very high. In Latosols [US: Oxisols] of the Cerrado region, NO emissions seem more significant than those of N2O (CARVALHO et al., 2006). Due to lack of consolidated information on N2O emissions by soil in Brazilian agriculture, the national inventory on this gas is based on use of emission factors provided by IPCC methodologies (EGGLESTON et al., 2006; IPCC et al., 1997), obtained in regions with contrasting climate and soil. Although in the short term this is the only possibility for developing inventories on greenhouse gases, it is quite possible that these methodologies may not be yielding appropriate results for Brazil.

Conceptual model on the system’s main sources of N2O and NO Nitrogen oxides are produced by nitrification and denitrification reactions (Figure 1). Nitrification produces relatively more NO than N2O, and denitrification is the dominant process in the production of N2O (DAVIDSON et al., 1993). Nitrification is favored by the presence of NH4+, by appropriate soil aeration conditions and by higher nitrogen cycling in the system (DAVIDSON et al., 2000). Soil pH is also an important variable, as nitrification rates increase with decreasing soil acidity. Nitrifying bacteria such as Nitrosomonas and Nitrosospira are the main genera that oxidize NH4+ into NO2−, and Nitrobacter is the main genus responsible for the second step in the process − producing NO3 . These organisms are favored by pH conditions above 5, which is common in agricultural areas which are usually corrected with lime. Well-drained soils favor nitrification, as it is an aerobic process, but humidity and temperature are important factors in optimizing it (PAUL; CLARK, 1996).

CHAP 5 - FIGURE 1 Restrição de O2

O2 restriction

Fixação química e biológica

Chemical and biological fixation

Desnitrificação

Denitrification

Temperatura

Temperature

N-nítrico

Nitric N

C-lábil

Labile C

Nitrificação

Nitrification

N-amoniacal

N-ammonia

Figure 1. Conceptual model of N2O production and variables involved. Biological and chemical fixation of N2 leads to increased ammonium concentrations in soil; through the nitrification process, ammonium is oxidized into nitrate, a process favored by aerobic conditions, pH over 5.0 and ammonium available in the soil, being accelerated as temperature increases. When O2 in the soil is limited, nitrate is denitrified, a process dependent on nitrate concentration and on reduction sources (mainly labile C), with higher temperatures accelerating the process. Both NO and N2O are produced in nitrification and denitrification. The thickness of the arrows indicates the relevance of each form produced within each of these processes.

Nitrate (NO3−) can accumulate in soil when its production exceeds demand by microorganisms and plants, thus promoting denitrification reactions (MATSON et al., 1999). Soil moisture, or the degree of water saturation of the soil, is essential in this process, as was demonstrated by Linn and Doran (1984). The denitrification process depends on diffusion of O2 from the atmosphere into the soil, and in locations where this diffusion is restricted and existing O2 is consumed, anaerobic sites are formed in the soil. Thus, the proportion of gases escaping during the denitrification process will also depend on the distance they will need to travel to reach the surface, or soil tortuosity. In conditions where the soil is heavily saturated with water, at close to 80% of its pore volume, high tortuosity decreases the chances of NO emissions by the soil, and N2O and N2 are the predominant forms emitted into the atmosphere. This situation is clearly reversed when the water saturation of the pores is low, normally below 50% (DAVIDSON et al., 1993; DOBBIE et al., 1999). Temperature is also a key factor in soil N2O emissions. For many biological processes, the Q10 temperature coefficient (the rate of increase in reaction speed due to an increase of 10 °C in temperature) can range from 2 to 3, with this range being wider for N2O production in soil (DOBBIE et al., 1999; SKIBA; SMITH, 2000), indicating that small variations in temperature have a major impact on emissions of this gas. When availability of NO3− and NO2− in soil is high, organic carbon is the limiting factor for denitrification reactions (DRURY et al., 1991; MCKENNEY et al., 1995). The results of a study conducted in controlled conditions, in a Dystrophic Latosol [Latosol = US: Oxisol] in the region of Piracicaba, São Paulo (SP), which did not receive plant waste in some years, but which was moistened to about 50% of pore saturation, and fertilized with N, demonstrated that the soil only showed high fluxes of N2O after treatment with a sugar solution (FLORES et al., 2007). These results indicate there is a limitation of the reducing source in the soil which is necessary for the denitrification process. N2O emissions from various production systems in Brazil will depend on how climate, soil type and crop management impact various factors controlling nitrification and denitrification in the soil and release of gas into the atmosphere.

Experimental results on N2O production from soil under various crops Quantification of N2O emissions from soil under agricultural use has been done mostly in closed static chambers. For the studies presented here, chambers were equipped with a metal base, inserted 5 cm to 7 cm into the soil, and kept at the sampling site for the entire evaluation period. The body of the chamber was removable, and when mounted on the base, it created a closed environment, at a height of 10 cm to 12 cm from the soil. Initial ambient air samples were taken as a reference for initial N2O concentration; and, after the predetermined incubation time, a new sample was taken from within the chambers using a manual vacuum pump. An adaptation to the vacuum pump allowed for automatic transfer of the sampled air into a sealed glass vial with a butyl rubber stopper, sealed with silicone film for high vacuum. Samples were sent to Embrapa Agrobiology for analysis of the N2O concentration using a gas chromatograph equipped with electron capture detector and backflush system, with separation of gases using Porapak Q (80–100 mesh) columns. Tests confirmed a linear increase in gas concentration in the chamber at incubation times between 20 and 30 minutes, which ensured reliable N2O flux estimates by taking samples on only two occasions (DOBBIE et al., 1999). Some results on NO emission are also presented, obtained using a dynamic chamber system coupled to a chemiluminescence detector (CARVALHO et al., 2006). In forest areas, studies were conducted with vented closed static chambers, as described by Keller et al. (2005).

Soybean and legumes for green manure Since 2004, soybean crops have been planted on over 20 million hectares, 80% of which in the South and Central-West regions of Brazil. The crop does not receive nitrogen fertilization, and the biological nitrogen fixation (BNF) contribution to soybean is between 70% and 80% (ALVES et al., 2003). On the other hand, green manure legumes are not often planted in the large grain production area, although they are recommended for crop rotations due to their potential for adding N from BNF to the system, which is essential for C sequestration in soil (SISTI et al., 2004). In the first national communication on greenhouse gas and N2O emissions by Brazilian agriculture, BNF represented 26.4 Gg, or 5% of emissions of this gas from agricultural soils – estimates based on the methodology proposed by the IPCC (IPCC et al., 1997), which considered that 1.25% of N derived from the BNF process was emitted as N2O. However, due to lack of evidence proving that N2O was produced by BNF, the recent revision of the methodology proposes disregarding it as a direct source of N2O for agriculture (EGGLESTON et al., 2006). Monitoring of N2O production in soils cultivated with soybean was conducted in Passo Fundo, Rio Grande do Sul (RS), and in Londrina, Paraná (PR), all in Dystroferric Red Latosol [Red Latosol = US: Rhodic Haplustox] having very clayey texture. The crop was managed under notillage and with light, leveling plowing and harrowing. The planting system did not affect production of N2O by the soil (Table 1). Mean daily emissions, calculated from the data in Table 1, varied from 0.67 g N ha−1 to 4.23 g N ha−1 in areas under no-tillage, versus 0.73 g N ha−1 to 4.20 g N ha−1 in areas with soil preparation, with lower values corresponding to the studies conducted in Londrina, Paraná (PR), which may be due to the lack of rain in this location, in the year 2006.

Table 1. Soil N2O emissions for soybean crops managed under no-tillage (NT) and conventional tillage (CT) in the areas of Passo Fundo, Rio Grande do Sul (RS), and Londrina, Paraná (PR). Crop cycle (days)

Soil use

(1)

(2)

Emission of N2O −1 (g N ha )

Passo Fundo, RS (600 g clay kg−1)(3) Soybean CT succession

168 / 187

671 / 658

Soybean NT succession

168 / 187

306 / 791

Soybean CT rotation

168 / 187

706 / 696

168 / 187

500 / 667

Soybean CT rotation

203

148

Soybean NT rotation

203

135

Soybean NT rotation −1 (4)

Londrina, PR (780 g clay kg )

(1)

Period in which N2O fluxes were measured (two numbers refer to two evaluation cycles at the same location).

(2)

Results of the integration of N2O fluxes in the measurement period.

Sources: (3) Jantalia et al. (2008); (4) Data provided by Bruno J.R. Alves, of Embrapa Agrobiology, in 2007.

Higher BNF rates occur in the period from around flowering to pod formation (ALVES et al., 2002), and should be accompanied by higher N2O fluxes, near 100 μg N m−2 h−1, which was not observed in either the study conducted in Londrina, Paraná (PR) (Figure 2), or the one conducted in Passo Fundo, Rio Grande do Sul (RS) (JANTALIA et al., 2008). Regardless of location, N2O fluxes showed no connection with the BNF associated with the crop, reinforcing the conclusion reached by the IPCC (EGGLESTON et al., 2006) to disregard this biological process as a direct source of N2O. In fact, the results obtained show that the largest N2O fluxes occurred at the beginning and end of the soybean crop cycle (Figure 2) and seem to be more associated with the decomposition of residue, especially in the final stage, when leaf senescence occurs (ZOTARELLI, 2000). Decomposition of residue is one of the processes considered by the IPCC (EGGLESTON et al., 2006) as being responsible for direct emissions of N2O from the soil. In this case, use of green manure legumes, which store high amounts of N in the soil, may lead to higher N2O emissions. However, a single study with green manure legumes in maize crops, with no-tillage or with incorporation of residue, conducted in Planaltina, Distrito Federal (DF), on Latosol [US: Oxisol], in the Cerrado region, showed little effects on N2O emissions (CARVALHO, 2005). CHAP 5 - FIGURE 2 Fluxo de N2O

N2O flux

Plantio direto

No tillage

Soja após aveia

Soybean after oats

Preparo convencional


Figure 2. N2O fluxes from a Latosol [US: Oxisol] having very clayey texture, observed throughout the monitoring period of soybean managed in no-tillage and conventional tillage systems, in Londrina, Paraná (PR). Source: Data provided by Bruno J.R. Alves, of Embrapa Agrobiology, in 2007.

The presence of legume residues was not enough to produce high N2O fluxes. NO fluxes, however, were significant (Table 2), suggesting that, in the experiment’s climate and soil conditions, nitrification was the process responsible for production of nitrogen oxides (CARVALHO, 2005). Table 2. Annual NO emissions in a clayey-textured Red-Yellow Latosol [Latosol = US: Oxisol; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox] cultivated with maize with no-tillage or incorporation of residues of two legumes and spontaneous vegetation. Soil preparation system

NO emission (g N ha−1)

Mucuna pruriens

No-tillage

80

Crotalaria juncea

No-tillage

60

Spontaneous vegetation

No-tillage

60

Mucuna pruriens

Incorporation

60

Crotalaria juncea

Incorporation

70

Spontaneous vegetation

Incorporation

40

Cover plants

Source: Adapted from Carvalho (2005).

Grain crops managed with N fertilizer Maize Maize is the second most planted crop in Brazil, occupying an area close to 12 million hectares, taking into account first and second harvest [safrinha]. The use of nitrogen fertilizer is critical to ensure high productivity, urea being the N source used most in the country. Nevertheless, nitrogen fertilizers are one of the major sources of N2O in agriculture. N2O fluxes were measured in areas planted with maize in the South region of Brazil, in Londrina, Paraná (PR), and in Brazil’s Cerrado region, in Santo Antônio de Goiás, Goiás (GO), both areas on clayey-textured Latosols [US: Oxisols], under various combinations of crops and planting systems. Measurements were also made of N2O emissions in an Argisol [US: Ultisol], in the region of Seropédica, Rio de Janeiro (RJ), with increasing doses of N in the form of ammonium sulfate. Measurements were made during the rainy season for all cases, but in Seropédica, Rio de Janeiro (RJ), and in the Cerrado area (Santo Antônio de Goiás), there was complementary irrigation. Higher N2O fluxes were observed in no-tillage areas after fertilization events, as seen in Figure 3, in the study conducted in Londrina, Paraná (PR). Emissions of N2O, however, were below 500 g N ha−1 in maize production systems in Latosol [US: Oxisol] areas, taking into account the period corresponding to the crop cycle and, in some cases, several weeks after harvest. Higher emissions were observed in the Argisol [US: Ultisol] in Seropédica, where the addition of only 50 kg N ha−1 resulted in emissions exceeding those observed in Latosols [US: Oxisols] (OLIVEIRA, 2009). In the region of Londrina, Paraná (PR), for two consecutive years, N fertilizer proportions were calculated by comparing fertilized areas with areas under the same soil preparation system but

without addition of N, which did not reach 0.1% of the total applied. In the area in Santo Antônio de Goiás, Goiás (GO), a higher emission of N2O was observed, with an estimated loss of 0.22% of applied N, but it should be mentioned that the maize crop was managed with irrigation. Low emissions observed in Latosols [US: Oxisols] suggest that denitrification events must occur for short periods of time after rainfall (Figure 3), as a result of these soils’ high drainage and the high rates of evaporation and evapotranspiration of tropical regions. These conditions are more conducive to nitrification, which is more associated with the production of NO. In fact, in a clayey-textured Red Latosol [US: Rhodic Haplustox], planted with maize, fertilized with 60 kg N ha−1 broadcast, it was not possible to detect production of N2O from the soil, even after irrigation events. However, production of NO was observed (CARVALHO et al., 2006) when the soil was moistened, which ensured favorable conditions for nitrifying activity.

CHAP 5 - FIGURE 3 Fluxo de N2O

N2O flux

Plantio direto

No tillage

MIlho após tremoço

Maize after lupin

Milho após aveia fertilizado com ...

Maize after oats fertilized with ...

Aplicação de fertilizante

Fertilizer application



Data

Date

Figure 3. N2O fluxes in no-tillage and conventional tillage systems in a Dystroferric Red Latosol [Red Latosol = US: Rhodic Haplustox] cultivated with maize, in Londrina, Paraná (PR). Source: Jantalia et al. (2007).

Upland rice Upland rice is a designation for old cultivars (e.g., cv IAC 47 and cv Rio Paranaíba) commonly used in opening new areas; it is tall, with low tillering and a mean yield potential of 4.5 t ha−1. In 1996, new, long-grain rice cultivars were released, which are more productive and more responsive to fertilization (e.g., cv Primavera, cv Maravilha). These new cultivars were named upland rice. Upland rice cultivation in Brazil covers nearly 2 million hectares (WANDER, 2006). Typically, upland rice cultivation requires 90 kg N ha−1. Production of nitrous oxide by upland rice crops was evaluated in a study conducted in a Dystrophic Red Latosol [Red Latosol = US: Rhodic Haplustox], with clayey-loam texture, in an area located at Embrapa Rice and Beans, Santo Antônio de Goiás, Goiás (GO). Fertilization of rice translated into in large fluxes of N2O, only with broadcast application of 60 kg N ha−1 of urea (Figure 4) coinciding with heavy rains. This effect from fertilization, however, was confined to the first five days after application. Integrating N2O fluxes for the 133-day period, 354 g N-N2O ha−1 were produced in the treatment fertilized with urea, versus 235 g N-N2O ha−1 in the control treatment which was not fertilized. Thus, 119 g N-N2O ha−1 was produced by applying

90 kg N ha−1 to rice crops, in the form of urea, which would represent a loss of 0.13% of the N added as fertilizer. A similar result was found by Metay et al. (2007) at the same location, studying a rice crop. In this study, NO production was also high, confirming that conditions are more favorable for nitrification in Latosols [US: Oxisols] of warmer regions. Wheat Wheat production areas are predominantly located in the South region of Brazil, ranging between 1.5 and 2.8 million hectares, in recent years. Use of N is not high, on average 40 kg N ha−1 when fertilized.


N2O flux

Controle (sem fertilizante)

Control (no fertilizer)

Fertilizado

Fertilized

Aplicação do fertilizante


Data de amostragem

Sampling date

Figure 4. N2O fluxes observed in samplings taken from upland rice crops under no-tillage in a Latosol [US: Oxisol], in Santo Antônio de Goiás, Goiás (GO). Source: Costa et al. (2007).

In the area studied, the wheat crop was planted in late July 2006, and fertilized with 50 kg N ha−1. N2O fluxes responded to increased availability of N in the soil resulting from the application of nitrogen fertilizer, with fluxes exceeding 60 μg N-N2O m−2 h−1. The effect of fertilization on N2O emissions lasted for 15 days, with no differences thereafter in relation to the unfertilized control area. Integrating the data for the period of wheat crop development, 309 g N ha−1 were produced in the fertilized area, versus 243 g N ha−1 in the control area. The difference, divided by the amount of N added (50 kg N ha−1), resulted in a loss percentage of 0.13% of fertilizer N in the form of N2O. Third-harvest common bean Areas occupied with common bean [Phaseolus vulgaris] crops cover approximately 4 million hectares in Brazil, and the areas where higher doses of fertilizer are used are in the Cerrado, using center pivot irrigation. In this case, common practices include application of nitrogen fertilizers without the use of rhizobia inoculant, rendering BNF negligible in the crop. Measurements of N2O fluxes in a common bean area, planted in Dystrophic Red Latosol [Red Latosol = US: Rhodic Haplustox] in Santo Antônio de Goiás, Goiás (GO), were higher after N fertilization followed by irrigation. Differences in fluxes in the fertilized area (295 g N-N2O ha−1) were observed compared to the control area (198 g N-N2O ha−1) (Figure 5), although of a small magnitude, which explains the small amounts of N volatilized as N2O, or 0.12% of the N applied.

CHAP 5 - FIGURE 5

Fluxo de N2O

N2O flux

Controle

Control

Fertilizado

Fertilized

Aplicação do fertilizante


Maio

May

Ago.

Aug.

Out.

Oct.

Mês

Month

Figure 5. N2O fluxes observed during sampling in common bean [Phaseolus vulgaris] crop under no-tillage in a Cerrado Red Latosol [Red Latosol = US: Rhodic Haplustox], in Santo Antônio de Goiás, Goiás (GO). Source: Madari et al. (2007).

Sugarcane Areas cultivated with sugar cane are increasing in Brazil, as a result of high demand and incentives for use of ethanol as an alternative fuel to gasoline. It is estimated that, in the next few years, planted areas will reach over 10 million hectares. Although characterized as an environmentally friendly fuel, little information is available about the impact of the sugarcane production system on N2O emissions. To evaluate only the effect of N fertilization on sugarcane crops, at ratoon stage, a study was conducted at Embrapa Agrobiology’s experimental area, in the municipality of Seropédica, Rio de Janeiro (RJ), in a Planosol, with application of 60 kg N ha−1 in the form of broadcast urea. An increase in N2O fluxes was observed in plots fertilized with N, when compared to the control area. Throughout the entire period, emission of N in the form of N2O reached 131 g N ha−1 in the fertilized area, and 36 g N ha−1, in the control area. Unlike in previous studies, the N2O fluxes in the fertilized area remained higher those of the non-fertilized area for a longer time (Figure 6), which can be associated with imperfect drainage of the soil in question. The study was conducted over a period of 48 days, and losses of N-N2O from fertilizer reached 0.16%. However, they could be higher, due to the behavior shown in Figure 6, which could be explained by the poor drainage of the soil in question. As sugarcane crops can benefit from BNF, with yields similar to sugarcane fertilized with N (URQUIAGA et al., 1992), the fact that the BNF-dependent sugarcane area produced approximately 4 times less N2O than the fertilized area highlights the importance of the BNF process to mitigate the greenhouse effect.


N2O flux

Sem adição de N

No N added

Aplicado no início do estudo

Applied at beginning of study

Date

Date

−1 Figure 6. Variations in N2O fluxes of soil under sugarcane, fertilized with 60 kg N ha in the form of urea, versus not fertilized.

Source: Data provided by Bruno J.R. Alves, of Embrapa Agrobiology, in 2007.

Table 3 summarizes the results of estimates of the proportions of N applied as fertilizer and crop residues, emitted as N2O, observed for maize, upland rice, wheat, sugarcane and third-harvest common bean (with sprinkler irrigation). Also taking into account the results obtained by Jantalia et al. (2008), which include crop residues, 21 estimates were obtained in Brazil The average of these numbers could be considered a first approximation for a N2O direct emission factor for Brazil. Table 3. Estimates of N2O emissions from soil and proportion of N emitted from fertilizer as N2O (PNENO) for various crops managed under prepared soil (PS), no-tillage (NT) and conventional tillage (CT), in various locations, treated with different sources of N and different rates of N fertilizer. Evaluation cycle(1) (days)

Soil use

Fertilizer N (source; kg N ha−1)

Londrina, PR SP maize rotation (year 1 and 2)(3) NT maize rotation (year 1 and 2)(3)

Soil type

PNENO(2) (%)

Dystroferric Red 136 / 141

Urea; 80

136 / 141

Urea; 80

140

Urea; 80

Latosol [Red Latosol = US: Rhodic

0.08–0.03

Haplustox] 0.12–0.08

Santo Antônio de Goiás, GO NT maize succession(4)

Dark Red

0.24

Latosol NT rainfed rice (year 1(5) and 2(4)) NT irrigated beans(6)

[US: Rhodic Haplustox]

133 / 132

Urea; 90

0.13–0.14

149

Urea; 80

120

Urea; 50

Red Yellow Argisol

0.05

[Argisol = US: Ultisol]

0.32

0.12

Seropédica, RJ PS maize(7) (7)

120

Urea; 100

(7)

120

Urea; 150

48

Urea; 60

PS maize PS maize

(4)

Sugarcane

0.44 Planosol

0.16

Passo Fundo, RS NT wheat rotation(4) NT wheat/soybean (year 1 and 2)(8) CT wheat/soybean (year 1 and 2)(8)

137 1 year

1 year

Urea; 40

0.13

Fert + Res;

0.56–0.81

120–116 Fert + Res;

0.47–0.52

126–133 Dark Red

NT wheat/maize(8)

1 year

CT wheat/maize(8)

1 year

CT wheat/sorghum(8)

1 year

CT wheat/sorghum(8)

1 year

Fert + Res;

Latosol

162

[US: Rhodic Haplustox]

0.41

Fert + Res;

0.70

141 Fert + Res;

0.24

193 Fert + Res;

0.29

193

Average (confidence interval)

0.31 (0.20–0.48)

(1) Period in which N2O fluxes were measured (two numbers refer to two cycles of the evaluated crop); (2) Percentage of fertilizer N emitted as N2O, calculated as the difference between emissions in the fertilized area and the non-fertilized area, divided by the total N added in the fertilizer.

Source: (3) Jantalia et al. (2007); (4) Data provided by Bruno J.R. Alves, of Embrapa Agrobiology, in 2007; (6) Madari et al. (2007); (7) Oliveira (2009); (8) Jantalia et al. (2008).

(5)

Costa et al. (2007);

Analysis of the frequency distribution shows a lack of normality, being right-skewed. Transforming the data into a natural logarithm generated a significant normal distribution. The average direct emission factor was estimated according to Olsson (2005). The confidence interval for the estimate of F was calculated using Cox’s method, modified (ZHOU; GAO, 1997), using the tabulated value of t for (n-1) degrees of freedom (OLSSON, 2005). The mean N2O direct emission factor was estimated at 0.31% with a confidence interval ranging from 0.20% to 0.48% (Table 3), or two thirds to a little over one and half times the mean value. Mean N2O direct emission factor for fertilizers The N2O direct emission factor found for various crops was estimated at 0.31%, assuming that all the added N remained in place, i.e., assuming there were no losses. In the recent review of the IPCC methodology (EGGLESTON et al., 2006), the N2O direct emission factor is 1% of the N deposited in the soil, and, like the estimate made in this study, this is a number obtained without considering losses in the system. Thus, it is possible that the lower direct emission factor estimated here, when compared to that of the IPCC, is a result of higher N losses, which will add to so-called indirect emissions of N2O. According to the IPCC (EGGLESTON et al., 2006), ammonia volatilization and nitrate leaching are loss routes that should be considered when calculating indirect emissions of N2O, the first estimated at 10%, and the latter at 30% of applied N, for locations where rainfall exceeds the soil’s

water storage capacity. That is the main limitation preventing better assessment of what occurs under conditions of the studies conducted in Brazil. There is no data in the literature regarding N losses by nitrate leaching, and there is little information on fertilizer ammonia volatilization, making it very difficult and uncertain to establish a global average value to estimate how much N from applied fertilizer is subject to losses through these processes. Lara Cabezas et al. (1997) measured losses of approximately 70% of N applied in the form of broadcast urea under no-tillage, whereas Hungria et al. (2006) measured losses of approximately 15% to 25% of N added as urea. These few data points suggest greater losses of N by ammonia volatilization in no-tillage systems when urea is applied to the surface; but when fertilizer is incorporated, losses by ammonia volatilization are minimal (LARA CABEZAS et al., 1997). In a study with fertilizer marked with isotope 15 N, Alves et al. (2006) showed that in a Cerrado Red Latosol [Red Latosol = US: Rhodic Haplustox], broadcast application of ammonium sulfate on maize and cotton crops resulted in total fertilizer N losses of around 30%. However, this source is not very representative of grain production systems in Brazil. Moreover, losses by volatilization of ammonia from sulfate are much smaller than those observed for urea (LARA CABEZAS et al., 1997). The lack of information on such losses makes the values recommended by the IPCC (EGGLESTON et al., 2006) the best that can be used. In any case, the mean emission factor estimated at 0.31% would be at the lower limit of the uncertainty interval proposed by the IPCC for that parameter (0.3% to 3%). However, it would be 3 times lower than the mean value of 1% used in inventories.

Animal excreta Areas under grazing pasture in Brazil occupy over 200 million hectares and, due to their predominantly extensive use, emissions occur mainly from animal excreta. According to Mosier et al. (1998), animal production systems may account for up to 50% of total N2O emissions attributed to agriculture. Fertilizers and cattle feces promote increases in mineral N concentrations in the soil and thus benefit nitrification and denitrification processes, which are responsible for producing N2O. Cattle excreta return to the system directly in extensive systems; however, in feedlots, excreta are applied to crops, such as occurs in pig farms, and the impact on N2O emissions also needs to be quantified. In the case of cattle, studies were conducted to evaluate the effect of applying urine and feces in pasture soils. In one of them, N2O fluxes were evaluated in pastures of Brachiaria brizantha cv Marandu, grown in a Planosol in the region of Seropédica, Rio de Janeiro (RJ) for milk production, immediately after the removal of the animals. Areas with and without addition of urine (2 liters in 0.19 m2), containing (~70 g N m−2), were established. N2O fluxes exceeded 1,000 μg N m−2 h−1 due to application of urine, and remained above those of the control area for a period of two weeks, returning to levels observed at the beginning of the study (JANTALIA et al. 2006). The soil, which had imperfect drainage, favored N2O production, especially in this study, which was conducted during a very rainy summer. Studies conducted in an Argisol [US: Ultisol], with better drainage, in a split-plot experiment, resulted in much lower N2O emissions (Figure 7). The study was conducted during autumn, with little rain, which would explain the observed behavior, in addition to the fact that the urine contained a third of the N of the urine used in the rainier season in the study conducted in a Planosol. The amount of N2O emitted during the study with urine application in the Planosol was

10 times greater than that observed in the study conducted in the Argisol [US: Ultisol], even considering the much longer evaluation period for the Argisol [US: Ultisol] (Table 4). Although factors such as soil type, time considered for integration of fluxes and quantity of N in the urine all contributed differently to the results, rainfall must have been the key to explain N2O emissions from urine. It is evident that emissions from feces are well below those of urine, which can be explained by the slow release of N from feces, not exceeding 5% of the total contained therein (FERREIRA et al., 1995).


N2O flux

Controle

Control

Fezes

Feces

Urina

Urine

Data

Date

Figure 7. N2O fluxes in pasture areas containing dairy cattle urine and feces, and control area without excreta. Source: Lessa et al. (2007).

Table 4. Estimates of N2O emissions from soil and of proportion of N from cattle and pig excreta emitted as N2O (PNENO) for various land uses, at various locations.

Soil use

Evaluation cycle(1) (days)

Excreta (type; kg N ha−1)

Emission(2) of N2O2 (g N ha−1)

PNENO(3) (%)

3,020

0.4

271

0.07

98

0

402

0.20

508

0.25

Seropédica, RJ(4) Brachiaria (summer)

30

Brachiaria (autumn)

50

Brachiaria (autumn)

50

Bovine urine; 700 Bovine urine; 250 Cattle feces; 820

Santa Maria, RS(5) NT bare soil

28

RT bare soil

28

Pig excreta; 154 Pig excreta; 154

(1)

Period in which N2O fluxes were measured.

(2)

Result of the integration of fluxes in the N2O flux measurement period.

(3) Percentage of fertilizer N lost as N2O, calculated as the difference between emissions in areas with and without excreta, divided by the total N added in the excreta. (4) Studies with application of excreta on Brachiaria pastures were conducted in the summer (JANTALIA et al., 2006) and autumn (LESSA et al., 2007), in Seropédica, Rio de Janeiro (RJ).

(5)

Studies with pig manure were conducted in Santa Maria, Rio Grande do Sul (RS) in areas with bare soil, previously managed with notillage (NT) and reduced tillage (RT) (GIACOMINI et al., 2006).

Considering the study period, the quantities of N2O emitted and the quantity of N in urine and feces, in both studies, losses of N as N2O were estimated to be 0.4% of the total deposited as urine, during rainy seasons, and 0.07% for drier conditions (Table 4). As emissions of N2O were not stimulated by the presence of feces, it was considered that, in the conditions of the study, which included depositions in drier periods, such emissions are not relevant. Studies with pig excreta (GIACOMINI et al., 2006) resulted in emission factors ranging from 0.20% and 0.25% of the applied N, being higher when excreta were applied in soil managed under no-tillage. Generally, the emissions of N2O resulting from deposition of N in excreta found in these studies are well below those proposed by the IPCC (EGGLESTON et al., 2006), of 2%. The studies presented are preliminary and need to be conducted on soils and climates that are more representative of the area under use by the national cattle industry. It is also urgent for an assessment to be made on N losses, especially by volatilization of urine ammonia, for a more complete study aimed at determining a more accurate N2O emission factor.

Secondary forests: recovery after land use Secondary forests are very important as modified biomes due to the large area they occupy, mainly in the Amazon region. In this region, 16% of the total original surface has already been converted in some way, resulting in 30% to 50% of areas occupied by plant succession at various stages of recovery after agricultural use. As the system for preparing areas in the region typically includes slashing and burning, together with the process of combustion of forest biomass, approximately 96% of the N contained in the above-ground biomass is emitted, resulting in soils poorer in this nutrient. In addition to the significant area occupied by secondary vegetation in the Amazon, this vegetation also is important because it serves as the basis for developing agroforestry systems with low inputs, typical in family farming. Using chronosequences including secondary vegetation aged 3 to 70 years after agricultural use and fragments of primary forest, it has been shown that, with increasing vegetation age, levels of mineral nitrogen (NO3− and NH4+) in the soil tend to increase, coinciding with increased emission of N2O, which is more pronounced during the rainy season (Figure 8).

CHAP 5 - FIGURE 8 Emissão de N2O

N2O emission

Úmido

Wet

Seco

Dry

Floresta

Forest

Avançada

Advanced

Intermediária

Intermediate

Jovem

Young

Pimental

Pepper area

Idade da parcela

Plot age

Figure 8. Emissions of N2O from soil in secondary forest chronosequence in the municipality of São Francisco do Pará, Pará (PA), during wet and dry seasons. The plots were classified according to recovery time as young (3–6 years), intermediate (10 to 20 years) and advanced (40 to 70 years). The pepper area refers to an area cultivated with piper nigrum [black pepper; pimenta-do-reino] and abandoned three years earlier. Source: Davidson et al. (2007).

Apparently, there is a change in the nitrogen cycle as secondary forests age, where soils of degraded areas that are poorer in this resource have more conservative cycles and, therefore, lower losses of N via emission of N2O, whereas soils under older forests have more open cycles and are more prone to emitting N2O (DAVIDSON et al., 2004, 2007). Alternative agriculture systems, without burning (Tipitamba) Approximately 75% of Brazilian greenhouse gas emissions are associated with land use in various exploitation systems involving deforestation and with biomass burning. Fire is used as a means to clear areas for planting and as a method for nutrient input and partial correction of soil acidity (SÁ et al., 2007). Alternatives to the traditional slash and burn method have been sought, initially for use in family and medium-scale farming, which involve substituting burning of biomass by mechanized grinding, resulting in a process which is currently in the validation stage (DENICH et al., 2004, 2005; SÁ et al., 2007). The balance of GHG emissions from the cutting and grinding system showed that there was an increase of 50% in N2O and NO emissions, basically resulting from the use of nitrogen fertilizers and the burning of fuel by the tractor. However, considering all emissions resulting from the process in one crop cycle, total emissions (in CO2 equivalents), resulting from the process using the plant burning method (Table 5) were at least five times higher than when using the alternative biomass grinding system (DAVIDSON et al., 2006). Table 5. Calculated values of N2O and NO emissions resulting from combustion of biomass using published emission factors (grams of gaseous compound emitted per kilogram of combustible material consumed) for tropical forest biomass (ANDREAE; MERLET, 2001), assuming that secondary vegetation has 100 Mg of dry matter per hectare and a combustion efficiency of 93% (SOMMER et al., 2004). Emission factor N2O NO

0.20 1.6

Emission by burning

CO2 equivalents

−1

5,600 kg CO2 ha−1

−1

–

19 kg N2O ha

130 kg NO ha

Source: Sommer et al. (2004).

Importance of soil and climate variables for NO and N2O fluxes

Application of nitrogen via fertilizers is a common practice in intensive and mechanized agriculture, except in the case of soybean, where nitrogen fertilizers are not recommended (VARGAS et al., 2004). Emissions of nitrogen gases from agricultural soils should be strongly associated with fertilization, but some soil variables, such as moisture and aeration, expressed as the percentage of pore space saturated with water (% Sw), in addition to temperature, significantly contribute to enhance the effect of this practice. A strong correlation between % Sw and soil N2O fluxes (Figure 9) was found in some of the studies discussed earlier in this chapter. However, as mentioned earlier, greater fluxes occur during a short period of time and overall emission is relatively small; in most cases, much lower than that estimated by the direct emission factor proposed by the IPCC (EGGLESTON et al., 2006).


N2O emission

% Epsa

% Sw

Figure 9. Variations in soil N2O fluxes according to percent of soil pore space saturated with water (% Sw) in pasture soils. Source: Lessa et al. (2007).

In Latosols [US: Oxisols], with high contents of iron and aluminum oxides, hydroxides and oxy-hydroxides (CARVALHO et al., 1995), the soil structure with microaggregation promotes pronounced aeration and drainage, which limits denitrification and formation of N2O. In this situation, in presence of water, either from rainfall or irrigation, the nitrogen gas predominantly produced in the soil should be NO. In a recently fertilized Cerrado Latosol [Latosol = US: Oxisol], it was possible to measure NO fluxes (6.9 μg N m−2 h−1 to 19.4 μg N m−2 h−1), whereas N2O production was not detected (CARVALHO et al., 2006). This same situation was observed by Metay et al. (2007), who also presented low estimates for N2O production and high NO fluxes in a Cerrado Latosol [Latosol = US: Oxisol] cultivated with rice. The results presented in Table 1 confirm the tendency towards relatively low N2O fluxes in this region, and in other Latosol [US: Oxisol] areas in Brazil. Davidson et al. (2001) presented annual N2O values ranging from 1.4 kg N ha−1 year−1 to 2.4 kg N ha−1 year−1 for the Amazon, whereas in Cerrado areas, mean annual fluxes are close to zero (< 0.4 kg N ha−1 year−1), as seen in some of the results in Table 1. These low annual N2O fluxes should result from the predominant soil aeration and drainage conditions in the Cerrado (Latosols [US: Oxisols]), under natural vegetation or a management system that favors aggregation. Good soil aggregation, associated with the seasonal pattern of this region’s climate, where the dry season lasts for about six months, promotes low mean annual N2O fluxes in the Cerrado (CARVALHO et al., 2006; PINTO et al., 2002). In addition to issues of moisture and N availability, availability of readily metabolizable organic matter in the soil should also be emphasized. Tropical climate favors high activity in the soil’s microbial biomass, and thus, consumption of simpler reduction sources should be intense, possibly limiting some processes such as denitrification. High soil N2O fluxes after treatment with sugar solution (FLORES et al., 2007), illustrate this situation.

Final considerations In Brazil there are few estimates based on field measurements for agricultural NO emissions. Understanding the dynamics between cultivation practices (e.g., fertilizer formula, tillage, irrigation, etc.) and NO fluxes (including frequency of high fluxes) is also fundamental for proposing strategies to mitigate greenhouse gas emissions. The use of N2 fixing legumes as plant species for green manure and soil cover (e.g., Mucuna pruriens and Crotalaria juncea) in agricultural systems may represent the incorporation of up to 230 kg N ha−1 (CARVALHO et al., 1995), and reduce the use of mineral fertilizers, which have a high impact on greenhouse gas production (both in the process of their synthesis, and in their use in agriculture). It is important to assess the magnitude of greenhouse gas production resulting from conventional use of legumes as green manure or ground cover. Associated to these factors, studies on the quality of residues, or their chemical composition in terms of their content of lignin, cellulose and other phenolic compounds, should be related to the gas emission potential of various plant materials. No information exists on greenhouse gas emissions for planted forests in Brazil, or for other crops under expansion, such as those with potential for biofuel programs, including sugarcane. Obviously, continuing to monitor agricultural and natural systems is critical to improve the information base on greenhouse gases. In addition to the variables commonly measured in studies on greenhouse gases, attention must be given to the need to monitor other variables such as mineralization potential and nitrification potential in the soil, enzyme activities related to C and N cycles, and characterization of the system in order to obtain essential parameters for mathematical models and simulators (Chapter 6 of this book). The use of stable isotopes as tools to evaluate C and N mechanisms and fluxes in the predominant land use systems has the potential to improve understanding of the processes involved in the soil-plant-atmosphere system.

References ALVES, B. J. R.; BODDEY, R. M.; URQUIAGA, S. The success of BNF in soybean in Brazil. Plant and Soil, The Hague, v. 252, p. 1-9, 2003. ALVES, B. J. R.; ZOTARELLI, L.; BODDEY, R. M.; URQUIAGA, S. Soybean benefit to a subsequent wheat crop in a cropping system under zero tillage. In: INTERNATIONAL SYMPOSIUM ON NUCLEAR TECHNIQUE IN INTEGRATED PLANT NUTRIENT, WATER AND SOIL MANAGEMENT, 2000, Vienna, AT. Abstracts... Vienna, AT: IAEA, 2002. p. 87-93. ALVES, B. J. R.; ZOTARELLI, L.; FERNANDES, F. M.; HECKLER, J. C.; MACEDO, R. A. T. de; BODDEY, R. M.; JANTALIA, C. P.; URQUIAGA, S. Fixação biológica de nitrogênio e fertilizantes nitrogenados no balanço de nitrogênio em soja, milho e algodão. Pesquisa Agropecuária Brasileira, Brasília, DF, v. 41, p. 449-456, 2006. ANDREAE, M. O.; MERLET, P. Emission of trace gases and aerosols from biomass burning, global biogeochemical cycles. Global Biogeochemical Cycles, Washington, DC, v. 15, p. 955–966, 2001. CARVALHO, A. M. de. Uso de plantas condicionadoras com incorporação e sem incorporação no solo: composição química e decomposição dos resíduos vegetais; disponibilidade de fósforo e emissão de gases. 2005. 199 f. Tese (Doutorado em Ecologia) - Universidade de Brasília, Brasília, DF. CARVALHO, A. M. de; BUSTAMANTE, M. M. C.; KOZOVITS, A. R.; MIRANDA, L. N. de; VIVALDI, L. J.; SOUSA, D. M. Emissões de NO e N2O associadas à aplicação de uréia sob plantio direto e convencional. Pesquisa Agropecuária Brasileira, Brasília, DF, v. 41, p. 679-685, 2006.

CARVALHO, A. M. de; FAGERIA, N. K.; KINJO, T.; PEREIRA, I. P. de. Resposta do feijoeiro à aplicação de fósforo em solos dos Cerrados. Revista Brasileira de Ciência do Solo, Campinas, v. 19, p. 61-67, 1995. COSTA, A. R.; MADARI, B. E.; MACHADO, P. L. O. A.; JANTALIA, C. P.; URQUIAGA, S.; BODDEY, R. M.; ALVES, B. J. R. Emissão de óxido nitroso derivado da cultura de arroz de terras altas sob plantio direto em um Latossolo dos Cerrados de Goiás, GO. In: CONGRESSO BRASILEIRO DE CIÊNCIA DO SOLO, 31., 2007, Gramado. Anais... Porto Alegre: Sociedade Brasileira de Ciência do Solo, 2007. 1 CD-ROM. DAVIDSON, E. A.; BUSTAMANTE, M. M. C.; PINTO, A. S. Emissions of nitrous oxide and nitric oxide from soils of native and exotic ecosystems of the Amazon and Cerrado Regions of Brazil. The Scientific World Journal, Newbury, v. 1, p. 312– 319, 2001. Suppl 2. DAVIDSON, E. A.; CARVALHO, C. J. R.; FIGUEIRA, A. M.; ISHIDA, F. Y.; OMETTO, J. P. H. B.; NARDOTO, G. B.; SABÁ, R. T.; HAYASHI, S. N.; LEAL, E. C.; VIEIRA, I. C. G.; MARTINELLI, L. A. Recuperation of nitrogen cycling in Amazonian forests following agricultural abandonment. Nature, London, GB, v. 447, p. 995-998, 2007. DAVIDSON, E. A.; KELLER, M.; ERICKSON, H. E.; VERCHOT, L. V.; VELDKAMP, E. Testing a conceptual model of soil emissions of nitrous and nitric oxides. BioScience, Washington, DC, v. 50, p. 667-680, 2000. DAVIDSON, E. A.; MATSON, P. M.; VITOUSEK, R.; RILEY, R.; DUNKIN, K.; GARCÍA-MÉNDEZ, G.; MAASS, J. M. Processes regulating soil emissions of NO and N2O in a seasonally dry tropical forest. Ecology, Tempe, v. 74, p. 130-139, 1993. DAVIDSON, E. A.; SÁ, T. D. A.; CARVALHO, C. J. R.; FIGUEIREDO, R. O.; ISHIDA, F. Y.; SABÁ, R. T.; SANTOS, E. B.; FREITAS, J. F. B. An integrated assessment of soil and fire emissions of greenhouse gases from slash-and-burn and chop-and-mulch agriculture in the Eastern Amazon. In: SCIENCE TEAM MEETING, 10., 2006, Brasília, DF. Abstracts... Manaus: LBA-ECO, 2006. p. 19. DAVIDSON, E. A.; SÁ, T. D. A.; CARVALHO, C. J. R.; FIGUEIREDO, R. O.; ISHIDA, F. Y.; SABÁ, R. T.; SANTOS, E. B.; FREITAS, J. F. B. Nitrogen and phosphorus limitation of biomass growth in a tropical secondary forest. Ecological Applications, Tempe, v. 4,p. S150-S163, 2004. DENICH, M.; VIELHAUER, K.; KATO, M. S. de A.; BLOCK, A.; KATO, O. R.; SÁ, T. D. de A.; LÜCKE, W.; VLEK, P. L. G. Mechanized land preparation in forest-based fallow systems: the experience from Eastern Amazonia. Agroforestry Systems, Holland, v. 61, p. 91-106, 2004. DENICH, M.; VLEK, P. L. G.; SÁ, T. D. de A.; VIELHAUER, K.; LÜCKE, W. A concept for the development of fire-free fallow management in the Eastern Amazon, Brazil. Agriculture, Ecosystems & Environment, Amsterdam, NL, v. 110, p. 43-58, 2005. DEPLEDGE, J.; LAMB, R. (Ed.). Caring for climate: a guide to the Climate Change Convention and the Kyoto Protocol. Bonn: Climate Change Secretariat-UNFCCC, 2005. DOBBIE, K. E.; MCTAGGART, I. P.; SMITH, K. A. Nitrous oxide emission from intensive agricultural systems: variations between crops and seasons, key driving variables, and mean emission factors. Journal of Geophysical Research, Washington, DC, v. 104, p. 26891-26899, 1999. DRURY, C. F.; MCKENNEY, D. J.; FINDLAY, W. I. Relationships between denitrification, microbial biomass and indigenous soil properties. Soil Biology & Biochemistry, Oxford, v. 23, p. 751-755, 1991. EGGLESTON, H. S.; BUENDIA, L.; MIWA, K.; NGARA, T.; TANABE, K. (Ed.). 2006 IPCC guidelines for national greenhouse gas inventories. Hayama: Institute for Global Environmental Strategies, 2006. FERREIRA, E.; URQUIAGA, S.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M. Perdas de N Derivado das fezes bovinas depositadas na superfície do solo. In: CONGRESSO ANUAL DA SOCIEDADE BRASILEIRA DE ZOOTECNIA, 32., 1995, Brasília, DF. Anais... Viçosa: Sociedade Brasileira de Zootecnia, 1995. p. 125-126. FLORES, R. A.; COLLIER, L. S.; URQUIAGA, S.; BODDEY, R. M.; ALVES, B. J. R. Impacto de algumas variáveis do sistema soloplanta na produção de N2O do solo. In: CONGRESSO BRASILEIRO DE CIÊNCIA DO SOLO, 31., 2007, Gramado. Anais... Porto Alegre: SBCS, 2007. 1 CD-ROM. FORSTER, P.; RAMASWAMY, V.; ARTAXO, P.; BERNTSEN, T.; BETTS, R.; FAHEY, D. W.; HAYWOOD, J.; LEAN, J.; LOWE, D. C.; MYHRE, G.; NGANGA, J.; PRINN, R.; RAGA, G.; SCHULZ, M.; DORLAND, R. van. Changes in atmospheric constituents and in

radiative forcing. In: SOLOMON, S.; QIN, D.; MANNING, M.; CHEN, Z.; MARQUIS, M.; AVERYT, K. B.; TIGNOR, M.; MILLER, H. L. (Ed.). Climate change 2007: the physical science basis: contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press, 2007. GIACOMINI, S. J.; JANTALIA, C. P.; AITA, C.; URQUIAGA, S.; ALVES, B. J. R. Emissão de óxido nitroso com a aplicação de dejetos líquidos de suínos em solo sob plantio direto. Pesquisa Agropecuária Brasileira, Brasília, DF, v. 41, p. 1653-1661, 2006. HUNGRIA, M.; FRANCHINI, J. C.; CAMPO, R. J.; CRISPINO, C. C.; MORAES, J. Z.; SIBALDELLI, R. N. R.; MENDES, I. C.; ARIHARA, J. Nitrogen nutrition of soybean in Brazil: contributions of biological N2 fixation and N fertilizer to grain yield. Canadian Journal of Plant Science, Ottawa, CA, v. 46, p. 927-939, 2006. IPCC. International Panel on Climate Change; UNEP. United Nations Environment Programme; OECD. Organization for Economic Co-Operation and Development; IEA. International Energy Agency. Revised 1996 guidelines for national greenhouse gas inventories. Paris, FR, 1997. v. 3. JANTALIA, C. P.; ALVES, B. J. R.; URQUIAGA, S.; FRANCHINI, J.; TORRES, E.; ZOTARELLI, L. Fatores de emissão de N2O do solo em diferentes sistemas de rotação de culturas sob plantio direto e preparo convencional, na região de Londrina, PR. In: CONGRESSO BRASILEIRO DE CIÊNCIA DO SOLO, 31., 2007, Gramado. Anais... Porto Alegre: Sociedade Brasileira de Ciência do Solo, 2007. 1 CD-ROM. JANTALIA, C. P.; OLIVEIRA, O. C. de; LENGRUBER, J. de A.; URQUIAGA, S.; BODDEY, R. M.; ALVES, B. J. R. O nitrogênio como elemento chave para evitar a degradação das pastagens e suas implicações na produção de gases de efeito estufa. In: ALVES, B. J. R.; URQUIAGA, S.; AITA, C.; BODDEY, R. M.; JANTALIA, C. P.; CAMARGO, F. A. O. (Org.). Manejo de sistemas agrícolas: impacto no sequestro de C e nas emissões de gases de efeito estufa. Porto Alegre: Gênesis, 2006. p. 201-196. JANTALIA, C. P.; SANTOS, H. P. dos; URQUIAGA, S.; BODDEY, R. M.; ALVES, B. J. R. Fluxes of nitrous oxide from soil under different crop rotations and tillage systems in the South of Brazil. Nutrient Cycling in Agroecosystems, Dordrecht, v. 82, p. 161-173, 2008. KELLER, M.; VARNER, R. K.; DIAS, J. D.; SILVA, H.; CRILL, P. M.; OLIVEIRA, R. C. de; ASNER, G. P. Soil-atmosphere exchange of nitrous oxide, nitric oxide, methane, and carbon dioxide in logged and undisturbed forest in the Tapajós national forest, Brazil. Earth Interactions, Washington, DC, v. 9, p. 1-28, 2005. LARA CABEZAS, W. A. R.; KORNDORFER, G. H.; MOTTA, S. A. Volatilização de amônia na cultura do milho: II Avaliação de fontes sólidas e fluídas em sistema de plantio direto e convencional. Revista Brasileira de Ciência do Solo, Campinas, v. 21, p. 489-496, 1997. LESSA, A. C. R.; LENGRUBER, J. A.; URQUIAGA, S.; ALVES, B. J. R.; BODDEY, R. M. Importância das excretas de bovinos na produção de óxido nitroso em pastagens. In: SEMANA CIENTÍFICA JOHANNA DOBEREINER, 7., 2007, Seropédica. Anais... Seropédica: Embrapa Agrobiologia, 2007. p. 65. LINN, D. W.; DORAN, J. W. Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and non-tilled soils. Soil Science Society of America Journal, Madison, v. 48, p. 1267-1272, 1984. MADARI, B. E.; COSTA, A. R.; JANTALIA, C. P.; MACHADO, P. L. O. A.; CUNHA, M. B.; MARTINS, D. R.; SANTOS, J. H. G.; ALVES, B. J. R. Fator de emissão de óxido nitroso (N2O) para a fertilização com N na cultura do feijoeiro comum irrigado no Cerrado. Santo Antônio de Goiás: Embrapa Arroz e Feijão, 2007. 4 p. (Embrapa Arroz e Feijão. Comunicado Técnico, 144). MATSON, P. A.; MCDOWELL, W. H.; TOWNSEND, A. R.; VITOUSEK, P. M. The globalization of N deposition: ecosystem consequences in tropical environments. Biogeochemistry, Amsterdam, NL, v. 46, p. 67-83, 1999. MCKENNEY, D. J.; WANG, S. W.; DRURY, C. F.; FINDLAY, W. I. Denitrification, immobilization and mineralization in nitrate limited and non-limited residue amended soil. Soil Science Society of America Journal, Madison, v. 59, p. 118-124, 1995. METAY, A.; OLIVER, R.; SCOPEL, E.; DOUZET, J.-M.; MOREIRA, J. A. A.; MARAUX, F.; FEIGL, B. J.; FELLER, C. N2O and CH4 emissions from soils under conventional and no-till management practices in Goiânia (Cerrados, Brazil). Geoderma, Amsterdam, NL, v. 141, p. 78-88, 2007.

MOSIER, A. R.; DUXBURY, J. M.; FRENEY, J. R.; HEINEMEYER, O.; MINAMI, K. Assessing and mitigating N2O emissions from agricultural soils. Climatic Change, Amsterdam, NL, v. 40, p. 7-38, 1998. MOSIER, A.; WASSMANN, R.; VERCHOT, L.; KING, J.; PALM, C. Methane and nitrogen oxide fluxes in tropical agricultural soils: sources, sinks and mechanisms. Environment, Development and Sustainability, Amsterdam, NL, v. 6, p. 11-49, 2004. OLIVEIRA, W. R. D. Quantificação das emissões de N2O resultantes de fertilizante, fixação de N2 e resíduos de colheita em Argissolo Vermelho Amarelo. 2009. 47 f. Dissertação (Mestrado em Ciência do Solo) – Universidade Federal Rural do Rio de Janeiro, Seropédica. OLSSON, U. Confidence intervals for the mean of a log-normal distribution. Journal of Statistics Education, Washington, DC, v. 13, n. 1, 2005. Available at: . Accessed on: 18 Oct. 2009. PAUL, E. A.; CLARK, F. E. Soil microbiology and biochemistry. San Diego: Academic, 1996. 340 p. PINTO, A. S.; BUSTAMANTE, M. M. C.; KISSELLE, K.; BURKE, R.; ZEPP, R.; VIANA, L. T.; VARELLA, R. F.; MOLINA, M. Soil emissions of N2O, NO and CO2 in Brazilian Savannas: effects of vegetation type, seasonality and prescribed fires. Journal of Geophysical Research, Washington, DC, v. 107, p. 8089-8095, 2002. PRATHER, J. M. Time scales in atmospheric chemistry: coupled perturbations to N2O, NO, and O3. Science, New York, v. 279, p. 1339-1341, 1998. SÁ, T. D. D. A.; KATO, O. R.; CARVALHO, C. J. R. de; FIGUEIREDO, R. de O. Queimar ou não queimar?: de como produzir na Amazônia sem queimar. Revista USP, São Paulo, v. 72, p. 90-97, 2007. SISTI, C. P. J.; SANTOS, H. P. dos; KOHHANN, R.; ALVES, B. J. R.; URQUIAGA, S.; BODDEY, R. M. Change in carbon and nitrogen stocks in soil under 13 years of conventional or zero tillage in southern Brazil. Soil and Tillage Research, Amsterdam, NL, v. 76, p. 39-58, 2004. SKIBA, U.; SMITH, K. A. The control of nitrous oxide emissions from agricultural and natural soils. Chemosphere - Global Change Science, Oxford, v. 2, p. 379-386, 2000. SMITH, P.; MARTINO, D.; CAI, Z.; GWARY, D.; JANZEN, H.; KUMAR, P.; MCCARL, B.; OGLE, S.; O’MARA, F.; RICE, C.; SCHOLES, B.; SIROTENKO, O. Agriculture. In: METZ, B.; DAVIDSON, O. R.; BOSCH, P. R.; DAVE, R.; MEYER, L. A. (Ed.). Climate change 2007: mitigation of climate change. Cambridge: Cambridge University Press, 2007. p. 497-540. SOMMER, R.; VLEK, P. L. G.; SÁ, T. D. A.; VIELHAUER, K.; COELHO, R. F. R.; FÖLSTER, H. Nutrient balance of shifting cultivation by burning or mulching in the Eastern Amazon: evidence for subsoil nutrient accumulation. Nutrient Cycling in Agroecosystems, Dordrecht, v. 68, p. 257–271, 2004. TEIXEIRA, M. A.; MURRAY, M. L.; CARVALHO, M. G. Assessment of land use and land use change and forestry (LULUCF) as CDM projects in Brazil. Ecological Economics, New York, v. 60, p. 260-270, 2006. URQUIAGA, S.; CRUZ, K. H. S.; BODDEY, R. M. Contribution of nitrogen fixation to sugar cane: nitrogen-15 and nitrogen balance estimates. Soil Science Society of America Journal, Madison, v. 56, p. 105-114, 1992. VARGAS, M. A. T.; MENDES, I. C. de; CARVALHO, A. M. de.; BURLE, M. L.; HUNGRIA, M. Inoculação de leguminosas e manejo de adubos verdes. In: SOUSA, D. M. G. de; LOBATO, E. (Ed.). Cerrado: correção do solo e adubação. 2. ed. Planaltina: Embrapa Cerrados, 2004. p. 97-127. VERCHOT, L. V.; DAVIDSON, E. A.; CATTÂNIO, J. H.; ACKERMAN, I. L.; ERICKSON, H. E.; KELLER, M. Land use change and biogeochemical controls of nitrogen oxide emissions from soils in eastern Amazonia. Global Biogeochemical Cycles, Washington, DC, v. 13, p. 31-46, 1999. WANDER, A. E. A competitividade do agronegócio brasileiro de arroz. Custos e Agronegócio on line, Recife, v. 2, p. 1-14, 2006. Available at: . Accessed on: 25 Oct. 2007. ZHANG, W.; ZHANG, X. A forecast analysis on fertilizers consumption worldwide. Environment Monitoring and Assessment, Dordrecht, v. 133, p. 427-434, 2007.

ZHOU, X.-H.; GAO, S. Confidence intervals for the log-normal mean. Statistics in Medicine, Chichester, v. 16, p. 783-790, 1997. ZOTARELLI, L. Balanço de nitrogênio na rotação de culturas em sistema plantio direto e convencional na região de Londrina-PR. 2000. 97 f. Dissertação (Mestrado em Ciência do Solo) - Universidade Federal Rural do Rio de Janeiro, Seropédica.

Chapter 6

Methane emissions in flooded rice cultivation Magda Aparecida de Lima, Rosa Toyoko Shiraishi Frighetto, Omar Vieira Villela, Falberni de Souza Costa, Cimélio Bayer, Vera Regina Mussoi Macedo, Elio Marcolin

Abstract: Evaluations of methane emissions in flooded rice cultivation were performed in the Southeast and South regions of Brazil, using the static chamber method. Experiments were conducted with conventional tillage, no tillage and minimum tillage in the South and Southeast regions. Crops under continuous and intermittent water regimes were compared. These areas are located in various climate regions, with different soil types and cultivars. Assessments on the effects of various crop and soil management systems on seasonal methane emissions are presented, as well as the impact of water regimes on crops. In general, lower seasonal methane emission rates were associated with no-tillage and intermittent water regimes, taking into account the three harvests analyzed in these regions of the country. Seasonal emission factors are presented for each crop in the three sites studied (Pindamonhangaba, São Paulo (SP), Cachoeirinha, Rio Grande do Sul (RS), and Uruguaiana, Rio Grande do Sul (RS)). Keywords: irrigated rice, methane, Pindamonhangaba, Cachoeirinha, Uruguaiana, no-tillage, water regime.

Introduction Cultivation of flooded rice represents one of the major global anthropic sources of methane (CH4), a major greenhouse gas that strongly influences atmospheric photochemistry. It is estimated that the overall emission rate of this gas in irrigated rice fields ranges from 20 to 100 teragrams (average of 60 Tg) per year, which accounts for 16% of total emissions from all sources (HOUGHTON et al., 1995). CH4 is produced in flooded soils by obligate (strict) anaerobic bacteria. Drainage decreases the emission (efflux) of CH4 into the atmosphere, as soil aeration inhibits production thereof by methanogenic bacteria. Concomitantly, there is a decrease of CH4 in soil due to aerobic oxidation by methanotrophic bacteria. Flooding the soil in a rice field interrupts entry of atmospheric oxygen into the soil, and decomposition of organic matter becomes anaerobic. In addition to CO2, methane is the main end product of anaerobic decomposition, which can escape into the atmosphere by ebullition, by diffusion through surface soil layers and flood water, and through aerenchyma tissue (air spaces) in rice plants. A substantial portion of the methane produced in the soil of irrigated rice is oxidized within the soil and water, prior to escaping into the atmosphere. The two main methane production pathways in submerged soils are: (1) Reduction of CO2 with H2 (derived from an organic compound) CO2 + 4H2 → CH4 + 2H2O (2) Decarboxylation (transmethylation) of acetic acid CH3COOH → CH4 + CO2

Brazilian national estimate of methane emissions in flooded rice Using the revised methodology of the Intergovernmental Panel on Climate Change (IPCC et al., 1997) to estimate annual greenhouse gas emissions, in which an average rate of global methane emissions from flooded rice fields is recommended (20 g m−2 per growing season), in 1994, emissions estimated for Brazil were around 283 gigagrams (Gg)4 of methane from irrigated rice cultivation (BRASIL, 2004, 2006). In that year, emissions totaled 261.08 Gg (92.2%) for rice cultivated under a continuously flooded regime, 0.58 Gg (0.2%) under an intermittently flooded regime and 21.38 Gg (7.6%) under a floodplain regime. Conducting studies to obtain real methane emission rates (efflux) in national conditions of Brazil and to understand rice-growing agricultural processes and practices that impact and are influenced by climate change was the main focus of this study within the Agrogases network. This chapter presents results obtained in experiments to measure methane emissions in flooded rice crops conducted in the harvests of 2002–2003, 2003–2004 and 2004–2005 at the experimental station of the São Paulo State Agribusiness Technology Agency (APTA) [Agência Paulista de Tecnologia dos Agronegócios] / Paraíba Valley Regional Agribusiness Technology Development Hub [Pólo Regional de Desenvolvimento Tecnológico do Vale do Paraíba], in Pindamonhangaba, São Paulo (SP) and at the experimental station of the Rice Institute of Rio Grande do Sul (IRGA) [Instituto Riograndense do Arroz], in Cachoeirinha, Rio Grande do Sul (RS). In the 2004–2005 harvest, an experiment was conducted at the IRGA experimental station, in Uruguaiana, Rio Grande do Sul (RS). In experiments conducted in Pindamonhangaba, methane emissions were evaluated in irrigated production systems under continuous and intermittent water regimes. In the South region of Brazil, three tillage systems were evaluated: conventional tillage (Cachoeirinha and Uruguaiana, Rio Grande do Sul (RS)), no-tillage (Cachoeirinha, Rio Grande do Sul (RS)), and minimum tillage (Cachoeirinha, Rio Grande do Sul (RS)). These are regions with quite distinct climates, and different soil types and predominant cultivation systems. In the 2002–2003 harvest in Pindamonhangaba, São Paulo (SP), an average seasonal emission of 32.84 g CH4 m−2 ± 0.24 g CH4 m−2 was observed in rice cultivation under a continuous flooding regime, and an average of 28.47 g CH4 m−2 ± 10.65 g CH4 m−2 was found under intermittent flooding. Mean daily methane emissions were estimated at 313.57 mg CH4 m−2 ± 13.98 mg CH4 m−2 under continuous flooding and at 389.74 mg CH4 m−2 ± 140.26 mg CH4 m−2 under an intermittent flooding regime. The 2003–2004 harvest presented the lowest seasonal emission averages recorded throughout the four harvests studied, with emissions of 8.92 g CH4 m−2 ± 1.05 g CH4 m−2 under continuous flooding and 5.68 g CH4 m−2 ± 2.12 g CH4 m−2 under intermittent flooding. Mean daily emissions were also lower than those found in the other harvests, with emissions of 79.01 mg CH4 m−2 ± 9.71 mg CH4 m−2 under continuous flooding and 52.17 mg CH4 m−2 ± 18.70 mg CH4 m−2 under an intermittent flooding regime. The 2004–2005 harvest registered mean seasonal emissions of 18.91 g CH4 m−2 ± 2.38 g CH4 m−2 under a continuous flooding regime and 16.20 g CH4 m−2 ± 2.54 g CH4 m−2 under an intermittent flooding regime. Mean daily methane emissions were higher in crops under a

4

1 Gg = 109 grams.

continuous flooding regime, with 188.97 mg CH4 m−2 ± 23.76 mg CH4 m−2, compared to an intermittent flooding regime, where they were 169.94 mg CH4 m−2 ± 17.24 mg CH4 m−2. In Cachoeirinha, Rio Grande do Sul (RS), in the 2002–2003 harvest, mean seasonal emissions were estimated at 49 g CH4 m−2 under conventional tillage, and 33 g CH4 m−2 under no tillage. In the 2003–2004 harvest, emissions were estimated at 59 g CH4 m−2 and 55 g CH4 m−2 under conventional tillage and no-tillage, respectively. In Uruguaiana and in Cachoeirinha, Rio Grande do Sul (RS), in the 2004–2005 harvest, emissions were estimated at 4.0 g CH4 m−2 and 4.7 g CH4 m−2 under conventional tillage and minimum tillage, respectively.

Variables affecting emission Factors such as temperature, soil pH and addition of organic matter influence the ratio of methane and carbon dioxide produced. The exact reasons for the variations in this ratio are not very clear, but they may be the result of variations in the availability of oxidants (Fe, Mn, NO3, SO4) and changes in methanogenic archaea population. A pH in the near-neutral range (6.9–7.1) is an optimum condition for CH4 production (OREMLAND, 1988 cited by GON, 1996; WANG et al., 1997). While soil flooding generally causes a stabilization of soil pH between 6.5 and 7.2, soil Eh falls to levels that depend on the electron acceptors present (PONNAMPERUMA, 1972). In laboratory incubation studies, methane production was only observed with Eh below −150 mV (WANG et al., 1997). Higher Eh values, however, have been observed in studies conducted by Yagi and Minami (1990) in some rice fields without organic fertilization. According to Gon (1996), there are several explanations for this observation: 1) methane production occurs at microsites where platinum electrodes cannot penetrate: 2) the platinum electrodes’ long extremities – 3 mm to 5 mm – are larger than the microsites where methane production occurs, and therefore, they touch sites with multiple redox potentials, with the higher value then being recorded; 3) methane production starts at Eh values higher than −150 mV. Fetzer and Conrad (1993) showed that methane production inhibition with redox potential higher than −150 mV is caused by introduction of free oxygen into the system. Sass and Fisher (1994) found a negative correlation between methane emission and clay content. This inverse relationship may be caused by the ability of clay minerals to protect organic matter from degradation (JENKINSON, 1977; OADES, 1988 cited by GON, 1996). On the other hand, calcareous soils show rapid methane formation during flooding (NEUE; ROGER, 1994) possibly because the presence of calcium carbonate raises pH in microsites where methanogenesis occurs (GON, 1996). The growing rice plant, with its developing root system and biomass, constitutes the main factor controlling the seasonal methane emission pattern. In the vegetative stage (up until tillering), soil organic matter residues left by previous crops are the main source of substrate for methanogens. In the intermediate stage (tillering to flowering) and at later development stages (flowering to harvest) of rice plants, the main probable sources of substrates are, respectively, root exudates and litter (GON, 1996). Climate variables, such as solar radiation, rainfall and temperature may indirectly influence methane production by affecting net primary production and root exudation (GON, 1996).

Temperature is one of the environmental factors that govern decomposition of organic substances in flooded soil, and in particular, the methane emission rate (PARASHAR et al., 1993; SCHÜTZ et al., 1990). Conrad et al. (1987) reported that methane production was 2.5 to 3.5 times higher at a soil temperature of 30 °C, in comparison with soil at 17 °C. Tsutsuki and Ponnamperuma (1987) recorded a significant increase in methane emissions with an increase in soil temperature from 20 °C to 35 °C. Rath et al. (2002) also reported an increase in methane production with an increase in incubation temperature from 15 °C to 35 °C, in which methane production was negligible in laterite and acid sulphate, compared to production in alluvial soil, even at a high temperature. These authors suggested that sensitivity to temperature depends on soil type and availability of substrates. In their experiment, the addition of rice straw to laterite and alluvial soils significantly increased the values of the temperature sensitivity coefficient. So they concluded that, even though tropical soils contain less organic carbon, residue generated in post-harvest and application of organic sources such as green manure in rice fields could contribute to acceleration of methane production during their rapid decomposition in tropical soils. Air temperature, besides affecting methane production through methanogenic archaea, also affects transference of methane from the rhizosphere to the atmosphere through rice plants (NOUCHI et al. 1994). Studies show that methane flux generally has two peaks in seasonal variations during growing season: a first peak before tillering, and a second one during the rice plants’ reproductive stage. The first would occur due to the breakdown of organic matter in the soil, and the second would be due to plant activity, which would provide exudates or litter for soil bacteria (SCHÜTZ et al., 1989).

Locations of the study The study was conducted in three locations, one in the Southeast region of Brazil (Pindamonhangaba, São Paulo (SP)) and two in the South region (Cachoeirinha, Rio Grande do Sul (RS), and Uruguaiana, Rio Grande do Sul (RS)). The original project proposal was to conduct studies in three physiographic areas in the South region. Because this was a pioneer study, in which many operational details depended on faithful monitoring of the study’s stages, in addition to the close presence of Embrapa Environment, an initial experimental test was developed in the city of Pindamonhangaba, São Paulo (SP). Afterward, due to annual variation observed between this experiment and those developed in Cachoeirinha, it was deemed important to continue quantification of methane emissions in rice fields in the Southeast region, also aiming for a possible comparison between various climate conditions and management systems in various areas. Cachoeirinha, Rio Grande do Sul (RS), was chosen due to it being located in a physiographic area interfacing between the South Coast and Depression region, where logistical conditions (proximity of Sedex 10 express mail) and human resources were found to carry out the work, with the support of IRGA and the Federal University of Rio Grande do Sul (UFRGS) [Universidade Federal de Rio Grande do Sul]. In the first harvest studied (2001–2002), a preliminary experiment was conducted in Pindamonhangaba, São Paulo (SP), to test the closed chamber method (INTERNATIONAL ATOMIC ENERGY AGENCY, 1992). In the 2002–2003 and 2003–2004 harvests, experiments were conducted in Pindamonhangaba and Cachoeirinha, Rio Grande do Sul (RS). In the 2004–2005 harvest, experiments were conducted in those areas and also in Uruguaiana, Rio Grande do Sul (RS).

Below is a description of the characteristics of the areas studied, and crop and water management systems for the harvests of 2002–2003, 2003–2004 and 2004–2005.

Experimental area of the Paraíba Valley Regional Agribusiness Technology Development Hub [Pólo Regional de Desenvolvimento Tecnológico do Vale do Paraíba], Pindamonhangaba, São Paulo (SP) This is located at latitude 22° 55' S and longitude 45° 30' W, at an average altitude of 560 meters. The soil is classified as a Gleysol with clayey to clayey-loam texture. The land has been planted for over nine years with flooded rice crops. Two systems of irrigated rice cultivation were studied: 1) rice cultivation system under a continuous flooding regime (average 10 cm sheet); and 2) rice cultivation system under an intermittent flooding regime (alternate wetting). For the treatment under an intermittent water regime, there were periodic interruptions of the water flow. In the harvest of 2003–2004, however, there was abundant rainfall, so it did not require a regular supply of water. The variety used in the three harvests was IAC 103, whose characteristics are described in Table 1. That is the trade name of the lineage IAC 1282 and it resulted from a cross between lineages LI 84-124 and LI 82-227, conducted at the Campinas Experimental Center in 1986. The rice was manually transplanted to furrows, with spacing between plants for this variety 30 cm between rows and 20 cm between clumps. Six seedlings were used per clump. The production of rough or “paddy” rice was determined with 13% moisture. The transplanting system was used in Pindamonhangaba, São Paulo (SP) because it was the predominant management type in the region, and, as it was a representative system, it was decided to keep it.

Experimental Station of the Rice Institute of Rio Grande do Sul (IRGA) [Instituto Riograndense do Arroz], Cachoeirinha, Rio Grande do Sul (RS) The IRGA experimental area in Cachoeirinha (central depression region) is located at latitude 29° 57' 02" S, longitude 51° 06' 02" W, at an average altitude of 7 m, with flat relief and humid subtropical climate, Cfa (Köppen). The soil is classified as Dystrophic Ta Haplic Gleysol, loamy texture (SISTEMA..., 1999). The annual historical average (1975–2002) of mean air temperature was 20 °C, rainfall was 1,394 mm and solar radiation was 357 cal cm−2 day−1 (COSTA, 2005).

Experimental station of the Rice Institute of Rio Grande do Sul (IRGA) [Instituto Riograndense do Arroz], Uruguaiana, Rio Grande do Sul (RS) The experiment was conducted at the IRGA Experimental Station in Uruguaiana, Rio Grande do Sul (RS), located in the West Border region, at an altitude of 74 meters above sea level, latitude 29° 45' 33" and longitude 57° 05' 37". The climate is classified as Cfa, according to the Köppen classification. According to the Embrapa classification (SISTEMA..., 1999), the soil is characterized as vertic carbonatic Ebanic Chernosol – MEk. The cultivar used in the experiment was Irga 417, a trade name of the lineage Irga 318-11-69-2B, selected from progeny of a cross between F1 of New Rex / IR19743-25-2-2 with BR-Irga-409,

developed in 1983 in the Rice Experimental Station (EEA) [Estação Experimental do Arroz], and launched in 1995.

Results and discussion Results are presented for experiments to measure methane emissions in flooded rice crops conducted in the harvests of 2002–2003, 2003–2004 and 2004–2005 at the experimental station of the São Paulo State Agribusiness Technology Agency (APTA) [Agência Paulista de Tecnologia dos Agronegócios] / Paraíba Valley Regional Agribusiness Technology Development Hub [Pólo Regional de Desenvolvimento Tecnológico do Vale do Paraíba], in Pindamonhangaba, São Paulo (SP) and at the experimental station of the Rice Institute of Rio Grande do Sul (IRGA) [Instituto Riograndense do Arroz], in Cachoeirinha, Rio Grande do Sul (RS). The results of an experiment carried out at the IRGA experimental station in Uruguaiana, Rio Grande do Sul (RS) are also presented for the 2004–2005 harvest.

Location: Pindamonhangaba, São Paulo (SP) 2002–2003 harvest In the 2002–2003 harvest in Pindamonhangaba, São Paulo (SP), an average seasonal emission of 32.84 g CH4 m−2 ± 0.24 g CH4 m−2 was observed in rice cultivation under a continuous flooding regime, and an average of 36.00 g CH4 m−2 ± 10.65 g CH4 m−2 was found for the intermittent flooding regime (Table 1). Mean daily methane emissions were estimated at 313.57 mg CH4 m−2 ± 13.98 mg CH4 m−2 under continuous flooding and at 389.74 mg CH4 m−2 ± 140.26 mg CH4 m−2 under an intermittent flooding regime. Until the middle of the growing season, mainly during the vegetative stage, emissions were higher under the intermittent flooding regime, dropping at a later stage below emissions under the continuous regime (Figure 1). It was noted that variations in emissions were higher under the intermittent flooding regime. Water management has a strong influence on methane emission rates, according to Aulakh et al. (2001), Cai et al. (2003), Sass et al. (1992) and Smith & Conen (2004). In the 2002–2003 harvest, no difference was observed between average methane emission rates in treatments with continuous and intermittent irrigation management. This can be explained by the small time interval for aeration under the intermittent water regime, given the frequent occurrence of rainfall. The two periods of water supply interruption in the intermittent regime amounted to a total of only 22 days without water supply. An increase in emissions was also observed at the beginning of both treatments, after fertilization with urea. Studies have linked urea application with increased methane emission in flooded rice fields (AULAKH et al., 2001; LINDAU, 1994; RATH et al., 1999). According to Sass and Fisher (1995), maximum emissions are obtained by applying 200 kg ha−1 to 300 kg ha−1 of urea N, with lower emissions at ranges from 0 kg ha−1 to 100 kg ha−1 of urea N. Apparently, methane emissions declined after urea applications done on February 14, 2003 and February 28, 2003. A more frequent collection period, close to the fertilizer application date, however, could clarify the effect of fertilizer application on emissions.

CHAP 6 - FIGURE 1 Fluxo de emissão

Emission flux

Contínuo

Continuous

Intermitente

Intermittent

Data de colheita das amostras

Sample collection date

−2 Figure 1. Methane emission fluxes, in mg m , under continuous and intermittent flooding regimes, in the 2002– 2003 harvest, in Pindamonhangaba, São Paulo (SP).

Source: Adapted from Emissão... (2008).

In this experiment, the occurrence of simultaneous methane flux increases was also observed in both treatments, after fertilizer applications performed on three dates (January 30, 2003, February 14, 2003 and February 28, 2003). Sass (personal communication, 2001) points out the occurrence of an initial large peak in the booting stage and a secondary peak in the anthesis (flowering) stage. Emission peaks may also occur also during the final stage of the season, during maturation. In the harvest studied, emission increases were observed after flowering in the treatment under the continuous water regime. During maturation, after the water supply was cut (April 4, 2003), emission peaks were recorded in both treatments, which probably occurred due to rainfall in this period. During the rice’s booting stage (the reproductive phase of growth), no samples were collected that would allow obtaining evidence of the influence of this stage on methane emissions. 2003–2004 harvest Seasonal emissions measured were only 8.92 g CH4 m−2 ± 1.05 g CH4 m−2 under continuous flooding, and 5.68 g CH4 m−2 ± 2.12 g CH4 m−2 under the intermittent flooding regime, (Table 2), well below the values found in the previous harvest. Mean daily emissions were also lower than those found in the other harvests, with emissions of 79.01 mg CH4 m−2 ± 9.71 mg CH4 m−2 under the continuous flooding regime and 52.17 mg CH4 m−2 ± 18.70 mg CH4 m−2 under the intermittent flooding regime. Initially, higher emissions of CH4 were observed under intermittent irrigation, in relation to the continuous flooding regime (Figure 2). The occurrence of heavy rain for a long period kept the soil under intermittent flooding in anaerobic conditions, similar to those observed under continuous irrigation management. After the first application of fertilizer (February 3, 2004), a methane emission peak was recorded in both the continuous and intermittent regimes. This pattern also occurred after the second fertilizer application (March 1, 2004). Table 2. Mean daily CH4 emissions, seasonal flux and grain yield, in Pindamonhangaba, São Paulo (SP), 2003– 2004 harvest. Continuous Regime

Intermittent Regime

Mean daily emission (mg m−2 day−1)

79.01 ± 9.71

52.17 ±18.70

Seasonal flux (g m−2)

8.92

5.68

Yield (kg of rough or “paddy” grain ha−1)

5,600

4,000



Emission flux

Contínuo

Continuous

Intermitente

Intermittent





Starting from the time of panicle initiation, emissions remained relatively close in both treatments, up until the booting stage (March 9, 2004), during which a new emission peak was recorded in both regimes. With broadcast fertilization (March 14, 2004) emissions under the continuous flooding regime rose until March 15, 2004 (75 days after flooding), after which emissions began to decline. In the intermittent flooding regime, without water sheet, emissions continued to drop even after this third fertilizer application. At the time of flowering (March 20, 2004), after 80 days of flooding, emission levels in the continuous flooding regime remained practically stabilized until March 24, 2004, and then decreased. At that plant stage, the main substrates for methanogenic archaea are composed of root exudates. Combined with this aspect, the occurrence of heavy and frequent rainfall in the period caused emissions in the intermittent regime to remain stabilized until April 8, 2004, with a new peak being observed on April 14, 2004, followed by a gradual reduction in emissions. In the continuous regime, after flowering, an emission peak was registered on March 31, 2004, followed by reduction until the water supply was cut off (April 10, 2004). On April 19, 2004, there was one last emission peak under the continuous regime, followed by a reduction until harvest. For the intermittent regime there was also an increase in emissions on April 14, 2004, with a subsequent reduction. Those peaks are likely to be associated with use of root litter by methanogens and/or release of methane trapped in the soil (GON, 1996; JAIN et al., 2000; WASSMANN et al., 1994). 2004–2005 harvest In the 2004–2005 harvest, in Pindamonhangaba, mean seasonal emissions recorded were 18.91 g CH4 m−2 ± 2.38 g CH4 m−2 under continuous flooding and 16.20 g CH4 m−2 ± 1.69 g CH4 m−2 under intermittent flooding (Table 3), and for the period of 19 to 25 April 2005, practically equivalent. Mean daily methane emissions were higher in crops under the continuous flooding regime, with 188.97 mg CH4 m−2 ± 23.76 mg CH4 m−2, compared to the intermittent flooding regime, where they were 169.94 mg CH4 m−2 ± 17.24 mg CH4m−2. In the initial stage of the plant, methane emissions were higher under the intermittent flooding regime (Figure 3). Table 3. Mean daily CH4 emissions, seasonal flux and grain yield, in Pindamonhangaba, São Paulo (SP), 2004– 2005 harvest.

Continuous flooding regime Mean daily emission (mg m−2 day−1) −2

Seasonal flux (g m ) Yield (kg rough or “paddy” grain ha−1)

Intermittent flooding regime

188.97

169.94

18.91

16.20

5,742.00

4,876.00



Emission flux

Contínuo

Continuous

Intermitente

Intermittent





Table 4. Mean daily CH4 emissions, total (seasonal) CH4, emissions, rice grain yield, relationship between CH4 released and grain yield and the CH4 emission percentage in development stages of rice in soil under conventional tillage (CT) and no tillage (NT) in the 2002–2003 and 2003–2004 harvests, in Cachoeirinha, Rio Grande do Sul (RS). Season 2002–2003

Mean daily CH4 emission (mg m−2 day−1) −2

Seasonal CH4 emission (g m ) −1

Grain yield (Mg ha ) CH4 / grain yield (g kg−1) Plant stage

2003–2004

CT

NT

CT

NT

435 ± 79(1)

324 ± 49

500 ± 92

455 ± 73

49 ± 2.6

33 ± 0.7

59 ± 4.1

55 ± 1.2

7.30

6.40

8.40

7.70

67

52

70

71

CH4 emission in rice plant development stages(2)

Vegetative

50%

47%

Reproductive

30%

28%

Maturation

20%

25%

(1)

Standard error of the mean. (2) Mean of CT and NT. Source: Adapted from Emissão... (2008).

Location: Cachoeirinha, Rio Grande do Sul (RS) Table 4 presents summary results of CH4 emissions and total amount of CH4 released from soil under conventional tillage (CT) and no tillage (NT), in the harvests of 2002–2003 and 2003– 2004.

During cultivation, CH4 emission rates under NT were lower than under CT, except in the first two samplings (Figure 4). Mean rates and ranges of daily CH4 emissions were 435 mg m−2 day−1 ± 79 mg m−2 day−1 and 500 mg m−2 day−1 ± 92 mg m−2 day−1 under CT, and 324 mg m−2 day−1 ± 49 mg m−2 day−1 and 455 mg m−2 day−1 ± 73 mg m−2 day−1 under NT. The mean emission rates are within the range of emissions cited in the international literature, which vary from 0 mg m−2 day−1 to 1,920 mg m−2 day−1 (LE MER; ROGER, 2001).

CHAP 6 - FIGURE 4 Metano

Methane

Vegetativa

Vegetative

Reprodutiva

Reproductive

Amadurecimento

Maturation

Nn

Nn

DP

PD

EP

PE

DS

SD

Colheita

Harvest



Plantio direto

No tillage

Dias após a inundação

Days after flooding

Figure 4. CH4 emission rates during rice plant development stages in conventional tillage and no-tillage, in a 105day period of the 2002–2003 harvest, in Cachoeirinha, Rio Grande do Sul (RS). The values represent the average of two replicates. Nn represents nitrogen applications, PD is panicle differentiation, PE is panicle emission and SD]is soil drainage. The bars on the symbols correspond to the standard error of the mean. Source: Adapted from Emissão... (2008).

Higher initial CH4 emission in NT in relation to CT may be due to the maintenance of winter crop residues on the soil surface. Working with two soils and with application of ryegrass residues, both incorporated and on the soil surface, Sousa (2001) demonstrated that there was an increase in production of organic acids during the first weeks of flooding when ryegrass residues remained on the soil surface, with acetic acid showing a higher concentration in the soil solution. This acid is one of the products of anaerobic decomposition of organic residue and is readily usable in methanogenesis (WASSMANN et al.,1998). Organic plant residues in flooded soils increase CH4 emission by reducing the oxi-reduction potential (Eh) of the soil and by serving as a source of organic compounds for methanogenesis (NEUE et al. 1996). In CT, decomposition of the residue may have occurred even before final flooding of the soil, due to the increase of the total specific surface area (SSA) of fractionated residue susceptible to microbial attack. As the residues were already partially decomposed at the time of flooding, the soil would take longer to reach the reduction levels required for methanogenesis, when compared to

NT. Over the course of cultivation, this increase in SSA could result in higher decomposition rates of the more recalcitrant fractions of crop residues under CT, when compared to NT, which did not have its residue fractionated, thus being less decomposable in terms of the size of particles being attacked by microbial decomposers. The tendency for deeper roots in the CT soil profile (data not shown), in relation to NT, could explain the CT soil profile’s higher emissions of CH4. In deeper layers of the soil, O2 concentration is lower. Consequently, this environment will be more reduced, and with no deficiency of labile CO, the potential for CH4 production will be greater (WASSMANN et al., 1998). Over the course of cultivation, CT always presented higher above-ground phytomass than NT. Greater phytomass may be associated with a bigger root system, with a possible increase in the release of C from roots, intensifying the production of CH4 (WANG et al. 1997). Both systems had two emission peaks, at 35 and 66 days after application (DAA). These peaks occurred at the end of the vegetative stage, and at the end of the reproductive stage and the beginning of the maturing stage (LINDAU et al., 1991). The first peak may be due to the release of organic compounds in decomposition of winter cover crop plant residues, coupled with the presence of easily metabolizable compounds in soil organic matter (NEUE et al., 1996). The second peak may be related to the release of root exudates, in addition to decomposition of rice roots. In this study, broadcast application of N (N1, Figure 4) was done 2 days before the peak detected at 35 DAA, which may have influenced the peak’s magnitude in both systems. Cai et al. (1997) noted that CH4 emission was reduced by 7% and 14% with application of 100 kg N ha−1 and 300 kg N ha−1, respectively, in relation to control treatment. However, Lindau et al. (1991) found that fertilization with urea increased CH4 emissions by 86% with application of 300 kg N ha−1, compared to control. In this experiment, after draining the soil, CH4 emissions decreased in both management systems. This is attributed to reduction of soil moisture after drainage (BRONSON et al., 1997), making it more oxidized and reducing methanogenic activity (LE MER; ROGER, 2001). The total amount of CH4 emitted over 105 days in CT (49 g m−2) was 48% higher than in NT (33 g m−2). Total emissions in the 2003–2004 harvest in Cachoeirinha were 59 g CH4 m−2 ± 4.1 g CH4 m−2 in conventional tillage and 55 g CH4 m−2 ± 1.2 g CH4 m−2 in no-tillage. No-tillage presented a methane emission 20% lower than conventional tillage only in the vegetative stage of the rice crop, as happened with the 2002–2003 harvest, and the first, and largest, emission peak also occurred at the end of the vegetative stage, as in the previous harvest (Figure 5). CHAP 6 - FIGURE 5 Metano

Methane

Vegetativa

Vegetative

Reprodutiva

Reproductive

Amadurecimento

Maturation

Nn

Nn

DP

PD

EP

PE

DS

SD

Colheita

Harvest



Plantio direto

No tillage


Days after flooding

Figure 5. CH4 emissions (effluxes) during rice plant development stages in conventional tillage and no-tillage systems, in a 111-day period of the 2003–2004 harvest, in Cachoeirinha, Rio Grande do Sul (RS). Values represent the average of three replicates. Nn represents nitrogen applications, PD is panicle differentiation, PE is panicle emergence and SD is soil drainage. The bars on the symbols correspond to the standard error of the mean. Source: Emissão... (2008).

Emissions in the no-tillage system were initially higher than those in the conventional tillage system, probably due to crop residues remaining on the soil surface. Methane emissions in this harvest were slightly higher than those observed in the 2002–2003 harvest, and were estimated at 49 g m−2 and 33 g m−2 in conventional tillage and no-tillage, respectively. In the 2004–2005 harvest, three management systems were evaluated: (1) conventional tillage (CT), (2) minimal tillage (MT) and (3) a natural system (NS). The variations in seasonal emissions observed between MT and CT (2004–2005), reflect modifications that the soil management system can cause to methane emissions, and demonstrate the potential of systems like MT and CT to mitigate methane emission into the atmosphere. It was found, however, that other important factors are the cultivar and the climate and soil conditions. Moreover, it should also be taken into account that, in years with low rainfall, there are difficulties in maintaining the level of the sheet of water in rice fields, which can strongly influence methane fluxes, and this was likely a determining factor for the lower emissions observed in 2004 -2005. Among the cultivation systems evaluated, conventional tillage was the one that showed higher emission values; on several days it showed twice the value of minimum tillage emissions (Figure 6). The evaluation of methane flux in the natural system (NS) was carried out to observe the methane emission contribution of a system not subject to agricultural use. It was found that there was no flux in the evaluated natural system. Quantifications made using chromatography indicate that methane concentration values ranged between 1.0 ppm and 1.3 ppm, values that are very close to the concentration of methane in the atmosphere. The low rainfall values in the region during the 2004–2005 harvest must be taken into account. Low rainfall in this atypical year may also have influenced the emissions measured in the conventional tillage (CT) system. In the 2004–2005 season it was found that the highest emission values for conventional tillage were approximately 320 mg m−2 day−1, whereas in previous years, during the seasons of 2002–2003 and 2003–2004, methane emissions reached 650 mg m−2 day−1 of methane in conventional tillage. Seasonal emissions obtained in the 2004–2005 harvest, in Cachoeirinha, Rio Grande do Sul (RS), showed a significant difference in methane emission contributions between conventional

tillage and minimum tillage. Conventional tillage in this year showed an emission three times higher than minimum tillage (Figure 7).

CHAP 6 - FIGURE 6 Taxa de fluxo de CH4

CH4 flux rate

Vegetativa

Vegetative

Reprodutiva

Reproductive

Maturação

Maturation

Preparo Convencional


Cultivo Mínimo

Minimum tillage

Sistema Natural

Natural system


Days after flooding

Figure 6. Methane fluxes in conventional tillage, minimum tillage and natural system during the 2004–2005 harvest. Conventional tillage and minimum tillage fluxes are the mean values of two chambers; and for the natural system, averages are obtained from measurements in three chambers, in Cachoeirinha, Rio Grande do Sul (RS). Source: Emissão... (2008).

CHAP 6 - FIGURE 7 Fluxo de CH4

CH4 flux

Preparo Convencional

Conventional Tillage

Cultivo Mínimo

Minimum Tillage

Sistema Natural

Natural System

Figure 7. Seasonal methane flux during the 2004–2005 harvest, in Cachoeirinha, Rio Grande do Sul (RS). Source: Emissão... (2008).

The 2004–2005 crop yield was similar in both tillage systems, with values of 7,200 kg ha−1 for minimum tillage and 7,021 kg ha−1 for conventional tillage.

Location: Uruguaiana, Rio Grande do Sul (RS) Emissions were measured in the 2004–2005 harvest in an irrigated rice cultivation area grown under conventional tillage (Figure 8). The estimated emission values in the Uruguaiana experiment were lower, reaching 100 mg m−2 day−1, when compared to emissions in Cachoerinha, which reached 300 mg m−2 day−1, for the same tillage system.

CHAP 6 - FIGURE 8

Taxa de fluxo de CH4

CH4 flux rate

Drenagem do solo

Soil drainage

Dias após a inundação do solo

Days after soil flooding

Figure 8. Methane fluxes in a conventional tillage system during the 2004–2005 harvest. Flux values were obtained from the average of two chambers in Uruguaiana, Rio Grande do Sul (RS). Note: N-application of urea. Source: Emissão... (2008).

The mean methane emission rate observed in Uruguaiana was lower than that observed in Cachoeirinha for both tillage systems. The seasonal emission value for Cachoeirinha was about 3 times greater than that of Uruguaiana for the conventional tillage system (Table 5). Table 5. Mean daily CH4 emission rate and seasonal emission, under conventional tillage (CT) and minimum tillage (MT) in the 2004–2005 harvest in Cachoeirinha, Rio Grande do Sul (RS), and Uruguaiana, Rio Grande do Sul (RS). Cachoeirinha Mean flux rate, mg CH4 m−2 day−1 Seasonal CH4 flux, g m−2

Uruguaiana

CT

MT

NS

CT

135.9(1)

52.9(1)

-0.4(1)

31.7(1)

13.2

4.7

-0.44

4.0

Note: CT: Conventional tillage, MT: Minimum Tillage, NS: Natural System. (1) Average from 14 evaluations. Source: Emissão... (2008).

Methane emissions from flooded rice fields under systems that are representative of the regions studied showed high variation throughout the harvests studied. The tillage systems most commonly used in Rio Grande do Sul are conventional tillage (CT) and no tillage (NT), both under the continuous flooding regime, representing, respectively, 41% and 14% of the cultivated area (INSTITUTO RIO GRANDENSE DO ARROZ, 2003). In the State of São Paulo, the most representative system is conventional tillage, under the continuous flooding regime. In Pindamonhangaba, São Paulo (SP), seasonal methane emissions from irrigated rice under the continuous flooding regime (the representative system for the region) varied from 32.84 g CH4 m−2 in the 2002–2003 harvest, to 8.92 g in the 2003–2004 harvest and 18.91 g in the 2004–2005 harvest, with an average of 20.22 g CH4 m−2 ± 12.01 g CH4 m−2 in that period. In Cachoeirinha, Rio Grande do Sul (RS), seasonal methane emissions from irrigated rice under the continuous flooding regime varied from 49.00 g CH4 m−2 in the 2002–2003 harvest, to 59.00 g CH4 m−2 in the 2003–2004 harvest, with an average of 54.00 g CH4 m−2 ± 7.07 g CH4 m−2. In a no-tillage system, seasonal emissions in Cachoeirinha varied from 33 g CH4 m−2 in the 2002–2003 harvest, to 55.00 g CH4 m−2 in the 2003–2004 harvest, with an average of 44.00 g CH4 m−2 ± 15.56 g CH4 m−2. Results from the 2004–2005 harvest obtained in Cachoeirinha and Uruguaiana, Rio Grande do Sul (RS), were not considered when calculating mean seasonal emissions, as the treatments were not repeated in other years. With the exception of the 2002–2003 harvest, results from Pindamonhangaba showed lower methane emissions in cultivation systems under the intermittent water regime in relation to systems under the continuous water regime, with reduction of methane emissions under the intermittent flooding regime ranging from 14% to 36%. In Cachoeirinha, the results showed a tendency to lower methane emissions in no-tillage (NT) cultivation systems, when compared to conventional tillage (CT) systems. In minimum tillage

systems (MT), it was also observed that methane emissions are lower when compared to those of conventional tillage. In Uruguaiana, Rio Grande do Sul (RS), low methane emission was observed under conventional cultivation conditions. However, given the unique climate conditions that occurred in this region during that harvest, it is not possible to say that this is the normal emission pattern associated with this region. Table 6 presents the final results of the study, with seasonal methane emission factors in g m−2 obtained in the locations studied, in the 2002–2003, 2003–2004 and 2004–2005 harvests. Table 6. Methane emissions in rice production systems, in Pindamonhangaba, São Paulo (SP) and Cachoeirinha, Rio Grande do Sul (RS). 2002–2003, 2003–2004 and 2004–2005 harvests. Seasonal methane emissions (g m−2) Area studied

Harvest

Management system 2002 –

2003 –

2004 –

2003

2004

2005

Average

Southeast Pindamonhangaba, SP

Conventional tillage, continuous regime

Pindamonhangaba, SP

Conventional tillage, intermittent regime

32.84

8.92

18.91

20.22

± 0.24

± 1.05

± 2.38

± 12.01

36.00

5.68

16.20

19.29

± 10.65

± 2.12

± 2.54

± 15.39

South Cachoeirinha, RS

CT – Conventional tillage

Cachoeirinha, RS

NT – No-tillage

Cachoeirinha, RS Uruguaiana, RS

49

59

± 2.6

± 4.1

13.2(1)

54.00 ± 7.07 44.00

33

55

± 0.7

± 1.2

MT – Minimum tillage

–

–

4.7(1)

–

MT – Minimum tillage

–

–

4.0(1)

–

-

± 15.56

(1)

Values not included in the average. Source: Emissão... (2008).

Final considerations The experiments conducted in this project indicate that no-tillage (NT) systems tend to produce less methane emissions in relation to conventional (CT) systems, and that intermittent flooding systems tend to produce lower emissions compared to continuous flooding systems. However, to know the real magnitude of these agricultural systems’ GHG contributions, it would be important to also assess CO2 and nitrous oxide fluxes during the harvest and inter-harvest periods. The use of different rice varieties should be tested, since, according to the literature, the variety can influence methane production, depending on the plant’s intrinsic characteristics, such as production of exudates, height, and aerenchyma, among other parameters. Experiments in the South and Southeast regions of Brazil used different varieties, in accordance with local agricultural recommendations, but this fact did not allow for a comparative analysis of the production systems between regions. Likewise, the impact of using different types of nitrogen fertilizers was not

assessed in the experiments mentioned above; thus, there is still an opportunity to evaluate mitigation potential as a function of fertilizer use. The need to monitor seasonal methane fluxes in irrigated rice areas for a longer period is emphasized, evaluating various harvests and tillage systems, considering the different responses of cultivation systems to climate conditions. In addition to the locations studied (Pindamonhangaba, São Paulo (SP), Cachoeirinha, Rio Grande do Sul (RS), Uruguaiana, Rio Grande do Sul (RS)), other experiments have since then been installed in the municipalities of Itajaí, Santa Catarina (SC), Santa Maria, Rio Grande do Sul (RS) and Tremembé, São Paulo (SP). Different varieties of rice should be studied to evaluate the potential effect of various cycles, the quantity and quality of exudates produced and the intrinsic characteristics of each species in methane emission from soil. The data obtained in this study is being used to validate the denitrification and decomposition (DNDC) model, with subsequent use of a geographical information and remote sensing system for estimating methane emissions in irrigated rice cultivation areas, thereby contributing to better methane emission estimates in Brazil.

References AULAKH, M. S.; WASSMANN, R.; RENNENBERG, H. Methane emissions from rice fields: quantification, mechanisms, role of management, and mitigation options. Advances in Agronomy, New York, v. 70, p. 193-260, 2001. BRASIL. Ministério da Ciência e Tecnologia. Brazil’s initial communication to the United Nations Framework Convention on Climate Change. Brasília, DF, 2004. 271 p. BRASIL. Ministério da Ciência e Tecnologia. Primeiro Inventário Brasileiro de emissões antrópicas de gases de efeito estufa: relatórios de referência: emissões de metano do cultivo de arroz. Brasília, DF, 2006. 56 p. BRONSON, K. F.; SINGH, U.; NEUE, H. U.; ABAO, E. B. Automated chamber measurements of methane and nitrous oxide flux in a flooded rice soil: II. Fallow period emissions. Soil Science Society of America Journal, Madison, v. 61, p. 988-993, 1997. CAI, Z.; TSURUTA, H.; GAO, M.; XU, H.; WEI, C. Options for mitigating methane emission from a permanently flooded rice field. Global Change Biology, Urbana, v. 9, p. 37-45, 2003. CAI, Z.; XING, G.; YAN, X.; XU, H.; TSURUTA, H.; YAGI, K.; MINAMI, K. Methane and nitrous oxide emissions from rice paddy fields as affected by nitrogen fertilisers and water management. Plant and Soil, The Hague, v. 196, n. 1, p. 7-14, 1997. CONRAD, R.; SCHÜTZ, H.; BABBEL, M. Temperature limitation of hydrogen turnover and methanogenesis in anoxic paddy soil. FEMS Microbiology Letters, Amsterdam, NL, v. 45, p. 281-289, 1987. COSTA, F. de S. Estoques de caborno orgânico e efluxos de dióxido de carbono e metano de solos em preparo convencional e plantio direto no subtrópico brasileiro. 2005. 127 f. Tese (Doutorado em Ciência do Solo) – Faculdade de Agronomia, Universidade Federal do Rio Grande do Sul, Porto Alegre. EMISSÃO de metano em sistemas de produção de arroz irrigado: quantificação e análise: relatório técnico final. Brasília, DF: Ministério da Ciência e Tecnologia: Embrapa, 2008. 86 p. FETZER, S.; CONRAD, R. Effect of redox potential on methanogenesis by Methanosarcina barkeri. Archives of Microbiology, New York, v. 160, p. 108-113, 1993. GON, H. A. C. van der. Methane emission from wetland rice fields. Wageningen: Agricultural University, 1996. 182 p. HOUGHTON, J. T.; MEIRA FILHO, L. G.; BRUCE, J.; LEE, H.; CALLANDER, B. A.; HAITES, E.; HARRIS, N.; MASKELL, K. (Ed.). Climate change 1994: radiative forcing of climate change and an evaluation of the IPCC IS92 emission scenarios. Cambridge: Cambridge University Press, 1995. 339 p.

INSTITUTO RIO GRANDENSE DO ARROZ. Available at: . Accessed on: 26 Mar. 2003. INTERNATIONAL ATOMIC ENERGY AGENCY. Manual on measurement of methane and nitrous oxide emissions from agriculture. Vienna, AT, 1992. 91 p. (IAEA. TECDOC, 674). IPCC. Intergovernmental Panel on Climate Change; UNEP. United Nations Environment Programme; OECD. Organization for Economic Co-Operation and Development; IEA. International Energy Agency. Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories. Paris, FR, 1997. JAIN, M. C.; KUMAR, S.; WASSMANN, R.; MITRA, S.; SINGH, S. D.; SINGH, J. P.; SING, R.; YADAV, A. K.; GUPTA, S. Methane emissions from irrigated rice fields in northern India (New Delhi). Nutrient Cycling in Agroecosystems, Dordrecht, v. 58, p. 75-83, 2000. JENKINSON, D. S. Studies on the decomposition of plant material in soil: V. The effects of plant cover and soil type on the loss of carbon from 14C labeled ryegrass decomposing under field conditions. Journal of Soil Science, Oxford, v. 28, p. 424-434, 1977. LE MER, J.; ROGER, P. Production, oxidation, emission and consumption of methane by soils: a review. European Journal of Soil Biology, New Jersey, v. 37, p. 25-50, 2001. LINDAU, C. W. Methane emissions from Louisiana rice fields amended with nitrogen fertilizers. Soil Biology and Biochemistry, Oxford, v. 26, p. 353-359, 1994. LINDAU, C. W.; BOLLICH,P. K.; DELAUNE, R. D.; PATRICK, W. H.; LAW, V. J. Effects of urea fertilizer and environmental factors on CH4 emissions from a Louisiana, USA rice field. Plant and Soil, The Hague, v. 136, p. 195-203, 1991. NEUE, H. U.; ROGER, P. A. Potential of methane emission in major rice ecologies. In: ZEPP, R. G. (Ed.). Climate biosphere interaction. New York: Wiley & Sons, 1994. p. 65-93. NEUE, H. U.; WASSMANN, R.; LANTIN, R. S.; ALBERTO, M. C. R.; ADUNA, J. B.; JAVELLANA, A. M. Factors affecting methane emission from rice fields. Atmospheric Environment, New York, v. 30, p. 1751-1754, 1996. NOUCHI, I.; HOSONO, T.; AOKI, K.; MINAMI, K. Seasonal variation in methane flux rice paddies associated with methane concentration in soil water, rice biomass and temperature, and its modelling. Plant and Soil, The Hague, v. 161, p. 195208, 1994. PARASHAR, D. C.; GUPTA, P. K.; RAJ, J. S.; DHARMA, R. C.; SINGH, N. Effect of soil temperature on methane emission from paddy field. Chemosphere, Amsterdam, NL, v. 26, p. 247–250, 1993. PONNAMPERUMA, F. N. The chemistry of submerged soils. Advances in Agronomy, New York, v. 24, p. 29-96, 1972. RATH, A. K.; KAMAKRISHNAN, B.; SETHUNATHAN, N. Temperature dependence of methane production in tropical rice soils. Geomicrobiology Journal, Philadelphia, v. 19, p. 581-592, 2002. RATH, A. K.; SWAIN, B.; RAMAKRISHNAN, B.; PANDA, D.; ADHYA, T. K.; RAO, V. R.; SETHUNATHAN, N. Influence of fertilizer management and water regime on methane emission from rice fields. Agriculture, Ecosystems and Environment, Amsterdam, NL, v. 76, p. 99-107, 1999. SASS, R. L.; FISHER, F. M.; WANG, Y. B.; TURNER, F. T.; JUND, M. F. Methane emission from rice fields: the effect of floodwater management. Global Biogeochemical Cycles, Washington, DC, v. 6, n. 3, p. 249-262, 1992. SASS, R. L.; FISHER, J. M. CH4 emissions from paddy fields in the United States Gulf Coast area. In: MINAMI, K.; MOSIER, A.; SASS, R. L. (Ed.). CH4 and N2O: global emissions and controls from rice fields and other Agricultural and Industrial Sources. Tsukuba: National Institute of Agro-Environmental Sciences, 1994. p. 65-77. (NIAES Series, 2). SASS, R. L.; FISHER, J. M. Methane emissions from Texas rice fields: a five year study. In: PENG, S.; INGRAM, K. T.; NEUE, H. U.; ZISKA, L. H. (Ed.). Climate change and rice. Berlin, DE: Springer-Verlag, 1995. p. 46-59. SCHÜTZ, H.; HOLZAPFEL-PSHORN, A.; CONRAD, R.; RENNENBERG, H.; SEILER, W. A three-year continuous record on the influence of daytime, season, and fertilizer treatment on methane emission rates from an Italian rice paddy field. Journal of Geophysical Research, Washington, DC, v. 94, p. 16405-16416, 1989. SCHÜTZ, H.; SEILER, W.; CONRAD, R. Influence of soil temperature on methane emission from rice paddy fields. Biogeochemistry, Amsterdam, NL, v. 11, p. 77-95, 1990.

SISTEMA brasileiro de classificação de solos. Brasília, DF: Embrapa Produção de Informação; Rio de Janeiro: Embrapa Solos, 1999. 412 p. SMITH, K. A.; CONEN, F. Impacts of land management on fluxes of trace greenhouse gases. Soil Use and Management, Oxford, v. 20, p. 255-263, 2004. SOUSA, R. O. Oxirredução em solos alagados afetada por resíduos vegetais. 2001. 164 f. Tese (Doutorado em Ciência do Solo) – Universidade Federal do Rio Grande do Sul, Porto Alegre. TSUTSUKI, K.; PONNAMPERUMA, F. N. Behaviour of anaerobic decomposition products in submerged soils, effects of organic material amendment, soil properties and temperature. Soil Science and Plant Nutrition, Tokyo, JP, v. 33, p. 1333, 1987. WANG, B.; NEUE, H. U.; SAMONTE, H. P. Effect of cultivar difference (“IR71”, “IR65598” and “DULAR”) on methane emission. Agriculture, Ecosystems and Environment, Amsterdam, NL, v. 62, p. 31-40, 1997. WASSMANN, R.; NEUE, H. U.; BUENO, C.; LANTIN, R. S.; ALBERTO, M. C. R.; BUENDIA, L. V.; BRONSON, K.; PAPEN, H.; RENNENBERG, H. Methane production capacities of different rice soils derived from inherent and exogenous substrates. Plant and Soil, The Hague, v. 203, p. 227-237, 1998. WASSMANN, R.; NEUE, H. U.; LANTIN, R. S.; ADUNA, J. B.; ALBERTO, M. C. R.; ANDALES, M. J.; TAN, M. J.; GON, H. A. C. van der; HOFFMANN, H.; PAPEN, H.; RENNENBERG, H.; SEILER, W. Temporal patterns of methane emissions from wetland rice fields treated by different modes of N application. Journal of Geophysical Research, Washington, DC, v. 99, p. 1645716462, 1994. YAGI, K.; MINAMI, K. Effects of organic matter application on methane emission from some Japanese paddy fields. Soil Science and Plant Nutrition, Tokyo, JP, v. 36, p. 599-610, 1990.

Chapter 7

Implementation of a generic model for sugarcane crops in the Southeastern region of Brazil Jônatan Dupont Tatsch, Osvaldo Machado Rodrigues Cabral, Marco Moriondo, Marco Bindi, Humberto Ribeiro da Rocha, Marcos Antonio Vieira Ligo, Helber Custódio de Freitas

Abstract: Sugarcane biomass has emerged as a source of renewable energy in developing countries and as a convenient and timely way of mitigating climate changes for Brazil. Given this outlook and the current expansion of sugarcane areas, the use of mathematical models to simulate the growth of agricultural systems can provide technical support to maximize productivity, considering the effects of climate, soil conditions and management system. This study reports on the first application of the CropSyst model on sugarcane crops, where comparison with experimental data obtained in two consecutive ratoon cane cycles showed that the model produced satisfactory estimates of the crop’s above-ground biomass. The model was used to simulate accumulation of above-ground biomass and evapotranspiration of a sugarcane crop in the northern region of the State of São Paulo. The set of parameters obtained in this calibration is an initial adaptation of the model to the Brazilian production environment, where only the effects of climate and soil were considered for sugarcane growth. Keywords: sugarcane, biophysical model, CropSyst, biomass, climate, soil.

Introduction Brazil is the world’s largest sugarcane producer (FAO, 2006), with a cultivated area of 6.2 million hectares and annual production of approximately 475 million tons (CONAB, 2007). The largest production areas are concentrated in the Central-South region of the country, particularly in the State of São Paulo, which has about 57% of production (IBGE, 2007). On average, 55% of sugarcane is harvested for ethanol production, with the remainder being used to produce sugar. Between 2000 and 2005, global sugarcane production increased at a rate of 13% per year, with Brazil accounting for 19% and 37% of the world’s sugar and ethanol production, respectively (MARTINES-FILHO et al., 2006). Brazilian ethanol has attracted a considerable interest in the international market due to sugarcane’s yield and efficiency in production of biofuel and energy compared to other crops. Thus, sugarcane biomass has emerged as an example of a new source of renewable energy among developing countries and as a convenient and timely way of mitigating climate changes for Brazil (GOLDEMBERG, 2007). The potential of applying mathematical models to simulate the growth of agricultural systems has been considered an issue of current relevant interest (EWERT et al., 2007; WALLACH, 2006). Given the current expansion of sugarcane areas, the use of these models can provide technical support to maximize productivity, considering the effects of climate, soil conditions and management system. The implementation of a generic crop model, particularly for sugarcane, provides an alternative analytical tool to study the economic and environmental viability of rotating sugarcane with other crops (e.g., peanut and soybean). The CropSyst model (STÖCKLE et al. 2003), based on concepts of efficient use of solar radiation and water, has been used to simulate growth, development and yield of a wide variety of

crops (maize, soybean, wheat, rye, sorghum, barley, irrigated rice, grapes, forage species) with generally good results (CONFALONIERI; BOCCHI, 2005; DONATELLI et al., 1996, 1997; FERRERALEGRE et al., 1999a, 1999b; FRANCAVIGLIA; BARTOLINI, 1997; FRANCAVIGLIA et al., 1999; GIARDINI et al., 1998; PALA et al.,1996; PANNKUK et al., 1998; RINALDI; VENTRELLA, 1997; STÖCKLE; DEBAEKE, 1997; STÖCKLE et al., 1994, 1997) for various regions (Western USA, Southern France, Northern and Southern Italy, Northern Syria, Germany, Spain, Western Australia, Russia and Africa). In Brazil, there are still no published results describing the performance of this model for sugarcane cultivation areas. Simulations of the growth of perennial crops, however, have been conducted for alfalfa cultivation areas in Northern Italy (CONFALONIERI; BECHINI, 2004). This study assessed the application of the CropSyst model to simulate the accumulation of above-ground biomass and evapotranspiration of a sugarcane crop in the northern region of the State of São Paulo, using experimental data obtained in two consecutive ratoon cane cycles.

Material and method Experimental data Data was collected in a 351 ha commercial sugarcane (Saccharum spp.) cultivation area located at the São José do Pulador Farm (Farm 27) of the Santa Rita Sugar Mill (latitude 21° 38' S, longitude 47° 47' W, altitude 552 m) in the municipality of Luiz Antônio, São Paulo (SP), in the center of the area containing sugarcane ratoon, variety SP83-2847, planted with a 1.5 m spacing between rows. The soil was classified as a Red-Yellow Latosol (LVA) [Latosol = US: Oxisol; RedYellow Latosol = US: Rhodic/Xanthic Haplustox] with sandy texture. Climate variables for rainfall (Prec, mm day−1), maximum and minimum temperature (Tmax, Tmin, °C), maximum and minimum relative humidity (RHmax, RHmin, %), incident global solar radiation (Srad, MJ m−2 day−1) and wind speed (U, m s−1) were measured by the automatic weather station (Campbell Systems) installed in the center of the plot in a micrometeorological tower. Data was sampled every 10 seconds and recorded as a 10 minute average in a data acquisition system (CR10X, Campbell Systems), from which daily values were calculated. The variables observed and used to calibrate the model were: above-ground biomass (defined here as the sum of the weights of dried stems, green leaves and senescent leafs) and evapotranspiration measured by an eddy covariance system (CABRAL et al., 2003). The measures of each component of above-ground sugarcane biomass were made at intervals of approximately 30 days during the period from April 2005 to June 2007.

The CropSyst model CropSyst (Cropping Systems Simulation Model) (STÖCKLE et al., 1994, 2003) is a deterministic biophysical model developed as an analytical tool to study the effect of climate, soil and tillage on productivity of a single crop or a crop rotation system. To that end, it simulates soil water balance, nitrogen balance in the soil-plant system, crop phenology, growth of roots and crown, biomass production, yield, production and decomposition of residue, soil erosion by water, and salinity. These processes are controlled by climate, soil and crop characteristics, management system used (includes crop rotation, which can be applied for multiple years), cultivar selection, irrigation, nitrogen fertilization, pesticide application, agricultural operations for soil preparation,

and waste management. The code was written in the C++ programming language using the Euler method for numerical integration with a daily time step. The CropSyst model has a graphical user interface that allows integration with a Geographic Information System (GIS) and a hydrological model, thus facilitating its use in the context of a spatial problem. The most important input data for the model is: daily meteorological data, crop characteristic parameters (morphology, phenology and growth), planting date, hydraulic characteristics of the soil profile, initial conditions of the soil profile (plant residues, water content, mineral nitrogen and organic matter) and the dates and amounts of products applied in each fertilization. The main outputs of the model are: above-ground biomass, expressed as dry matter, leaf area index, root depth, potential and actual evapotranspiration, soil moisture and nitrogen balance. The soil water balance of the model includes rainfall, irrigation, runoff, interception, infiltration, redistribution in the soil profile, transpiration and evaporation. Potential evapotranspiration can be estimated by the Penman-Monteith or Priestley-Taylor methods. Soil water dynamics are solved using the simple cascade system or using Richardson’s equation (finite difference method). Nitrogen balance includes transformations of N in the soil (mineralization, nitrification, denitrification and volatilization), biological nitrogen fixation (only considered in legume species simulations) and N demand and accumulation by the crop. Water and nitrogen balances interact to simulate the transportation of N into the soil, as well as chemical balances (pesticides and salinity). Furthermore, the model can simulate soil erosion using the method of Revised Universal Soil Loss Equation (RUSLE) (RENARD et al. 1997). Results and Discussion Weather observations over the two sugarcane plantation cycles are shown in Figure 1. The difference between them is due to the temporal distribution of rainfall and solar radiation. Although the total rainfall in the 2005–2006 cycle (1,270 mm) was lower than the total of the 2006– 2007 cycle (1,433 mm), the total rainfall during the first 100 days in 2005 was 195 mm, and in 2006, it was only 15 mm. The abundant rainfall and heavy cloud conditions observed in early 2007 promoted a reduction of 18% in total accumulated solar radiation within the photosynthetically active range in the last 180 days of the 2006–2007 cycle, in relation to the previous cycle.

CHAP 7 - FIGURE 1 Ciclo 2005-2006

2005-2006 cycle

Ciclo 2006-2007

2006-2007 cycle

URmax

RHmax

URmin

RHmin

Prec

Rain

(mm dia-1)

(mm day-1)

Fev.

Beb.

Abr.

Apr.

Ago.

Aug.

Out.

Oct.

Dez.

Dec.

Data

Date

Figure 1. Seasonal variation of maximum and minimum temperature (Tmax and Tmin, respectively, both in °C), maximum and minimum relative humidity (RHmax and RHmin respectively, both in %), daily totals of global solar incidence (Srad in MJ m−2 day−1) and rainfall (Prec, in mm day−1), and daily mean wind speed (U, in m s−1) in a sugarcane area of Luis Antônio, São Paulo (SP), from February 2005 to June 2007.

The textural analysis conducted up to a 2 m depth (Figure 2) indicated an average composition of 22% clay, 3% silt and 74% sand. More details on the experimental site, climatology of the region and climate conditions during the first observation period (February 2005 – March 2006) are described in Tatsch (2006).


Depth

Areia

Sand

Silte

Silt

Argila

Clay

Figure 2. Vertical soil profile (Red-Yellow Latosol (LVA) [Latosol = US: Oxisol; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox]) of the experimental area of Luis Antônio, São Paulo (SP).

Harvests occurred on May 10, 2006 and May 28, 2007, corresponding to the second and third ratoon cane cycle, respectively. The yield (green weight of stalks) obtained by the sugar mill was 102.4 t ha−1 in the 2005–2006 cycle and 63 t ha−1 in 2006–2007; the respective yields estimated through biomass sampling were 118 t ha−1 ± 23 t ha−1 and 85 t ha−1 ± 14 t ha−1; they were higher due to the inclusion of sugarcane tops (SCTs), which are usually discarded during harvesting. There was a significant reduction (~40%) of total above-ground biomass in the 2006–2007 cycle compared to the 2005–2006 cycle (Figure 3), which led to the reformation of the area after the 2007 harvest. Two factors contributed to the productivity decline:, the mean decrease in yield, which (according to the Santa Rita Sugar Mill) is around 5 t ha−1 year−1 during the course of regrowth; and the climate factor of the temporal distribution of rainfall and solar radiation discussed earlier.

CHAP 7 - FIGURE 3 Biomassa de cana-de-açúcar

Sugarcane biomass

Fev.

Feb.

Abr.

Apr.

Maio

May

Ago.

Aug.

Set.

Sep.

Out.

Oct.

Dez.

Dec.

Biomassa aérea


Folhas secas

Dry leaves

Folhas verdes

Green leaves

Colmos

Dried stems

Figure 3. Variation in biomass (dry matter) of sugarcane in Luis Antônio, São Paulo (SP), from April 2005 to June 2007. The area shaded in gray indicates the standard deviation of the biomass samples. The scale on the x axis is in days after cutting (DAC).

Model calibration The values of the crop’s characteristic parameters are described in Table 1. Most of these values were determined based on values observed experimentally, values suggested in the CropSyst manual itself for C4 plants (STÖCKLE; NELSON, 1999), references from the literature, or local experience. For some parameters, calibration was necessary (parameters highlighted in bold in Table 1). Table 1. Set of parameters determined to simulate sugarcane in the CropSyst model. Parameter

Value

Unit

Morphology Maximum root depth

1

m

Maximum leaf area index (LAImax)

3.1

m m−2

Specific leaf area (SLA)

14

m2 kg−1

Partition coefficient between leaves and stems (LSP)

2

–

Leaf sensitivity to water stress

1

–

Fraction of LAImax at physiological maturity

0.8

–

Solar radiation extinction coefficient (k)

0.5

–

1

–

Transpiration – above-ground biomass coefficient

10 (9)(1)

kPa kg m−3

Conversion coefficient of radiation into above-ground biomass

4 (3.5)(1)

g MJ−1

5.2

mm day−1

-500

J kg−1

-1,300

J kg−1

27

°C

Ratio between potential and actual transpiration limiting leaf growth

0.96

–

Ratio between potential and actual transpiration limiting root growth

0.8

–

Base temperature

18

°C

Limit temperature

38

°C

Degree days to emergence

150

°C day

Degree days to LAImax

1,020

°C day

Degree days to physiological maturity

2,000

°C day

0

–

Crop coefficient for fully developed crown

2

Growth

Maximum evapotranspiration Leaf water potential at permanent wilting point Critical leaf water potential Optimal temperature

Phenology

Physiological sensitivity to water stress (1)

Parameters changed in the second cycle (values in brackets).

After a few tests with methods for estimating evapotranspiration, it was decided to perform the simulations using the Priestley-Taylor method, due to the fewer variables needed (only Tmax, Tmin and Srad) and because it presented similar results to those determined using the PenmanMonteith method. This option enables the use of the same calibration in future validation of other experimental areas, since humidity and wind measurements are generally less frequent. Table 2 indicates the periods used in the simulations. In the CropSyst model, other options characterizing the sugarcane crop were also selected, such as type of plant (in this case, C4) and the perennial option, which establishes the cutting of the crop according to harvest dates. Table 2. Dates for beginning and end of the sugarcane plantation cycles used in the simulations, with the corresponding day of the year in brackets.

Cycle

Beginning (day of year)

Harvest (day of year)

2nd ratoon cane

April 14, 2005 (105)

May 10, 2006 (130)

3rd ratoon cane

May 11, 2006 (131)

May 29, 2007 (149)

Sensitivity tests were performed mainly for the growth module, revealing a need to modify some parameters in the second cycle, indicated by footnotes (1) in Table 1.

Model results Figure 4 shows production of above-ground sugarcane biomass observed and simulated by the CropSyst model using the parameters presented above (Table 1). It was verified that simulated growth of sugarcane was consistent with the pattern observed (Figure 4a). Simulated biomass was within one standard deviation of empirically measured above-ground biomass for most of the simulated period. The model, however, overestimated the initial growth stage in the first cycle and underestimated it in the second cycle. It also overestimated the final above-ground biomass of the second cycle (Figure 4a). These initial differences demonstrate the need for a better adaptation of the soil water balance module regarding the issue of water stress, associated to possible recovery of plant growth after the start of the rainy season. The regression between simulated and observed values resulted in a determination coefficient of 0.98 (Figure 4b) and an angular coefficient (1.035) not statistically different from unity, indicating that the calibration performed did not introduce any tendencies.

CHAP 7 - FIGURE 4 Fev.

Feb.

Abr.

Apr.

Maio

May

Ago.

Aug.

Set.

Sep.

Out.

Oct.

Dez.

Dec.

Biomassa aérea


Biomassa simulada

Simulated biomass

Biomassa observada

Observed biomass

Figure 4. (A) Above-ground biomass observed (red dots) and simulated (blue line), in tons per hectare, for a sugarcane area in Luis Antônio, São Paulo (SP), from April 2005 to June 2007. The area shaded in gray indicates the standard deviation of the biomass samples. (B) Comparison between observed and simulated values of above-ground sugarcane biomass; the determination coefficient was estimated at approximately 0.98.

Final considerations The CropSyst model produced satisfactory estimates on the crop’s above-ground biomass over two ratoon cane cycles. The set of parameters obtained in the first calibration is an initial adaptation of the model to the Brazilian production environment, where only effects of climate and soil were considered for sugarcane growth and development. The results obtained indicated the importance of the soil’s water conditions in calibrating the CropSyst model to sugarcane, which should be evaluated by applying it to other experimental data (CABRAL et al., 2003) obtained in clay soil (Sertãozinho, São Paulo (SP), Dark Red Latosol [US: Rhodic Haplustox]), contributing to a better understanding of the differences between evapotranspiration rates, which may be partially related to differences in soil and sugarcane varieties used. In this exercise, the nitrogen balance module was not used, but its incorporation seems to be important, based on fertilizer application data, adjusting the phenology module for a better synchronization of growth curves in the beginning of cycles (fine tuning) and introducing the decrease in productivity during regrowth. The fact that CropSyst uses the same approach to simulate the growth of various crops reveals its major potential as an analytical tool to assess the effect of rotating sugarcane with other crops (e.g., peanut and soybean) on productivity and the environment.

References CABRAL, O. M. R.; ROCHA, H. R.; LIGO, M. A. V.; BRUNINI, O.; DIAS, M. A. F. S. Fluxos turbulentos de calor sensível, vapor d’água e CO2 sobre uma plantação de cana-de-açúcar (Saccharum sp.) em Sertãozinho, SP. Revista Brasileira de Meteorologia, Rio de Janeiro, v. 18, p. 61-70, 2003. CONAB. Companhia Nacional de Abastecimento. Acompanhamento da safra brasileira cana-de-açúcar: safra 2007/2008: terceiro levantamento. Available at: . Accessed on: 19 Dec. 2007. CONFALONIERI, R.; BECHINI, L. A preliminary evaluation of the simulation model CropSyst for alfalfa. European Journal of Agronomy, Amsterdam, NL, v. 21,p. 223-237, 2004. CONFALONIERI, R.; BOCCHI, S. Evaluation of CropSyst for simulating the yield of flooded rice in northern Italy. European Journal of Agronomy, Amsterdam, NL, v. 23, p. 315-326, 2005. DONATELLI, M.; SPALLACCI, P.; MARCHETTI, R.; PAPINI, R. Evaluation of CropSyst simulations of growth of maize and of water balance and soil nitrate content following organic and mineral fertilization applied to maize. In: EUROPEAN SOCIETY FOR AGRONOMY CONGRESS, 4., 1996, Wageningen. Proceedings... [S.l.]: European Society for Agronomy, 1996. p. 342-343. DONATELLI, M.; STÖCKLE, C. O.; CEOTTO, E.; RINALDI, M. Evaluation of CropSyst for cropping systems at two locations of northern and southern Italy. European Journal of Agronomy, Amsterdam, NL, v. 6, p. 35-45, 1997. EWERT, F.; PORTER, J. R.; ROUNSEVELL, M. D. A.; LONG, S. P.; AINSWORTH, E. A.; LEAKEY, A. D. B.; ORT, D. R.; NOSBERGER, J.; SCHIMEL, D. Crop models, CO2, and climate change. Science, Washington, DC, v. 315, p. 459–460, 2007. FAO. Food and Agricultural Organization. Food and agricultural commodities production. Available at: . Accessed on: 18 Mar. 2006. FERRER-ALEGRE, F.; VILLAR, J. M.; CARRASCO, I.; STÖCKLE, C. O. Developing management decision tools from yield experiments with the aid of a simulation model: an example with N fertilization in corn. In: INTERNATIONAL SYMPOSIUM

MODELLING CROPPING SYSTEMS, 1., 1999, Lleida. Proceedings… Lleida: European Society for agronomy-Division Agroclimatology and Agronomic Modelling, 1999a. p. 179-180. FERRER-ALEGRE, F.; VILLAR, J. M.; CASTELLVÌ, F.; BALLESTA, A.; STÖCKLE, C. O. Contribution of simulation techniques to the evaluation of alternative cropping systems in Andorra. In: INTERNATIONAL SYMPOSIUM MODELLING CROPPING SYSTEMS, 1., 1999, Lleida. Proceedings... Lleida: European Society for agronomy-Division Agroclimatology and Agronomic Modelling, 1999b. p. 180-18. FRANCAVIGLIA, R.; BARTOLINI, D. Calibrazione del modello CropSyst su una rotazione mais da granella-frumento tenero nella bassa Pianura Padana. Agricoltura Ricerca, Rome, IT, v. 171, p. 73-80, 1997. FRANCAVIGLIA, R.; MECELLA, G.; SCANDELLA, P.; MARCHETTI, A. Model comparison to evaluate the soil moisture content in different pedoclimatic regions. In: INTERNATIONAL SYMPOSIUM MODELLING CROPPING SYSTEMS, 1., 1999, Lleida. Proceedings... Lleida: European Society for agronomy-Division Agroclimatology and Agronomic Modelling, 1999. p. 178179. GIARDINI, L.; BERTI, A.; MORARI, F. Simulation of two cropping systems with EPIC and CropSyst models. Italian Journal of Agronomy, Pavia, v. 1, p. 29-38, 1998. GOLDEMBERG, J. Ethanol for a sustainable energy future. Science, Washington, DC, v. 315, p. 808-810, 2007. IBGE. Instituto Brasileiro de Geografia e Estatística. Relatório de produção agrícola. 2007. Available at: . Accessed on: 12 Sept. 2008. MARTINES-FILHO, J.; BURNQUIST, H. L.; VIAN, C. E. F. Bioenergy and the rise of sugarcane-based ethanol in Brazil. Choices, v. 21, n. 2, p. 91-96, 2006. Available at: . Accessed on: 12 Sept. 2008. PALA, M.; STÖCKLE, C. O.; HARRIS, H. C. Simulation of durum wheat (Triticum durum) growth under differential water and nitrogen. Agricultural Systems, Barking, v. 51, p. 147-163, 1996. PANNKUK, C. D.; STÖCKLE, C. O.; PAPENDICK, R. I. Validation of CropSyst for winter and spring wheat under different tillage and residue management practices in a wheat-fallow region. Agricultural Systems, Barking, v. 57, p. 121-134, 1998. RENARD, K. G.; FOSTER, G. R.; WEESIES, G. A.; MCCOOL, D. R.; YODER, D. C. Agriculture handbook, 703, predicting soil erosion by water: a guide to conservation planning with the revised universal soil loss equation (RUSLE). Washington, DC: United States Department of Agriculture-Agricultural Research Service, 1997. 404 p. RINALDI, M.; VENTRELLA, D. Uso dei modelli EPIC e CropSyst in sistemi colturali del Sud Italia. Agricoltura Ricerca, Rome, IT, v. 171, p. 47-58, 1997. STÖCKLE, C. O.; CABELGUENNE, M.; DEBAEKE, P. Comparison of CropSyst performance for water management in southwestern France using submodels of different levels of complexity. European Journal of Agronomy, Amsterdam, NL, v. 7, p. 89-98, 1997. STÖCKLE, C. O.; DEBAEKE, P. Modelling crop nitrogen requirements: a critical analysis. European Journal of Agronomy, Amsterdam, NL, v. 7, p. 161-169, 1997. STÖCKLE, C. O.; DONATELLI, M.; NELSON, R. CropSyst, a cropping systems simulation model. European Journal of Agronomy, Amsterdam, NL, v. 18, n. 3/4, p. 289–307, 2003. STÖCKLE, C. O.; MARTIN, S. A.; CAMPBELL, G. S. CropSyst, a Cropping systems simulation model: water/nitrogen budgets and crop yield. Agricultural Systems, Barking, v. 46, n. 3, p. 335–359, 1994. STÖCKLE, C. O.; NELSON, R. L. CropSyst user’s manual. Pullman: Biological Systems Engineering DepartmentWashington State University, 1999. 186 p. Available at: . Accessed on: 19 Sept. 2007. TATSCH, J. D. Uma análise dos fluxos de superfície e do microclima sobre cerrado, cana-de-açúcar e eucalipto, com implicações para mudanças climáticas regionais. 2006. 112 f. Dissertação (Mestrado) - Departamento de Ciências

Atmosféricas, Instituto de Astronomia, Geofísica e Ciências Atmosféricas, Universidade de São Paulo, São Paulo. Available at: . Accessed on: 12 Sept. 2008. WALLACH, D. Evaluating crop models. In: WALLACH, D.; MAKOWSKI, D.; JONES, J. W. (Ed.). Working with dynamic crop models: evaluation, analysis and applications. Amsterdam, NL: Elsevier, 2006. p. 11-54.

Chapter 8

Greenhouse gas production in agricultural systems: Groundwork for an inventory of methane emissions by ruminants Odo Primavesi, Alexandre Berndt, Magda Aparecida de Lima, Rosa Toyoko Shiraishi Frighetto, João José Assumpção de Abreu Demarchi, Márcio dos Santos Pedreira

Abstract: In order to meet the national demand for presentation of greenhouse gas emission inventories to the United Nations Framework Convention on Climate Change (UNFCCC), this project determined ruminal methane (CH4) emission factors for some categories representing Brazilian cattle herds under tropical climate. The experiments were carried out with Holstein and crossbred zebu dairy cows, and Nelore beef cattle on experimental areas located in the Southeastern region of Brazil. The method of sulfur hexafluoride (SF6), known as the SF6 tracer method, was adapted and validated to determine ruminal methane in an open environment. It was found that the variables determining emission factors were feed composition and consumption, as well as the specific characteristics of the various animal categories (beef and dairy cattle). Based on experiments assessing methane emission rates under conditions of adequate nutritional levels and balanced diets, polynomial regression equations were developed for categories of beef and dairy cattle, allowing for determination of factors of methane emissions from enteric fermentation specific to the local conditions studied. Mitigation strategies are presented and research gaps are indicated. Keywords: ruminal methane, SF6, tropical climate, dairy cattle, beef cattle, mitigation.

Introduction Human activities are influencing global climate due to increased concentrations of greenhouse gases, particularly carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O), among others. These gases retain and redirect infrared or excess heat radiation generated (> 300 W m−2) onto the Earth’s surface (PRIMAVESI et al., 2007). According to the Intergovernmental Panel on Climate Change (IPCC), 20% of the increase in global radiative forcing is attributed to the agricultural sector, which accounts for 50% of CH4 production and 70% of anthropic N2O (IPCC et al., 1997). Intensive use of soils, burning of agricultural residue, large-scale ruminant livestock production, treatment of animal waste in liquid form and cultivation of rice in flooded fields are some of the agricultural activities that contribute to anthropogenic emissions of greenhouse gases (GHG). In order to understand the role of livestock in the global effort to mitigate the greenhouse effect, and more precisely within the sphere of the flexibility mechanisms of the Kyoto Protocol, in particular the Clean Development Mechanism (CDM), consideration should be given to the ways in which livestock contributes to anthropogenic emissions of GHG and to developing technologies and systems to reduce emissions.

In Brazil, most livestock is comprised of bovines (85% of the herd, not considering the number of poultry), with a total of 171 million head, according to the IBGE 2006 Agricultural Census (2010). The Brazilian cattle herd makes up approximately 18% of the world cattle herd (which was 979.176 million head in 2010 (ANUALPEC..., 2011)), making Brazil an important contributor to methane emissions from enteric fermentation (Table 1). The pasture production system is predominant in the country, and waste production represents a small contribution to methane emissions, as the number of confined systems is still small, representing about 2.4% of the Brazilian herd (IBGE, 2006), although Assocon (2008) estimated that there were 2,300,000 cattle head in confined feedlots that same year. Of Brazil’s total cattle herd, milk cows totaled 12.6 million head (7.4%). Methane is an important greenhouse gas, with a global warming potential 25 times greater than that of carbon dioxide and a lifespan of approximately 12 years in the atmosphere in a 100 year time horizon (DONG et al., 2006). The mean global methane atmospheric concentration is 1,780 ppbv, more than twice the value of the pre-industrial period (800 ppbv) (DLUGOKENCKY, 2001 cited by MOSIER et al., 2004). CH4 is thought to have a contribution of 15% in global warming potential (COTTON; PIELKE, 1995; HOUGHTON et al., 1990). About 70% of methane comes from anthropogenic sources, and approximately 30% comes from natural sources (Table 2). Microbiological activity in anaerobic environments (wetlands, flooded rice cultivation, enteric fermentation and anaerobic processing of waste) constitute the main source of methane, in addition to biomass burning and coal and natural gas industries (LIMA; DEMARCHI, 2007). Table 1. Main cattle herds in the world. Country

1996

2007

2010

Head (thousands)

India

299,802

281,700

304,000

Brazil

153,882

168,367

177,964

China

110,318

139,721

105,060

United States

101,656

96,669

92,550

European Union

84,526

87,650

87,500

Argentina

51,696

51,062

48,656

Australia

26,780

29,202

28,280

Russia

35,800

18,370

16,919

1,052,943

982,663

979,176

Total Source: Anualpec... (2005, 2011).

On the other hand, the largest methane sink in the atmosphere is its reaction with hydroxyl radicals (OH−) in the troposphere, estimated at 420 Tg year−1 to 520 Tg year−1 (Tg = million tons) (MOSIER et al., 2004). However, the burning of forests, which produces ozone, can reduce the concentration of hydroxyl radicals in the atmosphere. The largest biological methane sink in terrestrial ecosystems is composed of microorganisms in aerobic soils (STEUDLER et al., 1989). Soil methane can be oxidized by methanotrophic and nitrifying bacteria (15 Tg year−1 to 45 Tg year−1) (MOSIER et al., 1998). Mosier et al. (1991), measuring the effect of fertilizer application in pastures in Colorado, concluded that high recycling of nitrogen suppressed methane fixation. An important historical contribution of greenhouse gas emissions is attributed to developed countries. Estimates conducted for developing countries, however, also classify them as potential emitters of greenhouse gases (CHATFIELD, 1996; CRUTZEN; ZIMMERMAN, 1991; UNFCCC, 2005).

Emission of methane from livestock is directly related to ruminal fermentation efficiency, representing loss of carbon and loss of energy, resulting in lower animal performance. Fermentation of the food ingested by animals is a process carried out by the ruminal microbial population, which converts structural carbohydrates (fiber) into volatile short chain fatty acids, mainly acetic, propionic and butyric acids. In this fermentation process, heat is dissipated through the body surface and CO2 and CH4 are produced. The emission of methane corresponds to losses of 5% to 8% of gross energy intake, according to the classic work of Clapperton and Blaxter (1965), and of 2% to 12%, according to Johnson and Johnson (1995); the IPCC estimates an average of 6% (IPCC et al., 1997). As methane emission varies according to the quantity and quality of ingested food (ENVIRONMENTAL PROTECTION AGENCY, 1990a, 1990b), different types of domestic animal farming systems result in different levels of methane emission. Thus, indications for reducing methane emissions from livestock are linked to feed management and nutritional strategies (BERNDT, 2010; CARMONA, 2005; CLARK et al., 2001; HOLTER; YOUNG, 1992; KURIHARA et al., 1999; POSSENTI et al., 2008; TAMMINGA, 1992). Global methane emissions from enteric processes are estimated at about 85 Tg year−1, corresponding to 22.7% of total methane emissions generated by anthropogenic sources. Emissions from animal waste are estimated at about 25 Tg year−1 (ENVIROMENTAL PROTECTION AGENCY, 2000), corresponding to 7% of total emissions. For Brazil, an emission of 9 Tg of enteric methane was estimated for livestock production (BRASIL, 2006), considering both ruminants and monogastrics and waste production in 1995 (IBGE, 1998a, 1998b). In 2000, emissions rose to 10.77 Tg of methane (BRASIL, 2010), mainly due to an increase in cattle population. Those estimates were based on secondary data collected in the country and on reference values proposed by the International Panel on Climate Change (IPCC et al., 1997). These emissions account for 96% of all methane generated by agricultural sources in the country, also including irrigated rice cultivation and burning of agricultural residue in fields. Beef and milk cattle alone account for 96% of methane emissions resulting from enteric fermentation (eructation) in Brazilian livestock production (Table 2). Table 2. Brazilian and global estimates of average CH4 emissions from natural and anthropic sources. Annual CH4 emission (Tg year−1) Global

Sources Houghton et al., 1995

Mosier et al, 2004 160 (30%)

Natural Lowlands, swamps

115

100–200

Termites

20

10–50

Oceans

10

5–20

31.8(1)

1–25

15

0–5

Freshwater Sweet gases (e.g., coalbed methane)

375 (70%)

Anthropic Total fossil fuel

100

26.6%

70–120

Biospheric carbon

275

73.3%

160–430(1)

Enteric fermentation

85

22.7%

65–100

Animal waste

25

6.7%

10–30

Flooded rice fields

60

16.0%

20–150

Landfills

40

10.7%

20–70

Biomass burning

40

10.7%

20–80

Sewage treatment

25

6.7%

25 535 (100%)

Total sources Annual CH4 emission (Tg year−1) Anthropic (agribusiness) Total

(1)

Brazil (Reference year: 1994)(2) 10.24

Brazil (Reference year: 2000)(2) 10.77

Enteric fermentation

9.00

87.9%

9.60

89.1%

Animal waste

0.67

6.5%

0.68

6.3

Flooded rice fields

0.44

4.3%

0.39

3.7

Burning of agricultural residues

0.13

1.3%

0.10

0.9

Average updated according to Anderson et al. (2010), not included in the total; Tg = teragram or million tons.

Sources: Houghton et al. (1995); Mosier et al. (2004); (1) Anderson et al. (2010), (2) Brasil (2010).

Methane emission factors vary according to the animal production system and the animal’s characteristics. In the case of dairy cattle, for example, mean values of ruminal methane emissions (eructation) are 100 kg methane animal−1 year−1 in Eastern Europe countries (550 kg LW – live weight, lactation of 4,200 kg year−1 and dry matter intake of 13.8 kg day−1, or 2.5% of LW, based on forages) and 118 kg methane animal−1 year−1 in North America (600 kg LW, lactation of 6,700 kg year−1 and dry matter intake of 16.2 kg day−1 or 2.7% of LW, based on concentrated feed). It is estimated, from IPCC default values, that in African and Asian countries, emissions vary from 36 kg methane animal−1 year−1 to 56 kg methane animal−1 year−1 (HOUGHTON et al., 1995); and that in Brazil, emissions are around 54 kg methane animal−1 year−1 using pasture (CRUTZEN et al., 1986). In 2006, the IPCC presented new mean CH4 emission factors for dairy cattle (63 kg animal−1 year−1) and meat cattle (56 kg animal−1 year−1) for Latin America. In view of the need to obtain more accurate data on quantities of ruminal methane originating from cattle production systems in Brazil, several studies were conducted in São Carlos, Jaboticabal, Nova Odessa and Andradina, in the State of São Paulo (SP), aiming to obtain emission rates or emission factors (in kg animal−1 year−1) of ruminal methane from various categories of bovine cattle and with various food quantity and quality conditions. The methodology used was measuring methane emitted by animals, raised under pasture and confinement regimes with feed intake control, developed by Johnson and Johnson (1995), which uses sulfur hexafluoride (SF6) as a tracer gas, adapted in Brazil by Primavesi et al. (2004b).

Results: emission factors The emission factors referenced by the International Panel on Climate Change (IPCC) (DONG et al., 2006; IPCC et al., 1997) were estimated primarily based on work carried out in non-tropical regions. In temperate climate regions, most feed is composed of more digestible forages, and includes greater use of concentrates in the diet. It is inferred that energy intake losses in the form of methane would be greater in tropical countries, per kg of milk or meat, due to the lower crude protein content, higher fiber and lignin content and lower digestibility of tropical forages. Kurihara et al. (1999) studied the effect of food quality on methane emission rates by zebu cattle, evaluating temperate and tropical climate forages, using respirometry chambers. This study showed that higher methane emission was associated with greater dry matter intake by the animal and lower energy density of the diet. The same authors indicated different methane emission potentials associated with certain diets, with emissions decreasing in the following order: diet with tropical grasses (C4 metabolism, with more fiber and lignin), temperate climate grasses (C3 metabolism, with less fiber and lignin) and diets with some content of concentrates. These findings helped to improve the determination of emission factors for cattle production systems in Brazilian tropical conditions, as a continuation of experiments started with support from the National Climate Change Program [Programa Nacional de Mudança do Clima] under the “Avança Brasil” Program of Brazil’s Ministry of Science and Technology (MCT) [Ministério da Ciência e Tecnologia], as well as support from the Study and Project Funding Agency (FINEP) [Financiadora de Estudos e Projetos], through the “Redugas” Project, and the US Environmental Protection Agency (EPA). Early works adapted and validated the SF6 tracer gas method to measure ruminal methane losses, determining emission factors for various categories of Brazilian cattle in the Southeastern region of the country, which allowed for a preliminary estimate of methane emission rates from the dairy and beef cattle herd. Some strategies to mitigate ruminal methane emissions were also assessed. The following are two sets of results obtained within the Agrogases network, which meet the goals proposed by the project of obtaining methane emission factors for: a) baseline: beef and dairy cattle, and b) planning mitigation strategies.

Evaluating methane emissions to obtain baseline Beef cattle Demarchi et al. (2003a, 2003b) reported the first results on methane emission potential for beef cattle (Nelore breed) in Brazilian field conditions. This category (beef) contributes 81% of the total methane attributed to bovine enteric fermentation. It is worth mentioning that a large part of the national cattle population is composed of zebu breeds, in management conditions involving pastures composed mainly of forage plants of the Brachiaria genus. These authors, measuring ruminal methane emissions in Nelore beef steers, with an average live weight of 375 kg (217 kg LW to 604 kg LW) in pastures of Brachiaria brizantha cv marandu, found a mean emission value of 47.3 kg animal−1 year−1 (Table 3). Furthermore, they indicated a seasonal effect in methane emissions, reflecting qualitative pasture conditions during dry and wet seasons. In these studies, the methane conversion rate, or loss of gross energy intake, was estimated at an average of 6.8% (5.0% to 9.1%), close to the global average (default value), which is

6.5% (DONG et al., 2006). In view of the difficulties in properly estimating dry matter intake, studies were also conducted under controlled conditions, with confined animals receiving only forage, allowing for measuring food intake. Under these conditions, Nascimento (2007) found a gross energy loss rate of 6.2% to 9.0% in the form of methane. Table 3. Ruminal methane emission by Nelore beef steers, grazing on Brachiaria brizantha during the four seasons of the year (90 days/season), by live weight. Treatment (Season)

CH4 emission

DMI LW (kg) kg day−1

% of LW

g day−1

kg year−1

g day−1 kg−1 of LW

% of GEI

g kg−1 of DMI

Winter

318

6.5

2.0

102

33

0.34

5.0

16

Spring

333

6.4

1.9

132

34

0.41

6.3

21

Summer

411

7.3

1.8

220

59

0.54

9.1

30

Autumn

438

7.6

1.7

174

63

0.41

6.6

23

Average

375

7.0

1.9

157

47

0.43

6.8

23

Notes: CP (crude protein) % − NDF (neutral detergent fiber) % − IVDDM (in vitro digestibility of dry matter) %, respectively in: (1) winter (August 2002) = 3.3 − 82.1 − 41.4; (2) spring (December 2002) = 7.8 − 71.5 − 60.4; (3) summer (February 2003) = 5.4 − 81.6 − 62.5; (4) autumn (May 2003) = 5.6 − 82.5 − 56.0. LW = live weight, DMI = dry matter intake, % GEI = Ym (methane conversion rate) = percentage of gross energy intake lost, considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1. Data obtained in Nova Odessa, São Paulo (SP), latitude 22° 45' S, longitude 47° 16' W, altitude 603 m, tropical climate. Source: Adapted from Demarchi et al. (2003a, 2003b).

Energy loss in the form of methane depends on the quality and quantity of the forage available throughout the year. Differences in forage quality between seasons (rainy and dry) are significant, affecting ruminal methane production. Nascimento (2007), evaluating various ages of tropical forage, having C4 metabolism, and looking to simulate these differences in the quality of dry matter ingested throughout the year, observed a greater loss of gross energy at the most advanced development stage (Table 4). Table 4. Methane emission by castrated male Nelore cattle, in confinement, with a diet of Brachiaria brizantha at three development stages. DMI

Treatment (days)

LW (kg)

15

CH4 emission

kg day−1

% of LW

g day−1

kg year−1


% of GEI

g kg−1 of DMI

402

6.5

1.6

133

49

0.33

6.2

17

45

402

5.4

1.4

134

49

0.33

7.4

20

90

402

4.7

1.2

138

50

0.34

9.0

23

Average

402

5.5

1.4

135

49

0.33

7.5

20

Notes:

At 15, 45 and 90 days, respectively: CP (crude protein) was 10.7%, 4.5% and 4.3%; NDF (neutral detergent fiber) was 70.6%, 76.0% and 77.7%; IVDDM (in vitro digestibility of dry matter) was 64.2%, 63.0% and 63.1%. LW = live weight, DMI = dry matter intake, % GEI = Ym (methane conversion rate) = percentage of gross energy intake lost, considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1. Data obtained in Andradina, São Paulo (SP), latitude 20° 54' S, longitude 51° 22' W and altitude 400 m, tropical climate, conducted between September and December 2005. Source: Adapted from Nascimento (2007) and Nascimento et al. (2007).

With increased forage maturity there is progressive deterioration in quality, reduced intake and increased methane emission per unit of dry matter ingested. Dairy cattle Pedreira (2004) and Primavesi et al. (2004a, 2004c) reported the first results on the methane emission potential for dairy cattle in Brazilian field conditions. In extensive management of animals, with non-fertilized pastures, the data suggests that the quality of the forage ingested by the animals was similar to that of fertilized pastures, probably because of the forage selection possibility (Table 5), as suggested by the differences in in vitro digestibility and CP content, or, more likely, by an underestimation of dry matter intake in the equations used. Those equations should be adjusted to Brazilian characteristics of breed and forage and feed quality, as was done for beef cattle. Holstein and crossbred (Holstein and zebu) animals showed little difference in methane emissions between summer and autumn. In summer, lactating and dry Holstein cows emitted, respectively, 147 kg methane year−1 and 101 kg methane year−1, whereas lactating and dry crossbred cows emitted, respectively, 121 kg methane year−1 and 107 kg methane year−1. In autumn, lactating and dry Holstein cows emitted, respectively. 139 kg methane year−1 and 94 kg methane year−1, whereas lactating and dry crossbred cows emitted, respectively, 108 kg methane year−1 and 86 kg methane year−1 (PRIMAVESI et al., 2004a, 2004c). Table 5. Methane emission averages from dairy cattle (summer and autumn), in pastures. DMI Treatment

CH4 emission

LW (kg) kg day−1

% of LW

g day−1

kg year−1


% of GEI

g kg−1 of DMI

Holstein Lactating cows

571 ab

18.4 a

3.2 a

393 a

143 a

0.69 a

6.5 ab

21 ab

Dry cows

623 a

15.2 b

2.4 b

268 bc

98 bc

0.43 c

5.3 bc

18 bc

Heifer, intensive

511 bc

13.0 c

2.6 bc

233 cd

85 cd

0.46 c

5.4 bc

18 bc

Heifer, extensive

446 cd

11.9 cd

2.7 bc

178 d

65 d

0.40 c

4.5 c

15 c

Average

538

14.6

2.7

268

98

0.50

5.4

18

Crossbred Lactating cows

476 c

12.7 cd

2.7 bc

314 b

115 b

0.66 ab

7.5 a

25 a

Dry cows

501 bc

13.1 c

2.6 bc

265 bc

97 bc

0.54 bc

6.2 ab

21 ab

Heifer, intensive

382 d

10.4 d

2.7 b

209 cd

76 cd

0.55 bc

6.0 abc

20 abc

Heifer, extensive

381 d

10.5 d

2.8 b

181 d

66 d

0.47 c

5.2 bc

17 bc

Average

435

11.7

2.7

191

89

0.56

6.2

21

Notes: LW = live weight, DMI = dry matter intake, % GEI = Ym (methane conversion rate) = percentage of gross energy intake lost, considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1. Data obtained in São Carlos, São Paulo (SP), latitude 21° 57' S, longitude 47° 51' W, altitude 878 m, tropical climate, conducted in February and June 2002. LW = live weight; DMI = dry matter intake; GEI = calculated gross energy intake; intensive = fertilized pasture + concentrate; extensive = nonfertilized pasture. Difference between seasons only in DMI, %LW and CH4 in g day−1 kg−1 of LW. Means followed by the same letters do not differ among them (P > 0.05, Tukey). In summer, in vitro digestibility of organic matter and crude protein, respectively, of Panicum maximum cv. [guinea grass; tobiatã] fertilized with N = 54.5% and 15.4% (Holstein), Brachiaria decumbens fertilized with N = 48% and 7.2% (crossbred) and without N = 41% and 6.5% (heifers in extensive system); concentrate = 82%. Lactating Holstein and crossbred cows received, respectively, 40% and 32% of dry matter (DM) in the form of concentrate. Dry cows and heifers in intensive system received 20% of DM in the form of concentrate. In autumn, in vitro digestibility of organic matter (IVDOM) and crude protein (CP), respectively, of Panicum maximum cv. [guinea grass; tobiatã] fertilized with N = 55.8% and 12.5% (Holstein), Brachiaria decumbens fertilized with N = 52.8% and 6.2% (Crossbred) and without N = 49.9% and 6.3% (heifers in extensive system); concentrate = 71%. Maize silage fed to lactating Holstein cows and sorghum silage fed to lactating crossbred cows showed an IVDOM of 57.5% and 52.8%, and CP of 7.5% and 13%, respectively. Lactating Holstein and crossbred cows received, respectively, 40% and 30% of dry matter in the form of concentrate; for dry cows and heifers in intensive system this amount was 20%. Source: Adapted from Pedreira (2004), Pedreira et al. (2009) and Primavesi et al. (2004a, 2004c).

Considering the difficulties of accurately estimating the dry matter intake amount, a decision was made to conduct studies in confined feedlot with intake control. In light of the need for greater accuracy in obtaining consumption data to ultimately obtain emission factors, studies were conducted with additional feed (sugarcane) for the dry season.

Evaluation of methane emissions for mitigation purposes Beef cattle In a study using sorghum silage, corrected with urea or 60% of DM in concentrate, it was found that inclusion of concentrate in the diet, regardless of the sorghum hybrid used, promoted an increase in energy use efficiency, reflected by lower methane loss in relation to gross energy intake (Table 6). The lower methane production per unit of dry matter ingested, associated with the negative correlation between the ruminal digestibility coefficient and methane emission, showed that the food use by animals should be maximized by providing diets with higher nutritional quality (OLIVEIRA, 2005; OLIVEIRA et al., 2007; PRIMAVESI et al., 2004c). Testing increasing percentages of concentrate in the dry matter of diets based on sorghum silage, it was found that silage without concentrate resulted in lower methane emission in relation to the animal’s live weight, and that addition of 30% of concentrate to the diet led to a maximum increase of emissions, suggesting that other variables must influence the process of methane emission (BERCHIELLI et al., 2003; PEDREIRA, 2004; PRIMAVESI et al., 2004c), particularly consumption and animal performance. However, a reduction was noted in methane production per unit of dry matter, organic matter and digestible energy ingested. According to Pedreira (PEDREIRA, 2004; PEDREIRA et al., 2009), maximum methane production (150.7 g day−1) occurred with a ratio of 36.6% of concentrate in the diet. With 60% of concentrate, methane production tended to drop, due to lower fiber content in the diet caused by the addition of concentrate, besides possible changes in composition of the rumen microbial population. Regarding loss of

gross energy intake, there was a continuous reduction, as a function of the increase in energy density, lower fiber content and increased digestibility of the concentrate dry matter (Table 7). Table 6. Methane emission by Nelore steers fed with sorghum silage, supplemented with urea or with replacement of dry matter for 60% of grain concentrate, in confinement. CH4 emission

DMI Treatment

LW (kg) kg day−1

% of LW

g day−1

kg year−1


% of GEI

g kg−1 of DMI

Silage + 1.2% of urea

216 a

3.6 b

1.7 b

49 a

18 a

0.22 b

4.0 a

13 a

Silage + 60% of Concentrate

214 a

5.8 a

2.7 a

69 a

25 a

0.32 a

3.5 a

12 a

Average

215

4.7

2.2

59

22

0.27

3.8

13

Notes: Silage with urea or concentrate, respectively, with: CP 11.6% and 17.8% of DM, NDF 55.0% and 35.5%, IVDDM 62% and 75%. Silage = Sorghum silage; Concentrate = energy concentrate, with 14% of crude protein; LW = live weight; DMI = dry matter intake; GEI = Ym (methane conversion rate) = percentage of gross energy intake lost, gross energy intake, considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1; IVDOM of sorghum silage = 53.7%. Animals weighing 140 kg to 310 kg. Means followed by the same letters do not differ among them (P > 0.05, Tukey). Data obtained in Jaboticabal, São Paulo (SP), latitude 21° 15' S, longitude 48° 17' W, altitude 578 m, tropical climate. Source: Adapted from Oliveira (2005) and Oliveira et al. (2007).

From the joint analysis of these two experiments, using the same genetic groups of cattle and sorghum silage with concentrate, it was observed that a difference in response occurs between animal categories, youths and adults, and in function of live weight, with larger animals, ingesting more dry matter, releasing more methane (g day−1 and g / kg DMI) per animal; although from a certain weight, they emit less methane per kg of live weight, since there is a reduction in relative dry matter intake per live weight (PRIMAVESI et al., 2004c) (Figure 1). Table 7. Methane emission by zebu crossbred steers fed with sorghum silage with increasing substitution of dry matter for energy concentrate, in confinement. CH4 emission

DMI Treatment

LW (kg) kg day−1

% of LW

g day−1

kg year−1


% of GEI

g kg−1 of DMI

0

467 a

5.6 c

1.2 c

125 b

46 c

0.27 b

7.3 a

22 a

30

459 a

8.0 b

1.7 b

150 a

55 a

0.33 a

6.2 b

19 b

60

456 a

8.8 a

1.9 a

140 ab

51 b

0.31 a

5.4 c

16 c

Average

461

7.4

1.6

138

51

0.30

6.3

19

Notes: LW = live weight, DMI = dry matter intake, %GEI = considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1. Data obtained in Jaboticabal, São Paulo (SP), latitude 21° 15' S, longitude 48° 17' W, altitude 578 m, high-altitude tropical climate, conducted from October to November 2002. Grain concentrate replaced part of the sorghum silage dry matter. Crude protein in the diet with 0%, 30% and 60% of concentrate was 5.4%, 7.5% and 9.6%, respectively, and neutral detergent fiber (NDF) was 70%, 56% and 42%, respectively. LW = live weight, DMI = dry matter intake, GEI =

Ym (methane conversion rate) = percentage of gross energy intake lost, gross energy intake calculated; IVDOM of sorghum silage with 0%, 30% and 60% concentrate = 54%, 58% and 64%, respectively. Animals weighing 400 kg to 540 kg. Means followed by the same letters do not differ among them (P > 0.05, Tukey). Source: Adapted from Berchielli et al. (2003), Pedreira (2004) and Primavesi et al. (2004c).

CHAP 8 - FIGURE 1 Ingestão silagem sorgo + 60% concentrado, Intake of sorghum silage + 60% 144 kg PV a 525 kg PV concentrate, 144 kg LW to 525 kg LW g dia-1

g day-1

PV

LW

MSI

DMI

Figure 1. Ruminal methane emission, as a function of the animal’s live weight. DMI = dry matter intake, LW = live weight. Source: Primavesi et al. (2004c).

In turn, the improvement in the quality of the diet can occur without the use of grains, using early maturation stages of forage grasses (C4) or of C3 metabolism forages with less fiber and larger digestible fraction, such as legumes, promoting a better fermentation pattern and the reduction of methane emissions (POSSENTI, 2006; POSSENTI et al., 2008) (Table 8). Table 8. Methane emission by castrated crossbred bull, fed with coast-cross grass hay, leucaena hay (Leu), with and without yeast (Yea), in confinement. Treatment Leu (%D M)

DMI

CH4 emission

LW (kg) Yea

kg day−1

% of LW

g day−1

kg year−1


% of GEI

g kg−1 of DMI

20

with out

800

7.3

0.9

139

51

0.17

5.8

19

50

with out

800

7.3

0.9

131

48

0.16

5.5

19

20

with

800

7.4

0.9

156

57

0.20

6.4

20

50

with

800

7.6

0.9

127

46

0.16

5.1

17

800

7.4

0.9

138

51s

0.17

5.7

19

Avg Notes:

CP 17.0% of DM, NDF 70%, IVDDM 63%. LW = live weight, DMI = dry matter intake, % GEI = Ym (methane conversion rate) = percentage of gross energy intake lost, gross energy intake, calculated considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1. Data obtained in Nova Odessa, São Paulo (SP), latitude 22° 46' S, longitude 47° 16' W, altitude 561 m, tropical climate, from May to June, 2005. Source: Adapted from Possenti (2006).

In this study, the higher level of leucaena in the presence of yeast (Saccharomyces cerevisiae) promoted a better fermentation pattern, with increased production of propionic acid and reduction in methane emission. Dairy cattle

In a study conducted with crossbred animals (Holstein and Gir), of various categories, kept in confinement, fed with chopped cane corrected with urea or concentrate, Pedreira (PEDREIRA, 2004; PEDREIRA et al., 2009) and Primavesi et al. (2004c) observed no effect on methane due to different varieties of sugarcane production, but they found that the supply of concentrated feed promoted an increase in the intake of nutrients by animals and, consequently, higher methane production per day (Table 9). There was, however, a reduction in gross energy loss in the form of methane proportional to the increase in the intake of higher energy density and protein, less fibrous, dry matter, in the form of concentrate. Table 9. Methane emission by zebu crossbred dairy heifers fed with detrashed sugarcane, with various qualities and treatments in São Carlos, São Paulo (SP). CH4 emission

DMI Cane

Treat.

LW (kg) kg day−1

% of LW

g day−1

kg year−1


% of GEI

g kg−1 of DMI

Variety IAC 86-2480, with an NDF/Pol ratio of 2.3 1

+ urea

357a

6.9b

1.9b

113b

41bc

0.32b

5.4ab

17ab

1

+ conc.

372a

10.9a

2.9a

166a

61a

0.45a

4.9b

15bc

365

8.9

2.4

140

51

0.39

5.2

16

Avg.

Variety IAC 87-3184, with an NDF/Pol ratio of 3.0 2

pure*

370a

5.3c

1.4c

101c

37c

0.27b

6.4a

19a

2

+ urea

370a

7.3b

2.0b

122b

45bc

0.33b

5.3ab

17ab

2

+ conc.

399a

11.2a

2.8a

140b

51b

0.36b

4.4b

13c

385

9.3

2.4

131

48

0.35

4.9

15

Avg.

Notes: LW = live weight, DMI = dry matter intake, % GEI = Ym (methane conversion rate) = percentage of gross energy intake lost, calculated considering 4.38 Mcal of gross energy per kg of DM and 0.01334 Mcal / g CH4. Emission Factor = CH4 in kg animal−1 year−1. Data obtained in São Carlos, São Paulo (SP), latitude 21° 57' S, longitude 47° 51' W, altitude 840 m, high-altitude tropical climate, conducted in august 2002. Treat. = Treatment: pure* = pure chopped cane, tested to measure energy loss in restricting condition of dry matter intake, due to low protein content; + urea = with 1% urea; + conc. = 40% of dry matter (DM) in the form of concentrated grains, with 20% crude protein (CP); CP at end of diet = 10.5%. LW = live weight; DMI = dry matter intake; GEI = gross energy intake calculated; Pol = sucrose content measured with polarimeter. In vitro digestibility of sugarcane organic matter: 1 = 64%, 2 = 55%. Neutral detergent fiber of sugarcane: 1 = 41%, 2 = 50% Lignin in DM: 1 = 3.4%, 2 = 5.2%. Animals weighing 300 kg to 450 kg. Means followed by the same letters do not differ among them (P > 0.05, Tukey). Source: Adapted from Pedreira (2004) and Primavesi et al. (2004c).

When results are calculated as methane production per unit of digestive organic matter ingested, the intake of sugarcane with lower fiber content (cane 1) results in lower methane emission in relation to cane 2, suggesting an influence due to fiber content. The use of concentrate as part of the dry matter intake reduces methane losses per unit of nutrient. Figure 2 shows methane production as a function of digestive organic matter intake. Methane production per unit of digestible neutral detergent fiber was affected, both by cane variety (fiber content) and by the concentration of the concentrate. In Figure 3, 79% of methane emission (g day−1) is explained by an increase in the intake of digestible neutral detergent fiber.

It was found that gross energy intake losses in the form of methane, with supply of tropical forage-based feed, including chopped sugarcane, were around 6.5% of total gross energy intake, a value considered average for tropical pasture conditions (DONG et al., 2006). Only the emission rates of lactating cows, evaluated in this study, exceeded the IPCC default values for dairy cattle under temperate climate (DONG et al., 2006), which can be explained by the occurrence of a stimulus for maximum intake of fibrous dry matter (forage or silage), by adding concentrate to the diet, generating a methane emission peak. On the other hand, higher emissions by these animals are offset by higher productivity, reducing emission rates per unit of product (Table 10).

CHAP 8 - FIGURE 2 Emissão de metano (g kg-1 MODi)

Methane emission (g kg-1 DOMi)

MOD ingerida (kg dia-1)

DOM ingested (kg day-1)

−1 Figure 2. Digestible organic matter intake (DOMi) influencing ruminal methane emission (g kg DOMi) by zebu heifers.

Source: Pedreira (2004).

CHAP 8 - FIGURE 3 Emissão de metano (g dia-1)

Methane emission (g day-1)

FDN digestível ingerido (kg dia-1)

NDF ingested (kg day-1)

Figure 3. Effect of digestible neutral detergent fiber (NDF) intake on ruminal methane emissions (g day−1) by crossbred zebu heifers. Source: Pedreira (2004).

Table 10. Methane emission per unit of milk produced by cows in pasture, in São Carlos, São Paulo (SP). Lactating Cow

Concentrate grains / DMI (%)

Cow live weight (kg)

Milk production (L day−1)

CH4 / L of milk (g day−1)

Holstein

22.7

18.4

51

40

572

Crossbred

13.3

25.3

37

32

435

Average

18.0

21.9

44

36

503

IVDDM of forage (%) summer

autumn Holstein

21.8

23.1

54

45

570

Crossbred

8.8

35.8

50

28

475

Average

15.3

29.5

52

37

523

IVDDM= In vitro digestibility of dry matter. DMI = dry matter intake. Source: Adapted from Pedreira (2004), Pedreira et al. (2009) and Primavesi et al. (2004b).

Estimate of the methane emission factor for bovine cattle (inventory) Results of evaluations of methane emissions for dairy and beef cattle showed that energy losses due to feeding management in various cattle production systems have periods of gross

energy intake losses in the form of methane similar to or lower than the IPCC default value (DONG et al., 2006), which is 6.5%. Aiming to improve methane emission estimates for local conditions in the country, methane emission factors were developed for various categories of dairy and beef cattle. Table 11 presents methane emission factors by animal category and per kg of live weight, considering the results of the quantification of methane emissions by dairy cattle (São Carlos, São Paulo (SP)) and beef cattle (Nova Odessa, Jaboticabal and Andradina, São Paulo (SP)). Despite regional differences in production systems, it was found that intrinsic characteristics of food and animal categories are determining factors for CH4 emission rates. Based on this premise and on results of methane emission evaluations (g CH4 / kg LW) determined by animal category, two polynomial regression models were adjusted (1 = beef cattle, and 2 = non-lactating dairy cattle). These models were used to estimate emission factors (kg CH4 animal−1 year−1) for categories not evaluated, with the aid of a spreadsheet and with data on estimated evolution of monthly weight gain and live weight of animals. So, for beef cattle, the main model was split into two submodels: (1a) one for an exclusive pasture diet condition, and (1b) another for a diet containing concentrate, both developed based on data obtained in this study. The regression model for dairy cattle overestimates methane emissions to a certain degree, because in its data, there are no situations of food restriction, as occurs for beef cattle. Both models exclude beef and lactating dairy cows. Furthermore, it should be considered that, among European and Zebu breeds, feed conversion efficiency is similar, with a variation only in the daily dry matter intake capacity (SOEST, 1994). It was considered that a calf, at birth, with an average weight of 35 kg, does not emit methane, as it does not ingest significant amounts of forage at this stage. A. Beef cattle y (g CH4 / day kg LW) = 0.000000004x3 − 0.000007x2 + 0.0034x − 0.1211 (R2 = 0.94) where x = LW (live weight), in kg, with a data interval from 35 kg to 800 kg. (1a) Pasture beef cattle y (g CH4 / day kg LW) = 0.00000000192x3 − 0.00000424x2 + 0.002577x − 0.09 (1b) Beef cattle on diet containing concentrate y (g CH4 / day kg LW) = 0.000000002x3 − 0.000004x2 + 0.0023x − 0.0792 B. Dairy cattle, predominantly crossbreed y (g CH4 / day kg LW) = 0.000002x2 + 0.0021x − 0.0753 (R2 = 0.81) where x = LW (Live weight), in kg, with a data interval from 35 kg to 623 kg.

Table 11. Tier 2 preliminary estimates of ruminal methane emissions by the Brazilian cattle herd (2001), using emission factors obtained in the field (bold), with adjustments for various animal categories. Animal category

kg LW


Number of animals

kg CH4 year−1

Animal average

1,000,000 t year−1 CH4

Dairy cattle herd Bulls

700

540,000

0.41

106

0.06

Lactating cows

476

10,400,000

0.66

115

1.20

Dry cows

501

5,190,000

0.54

97

0.50

Heifers

2–3 years

381

3,100,000

0.47

66

0.20

Heifers

1–2 years

250

4,800,000

0.32

30

0.14

Female calves

160

5,436,000

0.21

12

0.07

Male calves

180

5,420,152

0.24

16

0.09

Average = 65

2.26

Subtotal

34,886,152 Beef cattle herd

Bulls

800

1,631,719

0.18

52

0.08

Cows

500

43,870,221

0.33

60

2.63

Heifers

2–3 years

350

9,198,246

0.38

49

0.45

Heifers

1–2 years

250

14,369,398

0.35

32

0.46

Female calves

160

16,576,294

0.26

15

0.25

Male calves

180

16,497,199

0.29

19

0.31

Steers

1–2 years

215

15,710,335

0.27

22

0.35

Steers

2–3 years

389

8,232,282

0.38

48

0.40

Steers

3–4 years

458

2,764,984

0.32

53

0.15

Steers

> 4 years

458

659,838

0.32

53

0.03

Average = 39

5.11

Subtotal

129,510,516

Grand total

164,396,668

7.37

Note: emission factors (kg CH4 animal−1 year−1) were generated by multiplying the mean values of g CH4 day−1 kg−1 of LW by 365 days, without necessarily taking seasonal variations into account, when these were not measured, and considering the dry and lactating period. Data in bold was determined under the conditions of the herd in the Southeastern region of Brazil. For data on European purebred dairy cattle, see Table 9. The default mean values (DONG et al., 2006) are 63 kg CH4 year−1 dairy cattle−1 (400 kg LW cows with production exceeding 800 kg milk−1 year−1) and 56 kg CH4 year−1 beef cattle−1 (excluding calves).

Variables that affect methane emission • Food consumption (dry matter, digestible dry matter, digestible organic matter): variable influenced by live weight (kg), average daily live weight gain (kg day−1), feeding system (confinement, intensive and extensive pasture), daily milk production (kg day−1) and fat content (%).

• Feed quality (NDF, CP and IVDDM concentrations, low 45%–55%, average 55%–75% and high 75%–85%; antinutritional factors): Diet composition may limit dry matter intake or reduce digestibility at high intake levels. • Type of diet (energy density = metabolizable energy, particle size, unsaturated oil content, concentrate percentage and starch content, etc.): Its influence on qualitative and quantitative variations of microorganisms in the digestive tract. • Genetic potential for production (zebu or taurine species; adult weight for early- to latematuring animals; feed efficiency, residual feed intake). • Animal category (young and adult animals, in growth or termination stage, milk production and dry period; reproduction). • Animal species: bovine, caprine, ovine, bubaline. • Energy losses in the form of methane: percentage of gross energy intake that is lost as eructed methane. • Pasture management (dry matter availability, management method, selectivity, fertilization (type, quantity and season), use of supplements, grazing and fallow periods, forage plant, etc.). • Production system (extensive or intensive), stocking rate, feeding management (periods of dietary restriction or supplementation, forage availability, etc.), reproductive management and health management.

Possible GHG emission mitigation/reduction actions In general, seeking greater production efficiency will be reflected in lower GHG emissions per unit of product, where mitigation will not necessarily be proportional to that reduction. For example, reduction of slaughter age from 48 to 24 months will probably not reduce emissions in the same proportion. This happens because there is a greater need for resources (food, energy, fertilizer, labor, etc.) to fatten the animals faster. It is estimated that emissions per unit of product (kg CH4 per kg of meat or milk) are lower in more efficient systems. Indications to reduce methane emissions per kg of bovine meat or milk are associated with improved diets (feeds with higher content of digestible gross energy, higher protein content and less fiber), improved pastures (higher forage supply, more digestible, without restrictions throughout the year), food supplementation (with additional energy and/or protein supplements), genetic improvement of animals (with higher production capacity and better feed efficiency), better management practices (less exposure of animals to heat, with nighttime grazing and daytime ruminating under shade, proximity to water sources, quality of water available) and other measures that result in better production efficiency, which as a whole, reduce the environmental impact of production systems, making them more profitable. The supply of feed with low digestibility and crude protein content below 7% in forage may result in increased rumen retention time, increased cellulose digestion, and increased methane emission, compared to feed with good crude protein content and a higher ruminal passage rate. The use of forages with low nutritional value may lead to higher food intake by heifers, which seek to meet their energy requirements for a more intensive metabolism, resulting in a higher methane emission rate by animals with lower live weight.

A possible alternative to reduce energy loss in the form of ruminal methane is the use of quality feeds, such as forages with higher digestibility using initial stages of grass maturation (C4) or plants with C3 metabolism, using sugarcane (verifying the NDF/Pol ratio or performing a simple detrashing) corrected with concentrate and/or urea and using feeds with less fiber and more digestible fraction, such as legumes. Using animals with greater potential for meat and milk production also reduces energy loss, as it is known that animals with lower productive capacity, selected for lower quality pastures, or sometimes with restricted supply, grow and develop more slowly. Those animals ingest less food, and, although they present gross energy intake conversion rates similar to those of productive animals (CRUZ et al., 2003; TULLIO, 2004), they require more time in confinement to attain the same degree of carcass finish, associated with the thickness of the outer fat layer. Supplying better quality feeds for those less efficient animals would increase dry matter intake, but it would direct excess energy to deposition of fat and not muscle. Some possible mitigation actions are listed below: • Improve supply and quality of feeds (less NDF, more CP, more IVDDM) throughout the year, through correct pasture management, use of better quality grain and forage (grasses, legumes) or alternative feeds (chopped sugarcane supplemented with minerals to correct its low protein content, silage, tall grasses, intercropped pastures, etc.). • Improve pasture management by applying concepts of light interception in order to determine the correct moment to begin grazing in rotational systems. • Use animals with higher genetic production potential (traditional herd versus improved herd). • Incorporate a feed efficiency factor (Residual Feed Intake) in genetic enhancement programs. • Increase the efficiency of the production system, increasing animal production, reducing slaughter age and methane production per unit of product (meat or milk). • Use nutritional additives such as essential oils, secondary plant compounds, probiotics, ionophores, yeast and others, in order to optimize ruminal fermentation. • Increase the herd’s reproductive efficiency, reducing age of first calving, increasing breed cow longevity, reducing the replacement rate and the need for many animals of various categories on the site. • Ensure the welfare of animals (animals produce more in the absence of heat stress, with shade, with no wind, etc.), as well as respect social relations (definition of hierarchy, avoid constant mixing and exchanges of animals from various lots and categories). • Increase the rate of nutrient bypass to the lower digestive tract (fat and protected and/or low rumen degradable protein). Studies evaluating management strategies of production systems aiming to mitigate methane emissions in the agricultural sector should also consider the following: • Implementing already known good management practices to increase the efficiency of production systems, aiming for a lower environmental impact and lower ecological footprint of the final product.

• Reducing and correcting production systems that lead to degradation of natural resources, which, besides increasing greenhouse gas emissions, significantly contribute to the generation of heat or intensifying of infrared radiation5.

Final considerations The results presented in this chapter may be used in future greenhouse gas emission inventories in Brazil, as they concern herds that are typically predominant in the country’s livestock production. It is essential, however, for this set of data, generated with the use of a high accuracy technique (SF6 tracer) to be accompanied by appropriate statistical data so that good methane emissions estimates are produced in inventories. Currently, it is still possible to find major distortions on the country’s cattle herd population data between official and unofficial sources. Thus, emphasis must be given to the need for better quantification and characterization of national herds. Apart from this, there are many perspectives for conducting new studies and experiments aiming to improve emission factor estimates, including the following: • Improve estimates on dry matter (DM) consumption at the field level (improve the use of indicators such as alkanes, chromium, lignin, titanium, LIPE® (lignin from Eucalyptus grandis isolated, purified and enriched), iNDF (indigestible neutral detergent fiber), iADF (indigestible acid detergent fiber), weight gain, etc.). • Test combinations of feeds (sugarcane, corn (maize) silage, concentrate levels; more digestible forages; by-products). • Test diets with unsaturated vegetable oils and by-products of biodiesel industries (glycerol and oil cakes) for changes in enteric methane emissions. • Field-test various categories of animals to establish equations for methane emission factors and allow estimating emissions by the national herd, under similar management conditions (supply and quality of feed throughout the year). • Measure methane production in calves and lactating beef cows. • Evaluate ruminal parameters along with methane production (volatile fatty acids (VFAs), NH3, microorganism populations). • Evaluate the effect of diets and additives on methane production: concentrate levels, use of ionophores or probiotics, effect of substances like tannin, mimosine (leucenol) and others. • Evaluate methane emissions from cattle excreta (potential use of biodigesters). • Improve available estimates on characteristics of the national herd (total number and stratification) by region/state. • Assess impact of various production systems on production of greenhouse gases by introducing different types of pasture management, supplement use, feeding strategies and integration of crops, livestock and forestry. 5

Heat generation and increased infrared radiation are consequences of reduced soil and air humidity, caused by lack of soil shading and elimination of vaporizing structures.

• Generate mean emission factors for the Brazilian herd to include in the IPCC database. • In an integrated manner, evaluate sources and sinks of greenhouse gas (N2O, CH4, CO2) emission at the level of the overall production system, including field production of sugarcane, maize for silage, extensive and intensive pasture management; in wet and dry soil, well and poorly ventilated, with or without crop residue cover, with low and high intake levels (N, lime, organic matter). A systematic survey may reveal consequences of any mitigation measure in a given animal production system. For example, increasing the use of grains in the diet and thus reducing methane emissions might transfer the problem to the maize production field, with increased nitrous oxide production. • Emission factors should be described for all animal categories by region. This requires determining the mean live weight by age range published by the Brazilian Institute of Geography and Statistics (IBGE) [Instituto Brasileiro de Geografia e Estatística]. Another interesting aspect is to determine the methane emission per unit of final product generated (in kg CH4 / kg meat or milk), rather than per animal per year, which, in principle, considers low efficiency production systems as being more beneficial, since animals with restricted feeding generate less methane. Indicating emission per kg of product incorporates production efficiency parameters to sustainability indexes.

References ANDERSON, B.; BARTLETT, K.; FROLKING, S.; HAYHOE, K.; JENKINS, J.; SALAS, W. Methane and nitrous oxide emissions from natural sources. Washington, DC: United States Environmental Protection Agency - Office of Atmospheric Programs, 2010. Available at: . Accessed on: 27 Aug. 2010. ANUALPEC 2005: Anuário da pecuária brasileira. São Paulo: FNP, 2005. 340 p. ANUALPEC 2011: Anuário da pecuária brasileira. São Paulo: FNP, 2011. 378 p. ASSOCON. O confinador: desafio de confinamento. AG: a Revista do Criador, Porto Alegre, ano 12, n. 120, p. 67-69, 2008. BERCHIELLI, T. T.; PEDREIRA, M. S.; OLIVEIRA, S. G.; PRIMAVESI, O.; LIMA, M. A.; FRIGHETTO, R. Determinação da produção de metano e pH ruminal em bovinos de corte alimentados com diferentes relações volumoso:concentrado. In: REUNIÃO ANUAL DA SOCIEDADE BRASILEIRA DE ZOOTECNIA, 40., 2003, Santa Maria. Anais... Santa Maria: Universidade Federal de Santa Maria: SBZ, 2003. 1 CD-ROM. BERNDT, A. Estratégias nutricionais para redução de metano. In: CONGRESSO LATINOAMERICANO DE NUTRIÇÃO ANIMAL, 4., 2010, São Pedro. Anais... São Pedro: CBNA: CLANA, 2010. p. 295-306. BLAXTER, K. L.; CLAPPERTON, J. L. Prediction of the amount of methane produced by ruminants. British Journal of Nutrition, Wallingford, v. 19, p. 511–522, 1965. BRASIL. Ministério da Ciência, Tecnologia e Inovação. Inventário brasileiro de emissões antrópicas por fontes e remoções por sumidouros de gases de efeito estufa não controlados pelo Protocolo de Montreal: parte 2. Brasília, DF, 2010. 154 p. Available at: . Accessed on: 15 Dec. 2011. BRASIL. Ministério da Ciência, Tecnologia e Inovação. Primeiro inventário brasileiro de emissões antrópicas de gases de efeito estufa: relatórios de referência. Brasília, DF, 2006. Available at: . Accessed on: 15 Dec. 2011. CARMONA, J. C. El gás metano en la producción ganadera y alternativas para medir sus emisiones y aminorar su impacto a nivel ambiental y productivo. Revista Colombiana de Ciencias Pecuarias, Medellín, v. 18, p. 49-63, 2005. CHATFIELD, R. The photochemistry of ozone. 1996. 26 p. Available at: . Accessed on: 30 Jan. 2007.

CLARK, H.; KLEIN, C.; NEWTON, P. Potential management practices and technologies to reduce nitrous oxide, methane and carbon dioxide emissions from New Zealand Agriculture. Ngaherehere: Ministry of Agriculture and Forestry, 2001. Available at: . Accessed on: 12 Jull. 2005. COTTON, W. R.; PIELKE, R. A. Human impacts on weather and climate. Cambridge: Cambridge University Press, 1995. 288 p. CRUTZEN, P. J.; ASELMANN, I.; SEILER, W. Methane production by domestic animals, wild ruminants and other herbivorous fauna and humans. Tellus, Boston, v. 38B, p. 271-274, 1986. CRUTZEN, P. J.; ZIMMERMAN, P. H. The changing photochemistry of the troposphere. Tellus, Boston, v. 43A-B, p. 136-151, 1991. CRUZ, G. M.; TULLIO, R. R.; ALLEONI, G. F.; BERNDT, A.; ALENCAR, M. M.; LANNA, D. P. D. Ganho de peso e conversão alimentar de bovino jovem não-castrado de quatro grupos genéticos em confinamento em relação ao status nutricional na fase de pastejo. In: REUNIÃO ANUAL DA SOCIEDADE BRASILEIRA DE ZOOTECNIA, 40., 2003, Santa Maria. Anais... Santa Maria: Universidade Federal de Santa Maria: SBZ, 2003. 1 CD-ROM. DEMARCHI, J. J. A. A.; MANELLA, M. Q.; LOURENÇO, A. J.; ALLEONI, G. F.; FRIGHETTO, R. S.; PRIMAVESI, O.; LIMA, M. A. Daily methane emission at different seasons of the year by Nelore cattle in Brazil grazing Brachiaria brizantha cv. Marandu: preliminary results. In: WORLD CONFERENCE ON ANIMAL PRODUCTION, 9., 2003, Porto Alegre. Proceedings... Porto Alegre: Sociedade Brasileira de Zootecnia: World Association of Animal Production, 2003a. 1 CD-ROM. DEMARCHI, J. J. A. A.; MANELLA, M. Q.; LOURENÇO, A. J.; ALLEONI, G. F.; FRIGHETTO, R. T. S.; PRIMAVESI, O.; LIMA, M. A. Preliminary results on methane emissions by Nelore cattle in Brazil grazing Brachiaria brizantha cv. Marandu. In: INTERNATIONAL METHANE AND NITROUS OXIDE MITIGATION CONFERENCE, 3., 2003, Beijing. Proceedings... [Beijing: China Coal Information Institute], 2003b. p. 80-84. DONG, H.; MANGINO, J.; MCALLISTER, T. A.; HATFIELD, J. L.; JOHNSON, D. E.; LASSEY, K. R.; LIMA, M. A.; ROMANOVSKAYA, A. Emissions from livestock and manure management. In: EGGLESTON, H. S.; BUENDIA, L.; MIWA, K.; NGARA, T.; TANABE, K. (Ed.). 2006 IPCC guidelines for national greenhouse gas inventories programme. Hayama: Institute for Global Environmental Strategies, 2006. v. 4, p. 10.1-10.87. ENVIROMENTAL PROTECTION AGENCY. Evaluating ruminant livestock efficiency projects and programs. In: PEER review draft. Washington, DC: EPA, 2000. 48 p. ENVIROMENTAL PROTECTION AGENCY. Greenhouse gas emissions from agricultural systems. In: WORKSHOP ON GREENHOUSE GAS EMISSIONS FROM AGRICULTURE, 1989, Washington, DC. Proceedings... Washington, DC: EPA, 1990a. v. 1. ENVIROMENTAL PROTECTION AGENCY. Methane emissions and opportunities for control: workshop results of Intergovernmental Panel on Climate Change. Washington, DC: EPA, 1990b. HOLTER, J. B.; YOUNG, A. J. Nutrition, feeding and calves: methane prediction in dry and lactating Holstein cows. Journal of Dairy Science, Savoy, v. 75, p. 2165-2175, 1992. HOUGHTON, J. T.; JENKINS, G. J.; EPHRAUMS, J. J. (Ed.). Climate change: the IPCC scientific assessment. Cambridge: Cambridge University Press, 1990. 365 p. HOUGHTON, J. T.; MEIRA FILHO, L. G.; BRUCE, J.; LEE, H.; CALLANDER, B. A.; HAITES, E.; HARRIS, N.; MASKELL, K. (Ed.). Climate change 1994: radiative forcing of climate change and an evaluation of the IPCC 1992 IS92 emission scenarios. Cambridge: Cambridge University Press, 1995. 339 p. IBGE. Instituto Brasileiro de Geografia e Estatística. Anuário estatístico do Brasil: 1997. Rio de Janeiro, 1998a. v. 57. IBGE. Instituto Brasileiro de Geografia e Estatística. Censo agropecuário 1995-1996, Brasil. Rio de Janeiro, 1998b. IBGE. Instituto Brasileiro de Geografia e Estatística. Censo agropecuário 2006: Brasil, grandes regiões e unidades da federação. Rio de Janeiro, 2006. 777 p. Available at: . Accessed on: 7 Oct. 2009.

IBGE. Instituto Brasileiro de Geografia e Estatística. Sistema IBGE de Recuperação Automática: SIDRA: banco de dados agregados. Rio de Janeiro, 2010. Available at: . Accessed on: 27 Aug. 2010. IPCC. Intergovernmental Panel on Climate Change; UNEP. United Nations Environment Programme; OECD. Organization for Economic Co-Operation and Development; IEA. International Energy Agency. Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories. Paris, FR, 1997. v. 3. JOHNSON, K. A.; JOHNSON, D. E. Methane emissions from cattle. Journal of Animal Science, Champaign, v. 73, p. 24832492, 1995. KURIHARA, M.; MAGNER, T.; HUNTER, R. A.; MCCRABB, G. J. Methane production and energy partition of cattle in the tropics. British Journal of Nutrition, Wallingford, v. 81, n. 3, p. 227-234, 1999. LIMA, M. A.; DEMARCHI, J. J. A. A. Emissão de metano pela pecuária ruminante: quantificação e estratégias de mitigação. Feed & Food, Sorocaba, v. 2, n. 7, p. 66-68, 2007. MOSIER, A. R.; DUXBURY, J. M.; FRENEY, J. R.; HEINEMEYER, O.; MINAMI, K. Assessing and mitigating N2O emissions from agricultural soils. Climatic Change, Amsterdam, NL, v. 40, p. 7-38, 1998. MOSIER, A. R.; SCHIMEL, D. S.; VALENTINE, D.; BRONSON, K.; PARTON, W. Methane and nitrous oxide fluxes in native, fertilized and cultivated grasslands. Nature, London, GB, v. 350, p. 330-332, 1991. MOSIER, A.; WASSMANN, R.; VERCHOT, L.; KING, J.; PALM, C. Methane and nitrogen oxide fluxes in tropical agricultural soils: sources, sinks and mechanisms. In: WASSMANN, R.; VLEK, P. L. G. (Ed.). Methane tropical agriculture in transition: opportunities for mitigating greenhouse gas emissions? Heidelberg: Kluwer Academic Publishers, 2004. p. 11-49. NASCIMENTO, C. F. M. Emissão de metano por bovinos Nelore ingerindo Brachiaria brizantha em diferentes estádios de maturação. 2007. 65 f. Dissertação (Mestrado em Nutrição e Produção Animal) – Universidade de São Paulo, Pirassununga. NASCIMENTO, C. F. M.; DEMARCHI, J. J. A. A.; BERNDT, A.; RODRIGUES, P. H. M. Methane emissions by Nelore beef cattle consuming Brachiaria brizantha with different station of maturation. In: GREENHOUSE GASES AND ANIMAL AGRICULTURE CONFERENCE, 3., 2007, Christchurch. Proceedings... Christchurch: GGAA, 2007. p. 64-65. OLIVEIRA, S. G. Caracterização nutricional de silagens de sorgo com variação no teor de tanino em bovinos de corte. 2005. 129 f. Tese (Doutorado em Zootecnia) – Universidade Estadual Paulista “Júlio de Mesquita Filho”, Jaboticabal. OLIVEIRA, S. G.; BERCHIELLI, T. T.; PEDREIRA, M. S.; PRIMAVESI, O.; FRIGHETTO, R.; LIMA, M. A. Effect of tannin levels in sorghum silage and concentrate supplementation on apparent digestibility and methane emission in beef cattle. Animal Feed Science and Technology, Amsterdam, NL, v. 135, n. 3-4, p. 236-248, 2007. PEDREIRA, M. S. Estimativa da produção de metano de origem ruminal por bovinos tendo como base a utilização de alimentos volumosos: utilização da metodologia do gás traçador hexafluoreto de enxofre (SF6). 2004. 162 f. Tese (Doutorado em Zootecnia) - Universidade Estadual Paulista “Júlio de Mesquita Filho”, Jaboticabal. PEDREIRA, M. S.; PRIMAVESI, O.; LIMA, M. A.; FRIGHETTO, R.; OLIVEIRA, S. G.; BERCHIELLI, T. T. Ruminal methane emission by dairy cattle in Southeast Brazil. Scientia Agricola, Piracicaba, v. 66, n. 6, p. 742-750, 2009. Available at: . Accessed on: 10 Dec. 2009. POSSENTI, R. A. Efeitos de dietas com Leucaena leucocephala com ou sem adição de Saccharomyces cerevisiae na digestão, fermentação, protozoários e produção de metano no rúmen em bovinos. 2006. 100 f. Tese (Doutorado em Zootecnia) - Universidade de São Paulo, Pirassununga. POSSENTI, R. A.; FRANZOLIN, R.; SCHAMMASS, E. A.; DEMARCHI, J. J. A. A.; FRIGHETTO, R. T. S.; LIMA, M. A. Efeitos de dietas contendo Leucaena leucocephala e Saccharomyces cerevisiae sobre a fermentação ruminal e a emissão do gás metano em bovinos. Revista Brasileira de Zootecnia, Viçosa, v. 37, n. 8, p. 1509-1516, 2008. PRIMAVESI, O.; ARZABE, C.; PEDREIRA, M. S. Mudanças climáticas: visão tropical integrada das causas, dos impactos e de possíveis soluções para ambientes rurais ou urbanos. São Carlos: Embrapa Pecuária Sudeste, 2007. 200 p. (Embrapa Pecuária Sudeste. Documentos, 70). Available at: . Accessed on: 19 Aug. 2008.

PRIMAVESI, O.; FRIGHETTO, R. T. S.; PEDREIRA, M. S.; LIMA, M. A.; BERCHIELLI, T. T.; BARBOSA, P. F. Metano entérico de bovinos leiteiros em condições tropicais brasileiras. Pesquisa Agropecuária Brasileira, Brasília, DF, v. 39, n. 3, p. 277-283, 2004a. PRIMAVESI, O.; FRIGHETTO, R. T. S.; PEDREIRA, M. S.; LIMA, M. A.; BERCHIELLI, T. T.; DEMARCHI, J. J. A. A.; MANELLA, M. Q.; BARBOSA, P. F.; JOHNSON, K. A.; WESTBERG, H. H. Técnica do gás traçador SF6 para medição de campo do metano ruminal em bovinos: adaptações para o Brasil. São Carlos: Embrapa Pecuária Sudeste, 2004b. 1 CD-ROM. (Embrapa Pecuária Sudeste. Documentos, 39). PRIMAVESI, O.; PEDREIRA, M. S.; FRIGHETTO, R. T. S.; LIMA, M. A.; BERCHIELLI, T. T.; OLIVEIRA, S. G.; RODRIGUES, A. A.; BARBOSA, P. F. Manejo alimentar de bovinos leiteiros e sua relação com produção de metano ruminal. São Carlos: Embrapa Pecuária Sudeste, 2004c. 21 p. (Embrapa Pecuária Sudeste. Circular Técnica, 39). SOEST, P. J. van. Nutritional ecology of the ruminant. 2nd ed. Ithaca: Cornell University Press, 1994. 476 p. STEUDLER, P. A.; BOWDEN, R. D.; MELILLO, J. M.; ABER, J. D. Influence of nitrogen fertilization on methane uptake in temperate forest soils. Nature, London, GB, v. 341, p. 314-316, 1989. TAMMINGA, S. Nutritional management of dairy cows as a contribution to pollution control. Journal of Dairy Science, Savoy, v. 75, n. 1, p. 345-357, 1992. TULLIO, R. R. Estratégias de manejo para a produção intensiva de bovinos visando à qualidade da carne. 2004. 107 f. Tese (Doutorado em Zootecnia) - Universidade Estadual Paulista “Júlio de Mesquita Filho”, Jaboticabal. UNFCCC. United Nations Framework Convention on Climate Change. Key GHG Data. Bonn, 2005.

Chapter 9

Computer simulations for the study of carbon and nitrogen dynamics and greenhouse gas emissions in agricultural production systems Luiz Fernando Carvalho Leite, Maria Conceição Peres Young Pessoa, Magda Aparecida de Lima, Beata E. Madari

Abstract: Analytical methods and computer simulators are widely used to quantify the effects of agricultural management systems on soil stocks of carbon (C) and nitrogen (N) and on greenhouse gas (GHG) emission estimates. Century, RothC and EPIC simulators have been widely used in temperate climate regions and, to a lesser extent, in tropical regions, to analyze C and N stock trends in several agroecosystems, given the various scenarios they promote. Due to the complexity of data input, however, other simpler simulators are being tested, such as CQESTR. For example, in Brazil, simulations using Century and CQESTR are continuously growing, probably due to the high accuracy resulting from the validations performed. Applying these computing resources to scenarios for areas under various tillage systems, in the medium and long term, a C stock increase was observed in areas under no-tillage, compared to conventional tillage. There have been reports of successfully obtaining estimates of GHG emissions by agricultural management systems using DayCent, the daily time step version of Century and, especially, DNDC (the DeNitrification-DeComposition model). DNDC has shown reasonable applicability for estimating methane from flooded rice cultivation for sites studied in Brazil. In all these applications, simulators made it possible to identify management strategies having less impact among the scenarios to be investigated after being validated. Keywords: simulator, greenhouse gases, carbon stock, Brazil, DNDC, Century, CQESTR, DayCent.

Introduction The study of agroecosystems having greater potential for greenhouse gas emissions may be accompanied by modeling activities based on system simulation. Identifying the main goal of scientific research is the key factor substantially guiding how these tools are used, in both the development phase, to support the decision-making process, and the phase of choosing and correctly using the simulators already available. Thus, the research goal, using mathematical modeling techniques and computer simulation, defines the level of complexity required for the expected response, and the information requirements, in terms of data input, are thus directly proportional to that complexity. There are several options for representing a given study subject, which is why there are generally various models and simulators proposed for representing it, each depending on the body of knowledge of the person who created it. Thus, a model is one of several possible descriptions of the real problem compatible with the goal of the study. The amount of knowledge regarding the process being modeled or being evaluated by simulation directly influences the quality of the results provided by the tool. Thus, if basic information needed to correctly represent the actual system is lacking, this limits the model’s development and further use and cause users to discredit it. Further details on mathematical modeling and computer simulation, including applications in agricultural research, can be found in Pessoa et al. (1997) and Pessoa & Scramin (2005).

Computational tools already available for use can be differentiated in terms of work scale and complexity. In this context, work scale is understood as whether the simulation is for a single site, a location, a region, or a broader scope (such as global). The work scale defines both the type of simulator to be used (depending on the detail needed) and the need to integrate these simulators with other computing resources, such as databases, specialized systems and Geographic Information System (GIS). Simulators of C and N dynamics, in particular, have been used to estimate changes in soil C at a local, regional, national or global scale. The modeling of C, especially on a regional scale, ranges from extremely simple approximations, where empirical relationships are taken and applied to large areas, to complex approximations, where results obtained from evaluating dynamics using simulators are then integrated with georeferenced data, typically using geoprocessing, in order to consider spatial differences in climate, soil and land use. In particular, for approaching C dynamics at a regional scale, with simulators integrated with GISs, there are applications for assessing management systems in diverse environments. In the United States, results obtained by the Century simulator (PARTON et al., 1987) were integrated with meteorological and soil databases generated in a GIS to estimate C sequestration potential for 44% of the country’s land area. From this information, it was found that conservation tillage practices and the use of cover crops could be utilized to increase soil C storage for about 40 years (DONIGAN et al., 1994). Applications of the RothC simulator (COLEMAN; JENKINSON, 1996), also associated with a GIS, in areas of natural ecosystems in New Zealand, showed that it is possible to assess impacts of climate change and land use on soil’s organic C and CO2 stocks. In this study, it was also estimated that the combined effect of ecosystem degradation and climate change can lead to a significant net release of CO2 for over 40 years of soil cultivation (PARSHOTAM et al., 1995). In Hungary, the C sequestration potential of various management practices was estimated using regression equations and RothC and Century simulators, associated with GIS, and it was found that there were differences between the methods, although estimates were on the same order of magnitude, and some management scenarios presented the same C mitigation potential as areas under reforestation (FALLOON et al., 2002). In Brazil, there are still few studies integrating these simulators with a geographic information system (GIS) to estimate C sequestration or stocks. In Rio Grande do Sul (RS), Century associated with a GIS was used to assess changes in carbon stocks occurring in two municipalities since the adoption of agriculture in 1900, up until 2050. In this study, a significant reduction in C stocks was observed since the introduction of the conventional system of agriculture, and a recovery of these stocks was observed after adoption of conservation practices, such as no-tillage and a complex crop rotation system (TORNQUIST et al., 2009). Soil C simulators, especially Century, EPIC (IZAURRALDE et al., 2006; WILLIAM; RENARD, 1985; WILLIAMS et al., 1989) and RothC, have also been used on a global scale to estimate the distribution of C and N in various land areas, as well as the effects of climate change (considering increases in temperature and CO2) on global soil C stock. In this context, the potential use of DNDC is also noteworthy (LI, 2000; LI et al., 1992) since this simulator represents processes involved in C and N dynamics in the soil, as well as in crop growth, thus making it possible to estimate both C sequestration and NO, N2O, CH4 and NH3 greenhouse gas emissions for agricultural ecosystems (LI, 2007; LI et al.,1992, 1994). For this reason, several studies highlight the application of DNDC to environmental conditions observed in various countries, including Australia, Canada, China, India, United States, New Zealand and Japan (BABU et al., 2006; BEHEYDT et al., 2004; GRANT et al., 2004; KIESE et al., 2005; LI, 1996; LI; SAGGAR, 2004; LI et al., 1996; MIEHLE et al., 2006; SAGGAR et al., 2003, 2007; SHIRATO, 2005; SMITH et al., 2002; ZHUANG et al., 2004).

The time scale of simulators is another factor to consider when choosing the tool for the intended application, as it can vary from hours to days, months and even several years. It should be noted, in particular, that the timescale for the application of C dynamics models ranges from shortterm daily and seasonal estimates of the amount of C and N in soil (e.g., DNDC) to long-term quantification of soil organic matter stocks and C sequestration (e.g., Century). Two classes of soil C and N dynamics simulators have emerged with different goals but with similar technology: shortterm simulators, which aim to predict soil C dynamics over a year or several years in a crop rotation system (e.g., Ceres) and long-term simulators, developed to estimate the status of C in soil throughout decades or centuries (e.g., Century). However, development of more advanced computers and demand for new technologies in agricultural and environmental sciences have enabled both types of simulators to be directed towards a common goal (e.g., DayCent simulator derived from Century) (DEL GROSSO et al., 2001). Depending on the complexity of the component models, simulators are usually classified as screening simulators (also known as reductive or summary simulators), intermediate simulators and research simulators. Screening simulators provide a preliminary view of the system’s behavior, with little detail, as they only use information that is highly relevant for the model representation. However, despite limited output, they have the advantage of requiring a reduced amount of input information to use. Intermediate simulators are situated between screening and research type simulators. They do not present details as meticulously as research simulators, but they incorporate heuristics and empirical data to processes in search of more information than what is offered by screening simulators. In this case the input data is a bit more sophisticated, but on a smaller scale than what is required by models classified as research simulators. Those simulators present rich detailing of mathematically modeled real processes and, therefore, become extremely complex in terms of input data requirements. However, they usually offer the user the option to select specific routines for the intended scenarios, thus decreasing (the extremely demanding) data input requirements. Furthermore, they allow for a more thorough investigation and reliable representation of the real system. DNDC can be mentioned as one of the simulators using models of this type. The increasing importance of multiple C pools in soil and in plant residue in the late 1970s and early 1980s represented a major change in the direction of research, resulting in a significant increase in the complexity of mathematical models and simulators. This complexity has developed in terms of the processes simulated, the rates estimated for each process, the role of soil microorganisms and the interaction between these processes (SHAFFER et al., 2001). The development of simulators is still mostly done using the FORTRAN, Pascal, C and C++ programming languages, although the use of specific computational package languages, such as Stella, linked to Java language resources, will tend to increase in upcoming years, considering its ability to interact with the internet. Furthermore, the significant increase in computer efficiency has spurred the development of more complex simulators, with the challenge of making them increasingly useful for the needs of its users. However, depending on the goal of the study, information available from databases or previously conducted research is not always usable in simulators producing estimates on greenhouse gas emissions, since there are environmental conditions (ecological, economic and social) that dramatically impact the management system locally used, and consequently, the results obtained. Additionally, the conditions under which such information was obtained must also be taken into account, as they do not always allow extrapolation for use in other areas or

situations, as already pointed out. As an example, there are conditions involving the physicalchemical nature of the soil, where, for example, organic matter content and density may vary for the same type of soil. Thus, the development of databases – in particular databases that are able to effectively bring together essential information currently scattered throughout Brazil in order to simulate scenarios to estimate greenhouse gas emission – should focus on providing local aspects which have low variability and which are related to specific features of crop varieties (development time in degree days, maximum grain yield, quantity of dry matter, (harvested) grain fraction of total biomass, N fixation index, quantity of C present in various parts of the plant, and grain C/N ratio) and animal breeds (digestibility rates as a function of feed supplied and development stage), as well as daily climate information (maximum and minimum temperature, rainfall, solar radiation, relative humidity and wind speed). The following is a presentation of more detailed considerations on some of the main simulators for evaluating carbon and nitrogen dynamics applied to the study of greenhouse gases.

Simulators for the study of C and N dynamics Contribution to understanding environmental aspects Simulators modeling C and N dynamics are important components of ecosystem simulators, which have been intensively used to expand knowledge about the impacts of land use on environmental quality. Carbon sequestration and soil quality Long-term stabilization of soil organic carbon (SOC) content has important implications for sequestration of atmospheric CO2 and maintenance of soil quality. Effects of land use on carbon stabilization can be estimated from changes in SOC content under various management systems and specific soil and climate conditions. Those estimates require data from long-term experiments and are thus restricted to a limited number of locations. Observations obtained from these experiments can be extrapolated to a wider variety of conditions through the use of ecosystem simulators that include transformations of C and N in the soil and in the plant. These simulators, such as Century or RothC, have been used to estimate changes in SOC in areas of pasture, organic and mineral fertilization and crops and tillage systems. Recently, simulators are being used to estimate changes in SOC in management practices as part of regional and national CO2 emission studies. As a result of changes in SOC representing differences between C inputs due to net primary productivity (NPP) and C losses due to heterotrophic respiration (HR), ecosystem simulators used for estimating C sequestration should be able to estimate the effects of land use on NPP and HR. Simulation of NPP must be clearly distinguished from simulations of phytomass growth, where NPP is represented using various models. NPP is the difference between gross primary production and autotrophic respiration spent on maintenance and growth of phytomass. NPP includes aboveground phytomass – and litter, which is largely in the roots. Accurate simulation of the dynamics of residues above and below ground is thus extremely important in order for a simulator to quantify changes in SOC.

An essential aspect of HR modeling is the rate at which labile C from litter becomes protected from rapid decomposition, thus reducing HR. This rate is typically dependent on the soil’s clay content (VEEN; KUIKMAN, 1990) and lignin content (PAUSTIAN et al., 1997), which determine the soil’s ability to protect labile C. Therefore, clayey soils show less HR than sandy soils with the same C input. Accurate simulation of the rate at which labile C is protected is essential, as this rate determines soil C sequestration in the medium and long term. Greenhouse gas emissions and air quality As a result of N2O and CH4 being radiatively active in the atmosphere, there is major interest in estimating the net exchange of these gases between terrestrial ecosystems and the atmosphere as part of climate change studies. Those estimates are important for evaluating management options to reduce emissions of these greenhouse gases, mainly due to national and international agreements to minimize their release into the atmosphere. Emissions of NO, N2O and CH4 are highly variable in space and time. Using system simulation techniques to research the environmental impact of greenhouse gases generation and emission processes by agroecosystems involves three main challenges: 1) some of the gases (e.g., NO and N2O) have multiple sources or origins, such as nitrification, denitrification and methanogenesis; 2) all gases are produced and consumed simultaneously in soils, controlled by the kinetics of a series of geochemical and biochemical reactions; and 3) there are a large number of environmental variables controlling biochemical reactions. Therefore, to simulate agroecosystems with potential to generate greenhouse gases, all factors must be considered, including ecological aspects, soil environment variables and biogeochemical reactions (Figure 1).

CHAP 9 - FIGURE 1 Reações bioquímicas/geoquímicas

Biochemical/geochemical reactions

Fatores ambientais

Environmental factors

Variáveis de controle ecológicas

Ecological control variables

Produção e consumo de gases-traços no Trace gas production and consumption in sistema solo-planta the soil-plant system Movimento mecânico

Mechanical movement

Dissolução/cristalização

Dissolution/crystallization

Combinação/decomposição

Combination/decomposition

Oxidação/redução

Oxidation/reduction

Adsorção/dessorção

Adsorption/desorption

Complexação/decomplexação

Complexation/decomplexation

Assimilação/dissimilação

Assimilation/dissimilation

Gravidade

Gravity

Radiação

Radiation

Temperatura

Temperature

Umidade

Humidity

Substrato

Substrate

Clima

Climate

Propriedade do solo

Soil properties

Vegetação

Vegetation

Atividades antropogénicas

Anthropogenic activities

Figure 1. Variables, factors and reactions involved in production and consumption of trace gases in the soil-plant system contemplated by the DNDC simulator. Source: Adapted from Li (2000).

The important soil environment variables for simulation are especially associated with temperature, O2 concentration and water content at depth, in addition to readily available C resulting from NPP. Those variables must be related to conditions of the specific site under study, as temporal and spatial emission patterns are highly variable and complex (GRANT, 2001). Consequently, the predictive value of short-term flux measurements for long-term estimates is limited to locations having similar soil and climate conditions. Soil management Land use practices involving soil changes and vegetation removal have caused releases of C into atmosphere. Those releases account for about 1/3 of atmospheric CO2 accumulated since the pre-industrial period and indicate an intense degradation process. Certain practices such as notillage, crop rotation, use of legumes and use of mineral and organic fertilizers have been adopted in order to increase C stocks and sequester atmospheric CO2. Ecosystem simulators have been used to estimate changes in SOC under various management practices and must be able to adequately represent the processes by which these practices affect the SOC. For example, the effect of tillage systems on SOC must be quantified using simulations that take into account the residue on the soil surface with its own microbial population and microclimate (temperature and water content), as contemplated in the Century simulator. These residues must also affect the energy balance on the soil surface and, consequently, the temperature fluxes and subsurface heat. Tillage practices available in simulators must influence the intensity and the depth at which surface residues are incorporated throughout the soil profile. This incorporation increases contact between the residue and the soil’s microbial population and microclimate, thus increasing its decomposition rate. Incorporation also reduces the heat flux from the residue, promoting faster heating and cooling processes. Soil tillage has an additional effect on release of CO2 due to increased contact between soil C and the microbial population, which also eed to be considered in simulators. In some of these tools (LI et al., 1992; MOLINA et al., 1983; PARTON et al., 1987) this effect can be simulated by redistributing part of the C from recalcitrant pools to labile pools (e.g., passive and active pool, respectively, in the Century model). Tillage alters the properties of the soil surface, such as roughness and residue coverage, and also of the subsurface, such as soil density. Such changes are also included in some available simulators (RICKMAN et al., 2001; SHAFFER; LARSON, 1987; WILLIAMS et al., 1989).

To simulate the effects of grasses and perennial legumes versus annual crops on SOC it is necessary to establish differences in C allocation between shoot and root. This allocation is heavily influenced by climate, including radiation, temperature and atmospheric CO2 concentration, and by soil properties, such as water and nitrogen. Ecosystem simulators must therefore estimate the root system’s NPP, including root recycling and rhizodeposition, separately from above-ground NPP, in a way that represents local effects on NPP and on allocation of C in roots without the need to calibrate them for a specific location (GRANT, 2001).

Biogeochemical processes in C and N dynamics simulation Several tools are used to understand alterations to an ecosystem resulting from land use changes. In this sense, it has been observed that understanding C and N dynamics is essential to improve the quality of simulators in performing accurate predictions of the effects of land use changes and responses to global climate change. Modeling of biogeochemical processes emerged in the 1930s and currently it is systematically addressed by an extensive list of simulators, whether empirical or mechanistic. There are, however, variations in terms of the complexity and mathematical description of the various biological and geochemical processes involved, which are represented in most simulators by primary productivity, mineralization, immobilization and humification. Currently, with the introduction of simulators of greenhouse gas emission dynamics – especially nitrous oxide and methane – nitrification and denitrification processes are also incorporated. Primary productivity Correct representation of processes related to NPP above and below the soil is essential for any simulator aiming to estimate effects of land use and climate change on SOC. NPP is represented at various levels of complexity in various simulators. Below-ground NPP is a determinant of SOC stocks, as, under certain conditions, up to 40% of the shoot’s C can be transferred to the roots (GRANT, 2001). In most simulators, below-ground NPP is represented as a fraction of above-ground NPP, which can be constant (PARTON; RASMUSSEN, 1994) or dependent on phenology (HANSEN et al., 1990). Spatial distribution of below-ground is sometimes represented in these computational tools using a logarithmic function of the soil’s depth. This NPP, however, is the largest fraction of above-ground NPP, and its spatial distribution is modified depending on species, perennial or annual, as well on water and nutrient content. In deeper root systems, there are higher rates of C accumulation, often measured in areas under forages and annual crops (BREMER et al., 1994; GRANT, 2001). Therefore, changes in the allocation of C between root and shoot caused by growth habit or soil conditions need to be represented in C dynamics simulators used to estimate changes in SOC (PARTON; RASMUSSEN, 1994). One approach to simulating allocation of C between root and shoot is by using the functional balance hypothesis proposed by Thornley (1995), where transport of C and N occurs as a function of a concentration gradient generated in the plant by C fixation and N uptake versus consumption of C and N in the root and shoot. This hypothesis allows root growth versus shoot growth to adapt to changes in environmental conditions and to be parameterized independently of location-specific root growth data (GRANT, 2001). Mineralization and immobilization

Several models use the same approach for HR simulation, i.e., the inclusion of first-order kinetics where metabolic demand of the soil biomass exceeds supply from the substrate. This assumption, however, becomes invalid when the supply from the substrate exceeds the metabolic demand from the biomass. This indicates that, in future development of simulators, a more mechanistic approach to soil microorganisms is essential. The role of these microorganisms is extremely important for C and N dynamics. They respond to environmental changes so dynamically that most simulators cannot produce adequate estimates. Not only do microorganisms compose C and N pools, they also catalyze most processes. It is, therefore, essential to properly simulate their role in the soil environment. These differences in approaches to microbial decomposition are necessary because C and N transformation rates may be inadequate to the first-order kinetics equation commonly used. An alternative to this equation is one where microbial activity is represented as a C and N transformation agent, using Monod’s kinetics (SMITH, 1982): ri = µm Ci / Ks + Ci * B where ri is the decomposition rate; Ci is the C content in the pool; B is the size of the microbial biomass; μm is the maximum mineralization rate; and Ks is a constant. For simulators that do not take into account the role of microorganisms, maximum mineralization rate and size of microbial biomass can be combined to adjust to Michaelis-Menten’s kinetics. The above equation can be simplified as first-order kinetics (ri = k1Ci) or as zero-order kinetics (ri = k0), in which k1 and k0 are the first- and zero-order coefficients, respectively, which can be modified by temperature, pH, O2 and microbial effects. The NLEAP simulator (SHAFFER et al., 1991), used to quantify nitrate leaching, takes a relatively simple approach for immobilization and mineralization processes. The net mineralization of N in the plant residues pool depends on its C/N ratio and decomposition rate: Mr = rr (1 / (C/N) r − 1/30), where Mr is the net nitrogen mineralization rate in the residues pool; C/N is the C-to-N ratio in the residues pool; and rr is the residues decomposition rate. Once the C/N ratio reaches between 6.5 and 12, depending on the type of residues, the C and N remaining in the residues are transferred to the fast cycling SOC pool. The decomposition rate is calculated by first-order kinetics: rr = kr Tf Wf Cr, where kr is the constant rate and can be adjusted based on the C/N ratio; Cr is the C content in the residues; and Tf and Wf, are coefficients associated with water and temperature. In RZWQM (AHUJA et al., 2000), which simulates soil water, plant growth and C and N dynamics in the soil, mineralization and immobilization processes are determined by the decomposition of organic pools and the growth of microorganism. There isn’t a previously defined C/N ratio to control net mineralization and immobilization, as is the case with the NLEAP simulator. Decomposition of organic matter in each pool is simulated by the first-order equation and modified by the effects of soil temperature, oxygen, heterotrophic aerobic microbial population, aerobic condition and ionic strength. In Century, surface residues are divided into metabolic and structural pools. The structural pool is subdivided into cellulose and lignin components. The metabolic C and cellulose pools are transferred to the fast cycling C pool (microbial biomass), whereas C bonded to lignin is directly allocated to the slow C pool (PARTON et al., 1994). The fast cycling (active) microbial pool has four destinations during the decomposition process: 1) slow C pool; 2) passive C pool; 3) pool of soluble C which can be leached; and 4) CO2. Decomposition of all pools occurs through first-order kinetics.

The decomposition rate is specific for each pool. For the structural pool, it is expressed by the equation: k = kce−3LsAtAw, where Ls is the lignin content in the structural C pool, Kc is the constant rate, and At and AW are factors related to temperature and water. For the active C pool, the decomposition rate is influenced by soil texture (Tm) in the equation k = kc AW At Tm, where Tm = 1 – 0.75 (Fsilt+Fcla). For the remaining C pools (metabolic, slow C and passive C), decomposition rates are functions of just At and AW. Humification Greater C stabilization in soils with high clay content was demonstrated by Sorenson (1981). Since then, soil clay content has been increasingly used in ecosystem simulators to reduce C decomposition and allocate products of that decomposition between pools with different decomposition rates. Verberne et al. (1990) divided the products of microbial decomposition into non-protected and protected organic fractions, using, in the case of the latter, coefficients ranging from 0.3 for sandy soils to 0.7 for clayey soils. Veen and Kuikman (1990) and Whitmore et al. (1991) suggested that efficiency of substrate use increases with clay stabilization. However, some studies concerning the effect of clay on substrate use have not endorsed this hypothesis (VEEN et al., 1985). Li et al. (1992) proposed that decomposition rates of organic substrates, using first-order kinetics, were reduced according to clay content. On the other hand, this hypothesis does not explain the marked increase in recovery of C as amino acid or microbial biomass, in soils with high clay content. It is more likely that microbial products and metabolites, as well as lignin degradation products, are stabilized by clay surfaces. Parton et al. (1987) thus adopt lignin content to allocate the products of decomposition of slow and active pools of the model incorporated into Century. Grant (2001) combined some of the products from lignin and protein hydrolysis and carbohydrates, following the stoichiometry proposed by Shulten and Schnitzer (1997), and allocated the resulting compounds in a complex of particulate organic matter, according to soil’s clay content. The rates of particulate organic matter formation were therefore a function of the lignin content in the residues, clay and heterotrophic microbial activity. Those rates contributed to long-term changes in the C of the simulated soil, under various management practices, which were corroborated by field measurements (GRANT, 2001). Nitrification and denitrification Nitrification is an important process for controlling N transformations between less (NH4+) and more (NH3) mobile forms. Accurate representation of nitrification in ecosystem simulators is therefore necessary to simulate losses of N through leaching and denitrification. Nitrification has currently been incorporated into simulators because it is an important cause of N2O emissions following fertilization. It is influenced, however, by several environmental factors, including substrate concentration (NH4+ and CO2), aeration, temperature and pH. In simpler simulators, nitrification rates have been represented by functions of variable orders (zero- and first-order) based on the rates of NO3− formation observed in field trials. In more complex simulators, first-order functions based on Michaelis-Mentem kinetics are associated to nitrifying biomass and specific activity. In most simulators, these functions can be modified by those associated with temperature, pH and porosity, which may be dependent upon soil texture. Some simulators consider that mineralization occurs at a much lower rate than nitrification, so the first inorganic N to be formed is nitrate (NO3−), instead of ammonium (NH4+). In addition, they

assume that the ammonium fertilizer applied is converted into nitrate immediately, not in the course of several days or weeks, as observed in the field (HANSEN et al., 1995). Century incorporates this assumption. Because it uses a monthly time-step, nitrification does not become important. In fact, the simulator does not differentiate ammonium from nitrate in the soil. DayCent (the daily version of Century) considers nitrification to be proportional to the soil N cycling rate (PARTON et al., 1996). In this simulator, N2O production resulting from nitrification is calculated both by the N cycling rate and by the excess of NH4+ in the soil (> 3 μg N g−1) as RN2O = NH2ONpHNt (kmx+NmxNNH4), in which RN2O is the flux of N2O; NH2O, NpH and Nt are factors related to water, pH and temperature, respectively; kmx is the N recycling coefficient; Nmx is the maximum flux of N2O based on the excess of H4+; and NNH4 is the effect of the soil’s NH4+ on nitrification. Denitrification is an important cause of N loss by gaseous emission, especially N2O. Simulating denitrification is extremely empirical, not only because of the relatively unknown nature of the process itself, but also due to the spatial and temporal variability of the soil’s anaerobic conditions. Mathematical simulation of denitrification can also use zero-order kinetics, first-order kinetics or Michaelis-Mentem kinetics. In the DayCent simulator, denitrification is simulated to calculate emissions of N2 and N2O. The total flux of N (N2 + N2O) is estimated by the following equation: Dt = min (Fd (NO3), Fd (CO2)) Fd(WFP), in which Dt is the total gas flux; Fd(NO3) and Fd(CO2) are the maximum flux of gaseous N for a given NO3− and the soil respiration rate, respectively; and Fd(WFP) is the effect of water on denitrification.

Simulators for studying C and N dynamics and greenhouse gases There is a large collection of simulators created by the agriculture and forestry sector to assess the dynamics of C and N and to study greenhouse gas emission (Table 1). Simulators like RothC, Century or EPIC have been widely used to estimate C sequestration in various agroecosystems, mostly in temperate regions, with few studies in tropical environments. However, due to the complexity of the models incorporated into these simulators, several researchers have been motivated to develop simpler models, such as the one available in the CQESTR simulator. Furthermore, the models incorporated into the DNDC and DayCent simulators are capable of estimating N2O and CH4 emissions for wetlands and flooded rice crops as well.

Table 1. Simulators for C and N dynamics and greenhouse gases. Simulators DNDC crop(1)

Ceres(2)

GePSI(3)

RothC(4)

Century(5)

Ecosys(6)

DayCent(7)

EPIC(8)

General Resources Time step

Daily

Daily

Hourly

Monthly

Monthly

Hourly

Daily

Daily

Simulation period (years)

1–102

1–10

1–10

1–102

1–102

10-1–10

1–102

1–102

Plant pool

4

4

5

0

3

>50

5

3

Organic C pool in the soil

8

4

0

5

8

~ 40

8

8

Inorganic N pool in the soil

7

2

2

0

2

1

2

2

Processes simulated Soil temperature

X

X

X

X

X

X

X

X

Soil moisture

X

X

X

X

X

X

X

X

Phenology

X

X

X

Leaf area index (LAI)

X

X

X

X

Photosynthesis

X

X

X

X

X

Respiration

X

X

X

X

X

Rooting process

X

X

X

X

N absorption

X

X

X

X

X

X

X

Effect of water on the crop

X

X

X

X

X

X

X

Effect of N on the crop

X

X

X

X

X

X

X

Effect of CO2 on the crop

X

Decomposition

X

X

X

X

CH4 emission

X

X

X

Mineralization

X

X

X

X

Nitrification

X

X X

X

X X

X

X

X

X

X

X

X

X

Denitrification

X

X

X

X

X

N trace gas emission

X

X

X

Output Variables Soil temperature

X

X

X

X

X

X

X

Soil moisture

X

X

X

X

X

X

X

Phenological stages

X

X

X

Leaf area index (LAI)

X

X

X

C pool in the plant

X

X

X

X

X

X

X

N pool in the plant

X

X

X

X

X

X

X

C pool in the soil

X

X

X

X

X

N pool in the soil

X

X

X

X

X

CH4 emission

X

X

X

X

X X

X

N trace gas emission

X

X

DeNitrification-DeComposition (DNDC) The DeNitrification-DeComposition (DNDC) simulator (UNIVERSITY OF NEW HAMPSHIRE, 2003) is a process-oriented tool developed by Li et al. (1992), initially to simulate C and N dynamics in soil and, thereby, estimate C sequestration as well as emission of NO, N2O, CH4 and NH3 for nonflooded agricultural lands. It was designed to help understand the effects of anthropic activities on increasing greenhouse gas emission rates, seen as a major factor influencing the balance of atmospheric trace gases. In light of this goal, primary processes controlling interactions between ecological variables, soil environment factors and biogeochemical reactions were incorporated. Subsequently, crop growth aspects were incorporated through use of growth curves (Li et al., 1994), while still not considering the effect of climate changes on crop growth, or the resulting interactions with biogeochemical cycles. Currently, the simulator allows for estimating greenhouse gas fluxes generated by the world’s major agricultural ecosystems through integrating detailed aspects on the production system – such as existence of crop rotations, soil management, use of fertilizer and natural manure, irrigation water management, weed control, among others – with biogeochemical processes of nitrification, denitrification, crop growth, water infiltration into the soil and litter production (BABU et al., 2005; LI, 2000). The simulator considers ecological variables governing environmental factors relevant in the focus of their development, such as: a) local abiotic factors (maximum and minimum temperature, rainfall and solar radiation); b) physical-chemical aspects associated at various depths the with soil type where the crop develops (soil density, field capacity, wilting point, saturation, pH, etc.); c) specific characteristics of cultivated varieties; d) anthropic activities (crop, soil, water, and fertilizer management). These variables impact the environmental factors resulting from the processes contemplated by the tool, which, in turn, impact the soil climate, the effect of temperature and moisture on decomposition, and decomposition itself. DNDC can be represented by two major components, as established by Li (2000) (Figure 2). The first component includes the soil climate, plant growth and decomposition submodels, which, in turn, allow for evaluation of trends in the effect of variables related to climate, soil properties, existing vegetation and anthropic activities on soil temperature, moisture, pH, redox potential (Eh) and substrate concentration. Generally, the three submodels interact to present information on the variables in a daily time-step. The soil climate submodel calculates temperatures, moisture, and Eh in the soil profile, based on information on air temperature, rainfall, the soil’s thermal and hydraulic properties and oxygen availability. This information is considered along with information on crop characteristics, climate and soil properties and management practices, enabling the evaluation of crop growth aspects and their effects on soil temperatures, moisture, pH, Eh, dissolved organic C and available N concentrations, through the crop growth submodel. The same information is also considered by the decomposition submodel to allow simulating substrate concentrations (dissolved organic carbon, NH4+ and NO3−). The mathematical equations used to represent the mathematical models of these components are detailed in Li et al. (1992, 1994). CHAP 9 - FIGURE 2

Variáveis ecológicas

Ecological variables

Climáticas

Climatic

Solo

Soil

Cultura

Crop

Manejos

Management practices

T média anual

Mean annual T

ET potencial diária

Potencial daily ET

IAF regulado albedo

Albedo regulated LAI

Evaporação

Evaporation

Transpiração

Transpiration

Fluxo de água entre camadas

Water flow between layers

T solo no perfil

Soil T in the profile

Umidade do solo no perfil

Soil moisture in the profile

Difusão oxigénio

Oxigen diffusion

Eh solo no perfil

Soil Eh in the profile

Clima do solo

Soil climate

Demanda diária de água

Daily water demand

Acúmulo diário de biomassa (IAF)

Daily biomass accumulation (LAI)

Absorção de água pela raiz

Water absorption through roots

Demanda de N

N demand

Grãos

Grains

Ramos

Branches

Raiz

Roots

Estresse de água

Water stress

Absorção diária de N pela raiz

Daily absorption of N through roots

Respiração da raiz

Root respiration

Efeito da temperatura decomposição

e

umidade

na Effects of temperature and moisture on decomposition

Muito solúvel

Very solluble

Solúvel

Solluble

Resistente

Resistant

Microbiano lábil

Labile microbial

Microbiano resistente

Resistant microbial

Hematos solúveis

Solluble hemato

Hematos resistentes

Resistant hemato

Humus passivo

Passive humus

Decomposição

Decomposition

Variáveis ambientais do solo

Soil environmental variables

Temperatura

Temperature

Umidade

Moisture

Substrato (NH4+, NO3- e Carbono Orgânico Substrate (NH4+, Dissolvido) Organic Carbon)

NO3-

Denitrificação

Denitrification

Denitrificador nitrato

Nitrate denitrifier

Denitrificador nitrito

Nitrite denitrifier

Carbono orgânico dissolvido

Dissolved organic carbon

Nitrificação

Nitrification

Nitrificadores

Nitrifiers

Argila

Clay

Fermentação

Fermentation

Eh do solo

Soil Eh

Aerênquima

Aerenchyma

Produção

Production

Oxidação

Oxidation

Transporte

Transport

and

Dissolved

Figure 2. DNDC simulator and its components. The first component consists of the soil, climate, plant growth and decomposition submodels; the second consists of the nitrification, denitrification and fermentation submodels. Source: Li (2000).

The second component of the DNDC conceptual model proposed by Li (2000) is represented by the nitrification, denitrification and fermentation submodels, which allow for estimating fluxes of NO, N2O, CH4 and NH3 generated during simulation of impacts on soil environmental conditions that are relevant to biogeochemical reactions. The factors controlling nitrification in DNDC are: soil temperature, moisture, pH and NH4+ concentration. The nitrification submodel estimates nitrification rates by following nitrifying activities and NH4+ concentration. NH4+ oxidizer growth and mortality rates are calculated based on concentration of dissolved organic C, temperature and moisture. DNDC calculates nitrification induced by production of NO and N2O as a function of the expected nitrification rate, and based on temperatures (LI, 2000). The denitrification submodel considers sequential reduction of nitrate into dinitrogen (N2) driven by the presence of denitrifying bacteria under anaerobic conditions (LI, 2000). Thus, in the

DNDC, denitrification rates are controlled by soil moisture and Eh, by temperature and by substrate concentrations (dissolved organic carbon, NO3−, NO2−, NO and N2O). The model contemplated by the simulator considers the soil matrix divided into two parts: aerobic and anaerobic, and only the substrate allocated to the anaerobic part is considered for the denitrification effect. The relative growth rate of denitrifiers is described as a multinutrient-dependent growth function. The mortality rate of denitrifiers is a constant fraction of the total denitrifier biomass. The models entered into the simulator also assume that relative growth rates for denitrifiers with various substrates are independent and that bacteria competition occurs via the common substrate of dissolved organic carbon. Substrate consumption rates are calculated using the Pirt equation (LI, 2000). DNDC also allows calculating NO and N2O diffusion rates in the soil matrix by means of a function which takes into account soil porosity, moisture, temperature and clay content. In the case of methane (CH4) emission, Li (2000) reports that DNDC takes into account important variables to describe methanogenesis and methanotrophy processes. Methanogenesis occurs by means of the reduction of available C in the soil to CH4 through anaerobic microbial activity, which will necessarily occur when the soil’s redox potential (Eh) is at low levels (−150 mV to −200 mV). Additionally, methane production is also strongly influenced by an increase in temperature (with optimal production registered in the 30 °C to 40 °C range). So, to estimate methane (CH4) emission rates, DNDC considers the following variables: carbon content available in soil (i.e., dissolved organic carbon), soil Eh and soil temperature. DNDC also calculates the methane oxidation rate as a function of CH4 concentration in the soil and Eh, as well as diffusion of this gas throughout the soil’s layers, as a function of its concentration gradients, and soil porosity and temperature. In addition, the simulator also incorporates estimates on the flux of CH4 carried by plants, as a function of gas concentration and presence of aerenchyma, throughout the cultivation period. In the absence of vegetated soil or of plants with well-developed aerenchyma, DNDC considers the ebullition factor as a source of methane emission, considering that it occurs only on the soil surface layer. Thus, the ebullition rate is evaluated by DNDC as a function of the soil’s CH4 concentration, temperature, porosity and the presence of aerenchyma in the plant. When considering NH3emission, it is known that the concentration of this gas in the soil is governed by chemical reactions in the soil’s liquid phase. DNDC calculates liquid NH3 concentration based on OH− concentrations (depending on soil pH and temperature) and NH4+ concentrations (provided by the decomposition submodel). Subsequently, NH3 concentration in the soil’s gas phase is calculated in proportion to the NH3 concentration found for the liquid phase and the soil temperature. In addition, the simulator assumes that the NH3 fraction emitted daily is related to the soil’s air-filled porosity and clay content, due to their effects on the diffusion of NH3. The simulator also incorporates aspects related to the potential for absorption and metabolization of NH3 by plants, since it has been demonstrated that there is a linear relationship between the dry NH3 deposition rate and the NH3 concentration in the air (LI, 2000). Thus, the concept of N deposition rate can be represented by the ratio of the N absorption rate (μg m−2 s−1) to the concentration of NH3 in the air (μg m−3), and it ranges from 0.003 m s−1 to 0.034 m s−1 for various crops. For this reason, DNDC adopts 0.034 m s−1 as a default value to calculate the rate of NH3 absorption by crops, although it also considers other factors, such as availability of N in the plant and moisture on the leaf surface (LI, 2000). Further details are available in Li (2000), which notes that DNDC makes it possible to evaluate total N content in the crop for the entire growing season. Upon detecting a decrease in total N content, DNDC reports the reduced part as NH3 flux released from plants.

The fermentation submodel contains equations related to methane and calculates production, oxidation and transport of this gas under submerged conditions. The denitrification submodel calculates production, consumption and diffusion of N2O and NO during flooding, irrigation or rainfall events; meanwhile, the nitrification submodel also includes functions to estimate NH3 production and volatilization. DNDC data entry for simulation in local mode requires the following variables: • Daily temperatures (maximum and minimum). • Daily rainfall. • Soil density. • Soil texture. • Organic carbon content. • Soil pH. • Management practices: crop type, rotation, no-tillage or conventional tillage, fertilization, incorporation of organic manure, irrigation, flooding, pasture and weeds. Based on the input data, DNDC first calculates soil temperature, moisture, Eh, pH and substrate concentration, on a daily basis, and then uses environmental variables in the nitrification, denitrification, CH4 production/oxidation submodels and in other relevant biogeochemical reactions to present daily estimates of NO, N2O, CH4 and NH3. The soil climate and denitrification submodels run on an hourly time step, unlike the others, which run on a daily time step. To use the simulator in regional mode, Li (2000) mentions that there is also a need for information previously scanned into databases (data planes) made available in a georeferenced information system (GIS), to allow for spatio-temporal assessment of gas emissions at a regional scale (LI et al., 1996). What differentiates the DNDC simulator is the ability to create management scenarios, both current and future, which allow for estimating greenhouse gas emissions from agricultural sources resulting from the practices reported. Therefore, provided that it is supported by previous stages of validating the simulator for the Brazilian environment, at the end of the simulated period it provides information that allows for a view of the direct effect of adopted practices on current emissions, as well as proposing management practices combining mitigation with productivity. CQESTR CQESTR is a simulator that has been calibrated and validated for temperate regions and that allows evaluating effects of management practices on carbon stocks (RICKMAN et al., 2001) (Figure 3). The model included in the tool estimates additions of C originating from crop residues, and losses of C resulting from microbial oxidation. In nature, residues added to the soil are decomposed and slowly incorporated into soil organic matter (SOM). In the CQESTR simulator, the period associated with this transformation is approximately four years, obtained from calibrations of the simulator with long-term soil C data.

CHAP 9 - FIGURE 3 Resíduos da planta

Plant residues

Decomposição

Decomposition

Fase

Stage

Resíduo superfície

Surface residues

Preparo do solo

Soil preparation

Resíduo incorporado (Camadas 1,…n)

Incorporated residue (layers 1,…n)

Resíduo na raiz (Camadas 1,…n)

Residue on roots (layers 1,…n)

Transição

Transition

Matéria orgânica do solo (Camadas 1,…n)

Soil organic matter (layers 1,…n)

Figure 3. CQESTR simulator and its components. Source: Rickman et al. (2001).

To estimate total C stocks (CT), CQESTR considers, for a daily time step and for multiple layers, the unit of weight of crop residues per unit of area within each layer: CT = (COM − DOM) + (CR − DR) + (CA − DA). COM is the amount of organic matter present in the soil at a given initial point, to which organic matter in the form of crop residues (CR) or organic amendments (CA) are added. Soil carbon is lost through daily decomposition of the organic matter (DOM), decomposition of crop residues (DR) and decomposition of organic amendments (DA). The net gain or loss of COM is determined by the cumulative daily loss (DOM) and the periodical contributions (in the simulator) due to CR − DR and CA − DA. CT is a dynamic value, varying with monthly additions of residues and daily losses through decomposition. The organic matter in the soil, COM, is a relatively static variable. The daily amounts to be decomposed (DR and DA) are small, but ever-present, and both CR − DR and CA − DA are small after the four year composting period used in the simulator. In CQESTR, the decomposition equation contains a decomposition rate k, the quantity of heat accumulated (in cumulative degree days, CDD), and four terms related to residues or environment, which modify the decomposition rate. These terms include a factor of nitrogen content in the residue (fN), a water factor (fW), a soil texture factor (fX) and a biomass factor (fB). The term fN provides different decomposition rates for residues rich and poor in nitrogen. The value for fW is determined by the location of the residue, incorporated or on the surface of the soil and by the presence or absence of crops in growth stage. The texture factor is not yet functional in the current version of the simulator. Its value varies between 0 and 1, depending on the impact of clay and sand contents, as advocated by Parton et al. (1987). The biomass factor differentiates fresh residues, roots, decomposed material and native soil organic matter. Values for fB were determined by calibration with long-term observations of SOM, obtained at the USDA Research Center, Oregon. The remaining residue stock (in units of weight of residue per unit of area) is computed through the addition to each layer from the initial amount of residue (Ri) and amount of heat accumulated at the time of the addition of the residue to the soil: Rr = Ri * exp (k * fN * fW * fX * fB * CDD) where exp is an exponential function.

CQESTR is considered a simple simulator, primarily due to the lesser number of input variables required for its use. The main input variables are initial SOM content, bulk density, textural class, N content in crop residues and mean monthly temperatures and rainfall. Century – DayCent Century incorporates a mechanistic model to simulate, in the long term (10,000 years), the dynamics of soil organic matter (SOM) and nutrients in the soil-plant system. Thus, it makes it possible to evaluate the impact of various management practices on C and N dynamics, as well as P and S dynamics. The Century simulator is composed of several submodels: the submodel for soil organic matter dynamics, the water submodel and the plant production submodel. The water and plant production submodels include most of the variables (soil temperature and moisture, nutrient uptake by plants and quantity and quality of plant residues) required for the SOM dynamics submodel. The simulator runs on a square-meter scale and simulates the 0 cm to 20 cm surface layer, using a monthly time step. The main input variables are: mean, minimum and maximum monthly air temperature, monthly rainfall, soil texture (sand, silt and clay content), nitrogen content, plant material lignin content, N intake from the atmosphere and soil and initial contents of C and N in various pools of the soil. Century includes three SOM pools (active, slow and passive), with different decomposition rates, plant residue pools above and below the soil and a surface microbial pool (Figure 4). Plant residues are divided into: surface – comprising shoot residues; and soil – comprising root system residues. These fractions are partitioned into two pools: structural, which presents a recycling time of 1 to 5 years; and metabolic, immediately decomposed by microbial action, with a recycling time from 0.1 to 1 year. The partitioning into these pools is done in accordance with the tissues’ ligninto-nitrogen ratio (L/N). As the ratio increases, most of the residue is allocated to the structural pool. The SOM is divided into three pools: active – consisting of soil microbial biomass and its products, easy decomposable and with short recycling time (1-5 years) depending on environment and sand content; slow – derived from resistant plant material (lignin) and from chemically and physically protected OM, with an intermediate recycling time (20 to 40 years); and passive – material very resistant to decomposition, being chemically recalcitrant and physically protected, with long recycling time (200 to 500 years).

CHAP 9 - FIGURE 4 Parte aérea

Above-ground part

Estrutural Lignina | Celulose

Structural Lignin | Cellulose

Metabólico

Metabolic

Microbiano da superfície

Surface microbial

Parte raiz

Root

Ativo

Active

Lento

Slow

Lixiviado

Leachate

Passivo

Passive

L/N = Lignina / Nitrogénio

L/N = Lignin / Nitrogen

A= Fator de decomposição abiótico

A= Abiotic decomposition factor

T= Teor silte + argila

T= Sand + clay content

Ts= Teor de areia

Ts= Sand content

Tc= Teor de argila

Tc= Clay content

Ls= Fração do carbono estrutural que é Ls= Fraction of structural carbon that is lignin lignin Água lixiviada na camada abaixo de 30 cm -1

Leachate water in layer below 30 cm

Taxa de decomposição maxima (ano )

Maximum decomposition rate (year-1)

Fração do resíduo metabólico

Metabolic residue fraction

Taxa de decomposição do carbono (ano-1)

Carbon decomposition rate (year-1)

Figure 4. Century simulator pools and fluxes. Source: Parton et al. (1987).

DayCent is the daily version of the Century simulator and it estimates C and N fluxes between the atmosphere, vegetation and soils. The Century simulator operates on a monthly time step because this degree of resolution is suitable for simulating medium- and long-term changes (10–100 years) in SOM, plant production and other ecosystem variables, in response to changes in climate, land use and concentration of atmospheric CO2. Simulations of gas fluxes, however, require a shorter time scale, as the majority of total gas flux resulting from short-term rainfall or irrigation and the processes resulting in emission of these gases are constantly responding to changes in soil water content in a non-linear way. Submodels for plant production, residue decomposition and soil organic matter, in addition to soil water, temperature dynamics and trace gas fluxes, are included in the DayCent simulator (Figure 5). Plant growth is controlled by nutrient availability, water and temperature. Carbon and nutrients are allocated to leaves, stem and root biomass, depending on the type of vegetation. Transfer of C and nutrients from dead material to SOM pools and transfer of available nutrients is controlled by material’s lignin content and C/N ratio, decomposition factors related to water and temperature, and soil texture. Maximum and minimum temperatures and daily rainfall, description of management practices and soil texture data are required as input variables. Recent modifications made to DayCent have included effects of solar radiation on plant growth (DEL GROSSO et al., 2002). Comparisons of the simulator’s results with measured data have shown that the model adequately simulates crop production, SOM stocks and trace gas fluxes for various native and managed systems (DEL GROSSO et al., 2002, 2005).

CHAP 9 - FIGURE 5 Absorção N

N absorption

Componentes da Planta

Plant components

Folhas

Leaves

Raízes finas

Fine roots

Galhos

Twigs

Tronco

Trunk

Raízes grossas

Thick roots

N entrada

N intake

morte

death

Material da planta morto

Dead plant material

Estrutural

Structural

Metabólico

Metabolic

MOS

SOM

Ativo

Active

Lento

Slow

Passivo

Passive

Figure 5. DayCent simulator and its components. Source: Del Grosso et al. (2001).

In DayCent, the submodel for gaseous N simulates emissions of N2O, NOx and N2 originating from the soil and resulting from nitrification and denitrification processes (Figure 6). The simulator assumes that release of gaseous N from the soil due to nitrification is proportional to nitrification rates and that these rates are controlled by NH4+ concentration, water content, temperature, pH and texture. Nitrification is limited by water stress on microbial activity when the amount of pore space occupied by water (PSOW) is low, and by O2 availability when this amount is high. Nitrification peaks occur when soil water content is approximately 50% of PSOW. Nitrification is not limited when pH is greater than 7, but it decreases exponentially as pH drops below 7, due to acidity. The denitrification submodel first calculates the total flux of gaseous N from denitrification (N2 + N2O) and then uses a N2 / N2O ratio to estimate N2O and N2 emissions of. Denitrification is controlled by availability of labile C (electron donor), concentration of NO3− in the soil (electron acceptor) and O2 availability.

CHAP 9 - FIGURE 6 Nitrificação

Nitrification

solo

soil

Textura

Texture

Desnitrificação

Denitrification

Figure 6. Nitrogen gas flux in the DayCent submodel. Source: Del Grosso et al. (2001).

Application of DNDC, Century and CQESTR simulators to estimate greenhouse gas emissions Using DNDC to estimate methane emissions from flood-irrigated rice cultivation in a Gleysol in Pindamonhangaba, São Paulo (SP) Rice cultivation in a flood-irrigated planting system is a source of methane emissions (CH4). Anaerobic decomposition of organic matter in rice fields under this cultivation system produces methane that escapes into the atmosphere, initially by diffusion-controlled transport, through aerenchyma of rice plants during the growing season. These emissions also depend on climate factors, plant varieties used and types of soil where they are grown (where soil physical-chemical factors and management factors are also important), as well as on organic matter added, fertilizers applied and especially water management (ALBRITTON; MEIRA FILHO, 2001; BABU et al., 2005; EGGLESTON et al., 2007; LI; FENG, 2002; LI et al., 2004; LIMA et al., 2001; YAN et al., 2005). Therefore, local practices and climate aspects of the crop environment must be considered concurrently when evaluating factors that contribute to methane emissions during crop development under flooding conditions (LIMA et al., 1999, 2003, 2006). Embrapa Environment, under the Agrogases and Carboagro projects, conducted studies aimed at generating scenarios in the DeNitrification-DeComposition (DNDC) simulator to represent the irrigated rice crop management system in a cultivation area located in Pindamonhangaba, São Paulo (SP). This work was done in order to allow using simulation to evaluate trends in effects of management practices adopted in a continuous rice flooding regime on seasonal methane emissions, as well as variations in daily rates of seasonal methane emissions over various crop cycles. Version 8.9 of DNDC (of June 2006) was used with simulator input data including information gathered in the field combined with other data from laboratory tests on these experiments, conducted by Embrapa Environment in an area of the São Paulo State Agribusiness Technology Agency (APTA) [Agência Paulista de Tecnologia dos Agronegócios] / Paraíba Valley Regional Agribusiness Technology Development Hub (PRDTA Vale do Paraíba) [Pólo Regional de Desenvolvimento Tecnológico do Vale do Paraíba], in the 2003–2004, 2004–2005 and 2005–2006 harvests in Pindamonhangaba, São Paulo (SP). The most important data on these crops used for the purpose of the simulations conducted is presented in summary form in Table 2. Additional information used included local environmental factors measured daily (maximum and minimum temperature, rainfall, solar radiation), as well as crop management and irrigation water and physical-chemical characteristics of the soil (classified as a Gleysol with clayey to loamy-clayey texture). Daily climate data was provided by the APTA / PRDTA Vale do Paraíba Experimental Station and the organizing of information, as well as treatments and conversions of resulting values were performed using Microsoft Excel and SAS. Specific aspects of rice variety IAC-103, available in a publication on the cultivar by Agronomical Institute of Campinas / São Paulo Agency of Agribusiness Technology (IAC/APTA) [Instituto Agronômico de Campinas / Agência Paulista de Tecnologia dos Agronegócios], were

incorporated into the simulator database. This variety was chosen because it was one of the most widely used in the region during the period studied. For each harvest evaluated, a specific scenario was created in DNDC, considering the information presented in Table 2. At the end of the simulated period, the estimated seasonal methane emission was compared to that observed in the field. Table 2. Management and climate factors for harvests of flooded rice evaluated at Pindamonhangaba, São Paulo (SP). Harvest

Sowing and harvest dates

2004

Dec. 5 and Apr. 14

Fertilizer application (urea) Feb. 3 (30 kg)

Flooding period considered

Mar. 1 (30 kg)

Jan. 1 to Apr. 10

Mar. 14 (30 kg) 2005

Dec. 6 and May 9

2006

Dec. 17 and May 5

Jan. 10 (33 kg) Feb. 14 (33 kg)

Dec. 22 to Apr. 20

Jan. 27 (40 kg) Feb. 14 (40 kg)

Jan. 2 to Apr. 11

Mar. 8 (40 kg)

As a reference standard for generated emissions, both for quantified local information and information estimated by DNDC, the default emission factor provided by the Intergovernmental Panel on Climate Change (IPCC) was used for flooded rice systems under a continuous water regime and without addition of organic fertilizer, obtained by Yan et al. (2005) for the IPCC in 2006, of 1.3 kg CH4 ha−1 day−1 (0.80 – 2.20 error limit). The first considerations focusing on studying the DNDC software and its requirements in order to applied in this environment were presented by Plec et al. (2007) and later evolved to include more reliable field data, carefully entered into the simulator, This made it possible to present considerations focusing on using the DNDC simulator in flooded rice production systems in local mode (Pindamonhangaba, São Paulo (SP)), and on aspects related to validating it for the Brazilian environment. Analysis of results provided by DNDC also made it possible to identify aspects management practices which contributed to increase methane emission rates throughout the harvests analyzed. For the 2003–2004 harvest, DNDC estimated seasonal methane emissions of 20,560.87 mg CH4 m−2, whereas the measured value of gas collections made in the field for that same period was 8,923.02 mg CH4 m−2 ± 1,048.52 mg CH4 m−2. Considering the 101-day flooding period, it was found that the simulated value was about 2.04 kg CH4 ha−1 day−1, whereas the observed value was 0.88 kg CH4 ha−1 day−1 and, therefore, both are within the limit error range reported by Yan et al. (2005). It was found that, in this harvest, the simulator overestimated seasonal emissions by 130.42% compared to the quantified value. For the 2004–2005 harvest, the DNDC estimate was 18,765.22 mg CH4 m−2 whereas the measured value of gas collections made in the field for that same period was 18,906.13 mg CH4 m−2 ± 2,384.47 mg CH4 m−2. It was observed that the simulated value was about 1.56 kg CH4 ha−1 day−1 whereas the observed value was 1.58 kg CH4 ha−1 day−1 for that harvest, considering the flooding period of 120 days, both within the limits reported by Yan et al. (2005). In this harvest, DNDC underestimated the seasonal emission potential by 0.77%. Meanwhile, for the 2005–2006 season the seasonal gas emission estimate simulated by DNDC was 11,578.26 mg CH4 m−2, whereas the value of quantified samples collected in the field for the same period was 16,631.45 mg CH4 m−2 ± 2,328.51 mg CH4 m−2. Thus, considering the 100-day flood period of this harvest, it was observed that the simulated value was 1.16 kg CH4 ha−1 day−1,

whereas the observed value was 1.66 kg CH4 ha−1 day−1 for this harvest, and therefore, also both within the limits reported by Yan et al. (2005). In the case of this harvest (2005–2006), DNDC underestimated the seasonal emission potential by 30.38%. Generally, it was found that DNDC reflects seasonal emissions within expected international reference limits for irrigated rice cultivation, as it also presented them in the DNDC-to-quantified ratios of 2.30, 0.99 and 0.70, respectively for the harvests of 2003–2004 (year 2004), 2004–2005 (year 2005) and 2005–2006 (year 2006) (Figure 7). The DNDC-to-quantified ratio observed for 2004 can be justified as a result of necessary adjustments to the collection and transport method used for quantifying field gases, as the value of quantified emissions for that year was well below the values observed in measurements recorded for the following harvests.

CHAP 9 - FIGURE 7 Emissão de metano

Methane emission

Ano

Year

Quantificdo

Quantified

Figure 7. Comparison between estimates of seasonal methane emissions quantified (observed in the field) and estimated by DNDC for flooded rice harvests in Pindamonhangaba, São Paulo (SP). Source: Pessoa et al. (2010).

It was also found that the simulator constitutes a potential tool for identifying management strategies having fewer impacts in terms of seasonal methane emissions. There is also an intention to carry out sensitivity tests on DNDC, since there are variations in climate and in aspects related to carbon content, bulk density and soil nitrogen that should be further investigated in order to explain the variations observed. Thus, new quantification studies, being conducted under the Carboagro project, address a more detailed tracking of parameters potentially related to methane fluxes emitted during the crop growing season. They also contemplate activities aiming to assess the behavior of the DNDC simulator in other major areas of flooded rice production in Brazil, such as those found in Santa Catarina (SC), which will be critical to present increasingly reliable results. These results will also contribute to preparing estimates of greenhouse gas emission factors as part of reference reports for nationwide greenhouse gas emission inventories.

Use of Century and CQESTR to estimate carbon emissions in Latosols [US: Oxisols] under conventional tillage and no-tillage with soybean and maize cultivations in Viçosa, Minas Gerais (MG), and Baixa Grande do Ribeiro, Piauí (PI) In Brazil, excessive soil tillage has favored emergence of erosion and compaction processes, and medium- and long-term physical, chemical and biological degradation of the soil. It has also increased biological oxidation of organic carbon into CO2, through increased microbial activity due to increased soil aeration, and increased contact between soil and plant residue. This has resulted in more favorable conditions for decomposition and increased exposure of carbon protected in soil aggregates to microbial attack, causing decreased organic matter stocks and, consequently, increased CO2 concentration in the atmosphere. These factors, in isolation or in combination, have

contributed to decrease crop productivity and create an unbalanced environment. Therefore, in recent years, adoption of no-tillage has been encouraged, as it is premised on sustainability of the production process, maintaining or restoring organic matter stocks by avoiding soil movements and incorporating residues, which ensures a lower decomposition rate (LEITE et al., 2003; ROSCOE, 2006). Despite these benefits, studies using simulators have shown that no-tillage, when not associated with other practices, such as inclusion of cover crops with significant biomass yield, may, in the long run, fail to increase C stocks. In an Red-Yellow Argisol [Argisol = US: Ultisol], in Viçosa, Minas Gerais (MG), the Century simulator estimated the dynamics of total organic carbon (TOC) and its pools, since the cutting of the Atlantic Forest in 1930, and subsequent adoption of conventional agriculture, up until the experiment period (1984–2000), with application of treatments (no-tillage, disc plow, heavy harrow + disc plow and heavy harrow), and extending to the year 2050 (LEITE et al., 2004). In all systems, including no-tillage, cultivated with a maize – beans succession, there was a decrease in simulated C stocks, which indicated the need for changes in management strategies, such as inclusion of cover crops with high inputs of residues (Figure 8). In the same study, the CQESTR simulator also estimated, for all systems, decreases in TOC stocks (LEITE et al., 2009) (Figure 9A).

CHAP 9 - FIGURE 8 COT

TOC

C ativo

Active C

C Lento

Slow C

C passivo

Passive C

Ano

Year

PD

NT

AD

DP

GPAD

HH+DP

GP

HH

Período anterior ao experimento

Period before the experiment

Início do experimento

Beginning of experiment

Figure 8. Century simulation of the dynamics of total organic carbon (TOC) and carbon pools (C) in a Red-Yellow Argisol [Argisol = US: Ultisol], in Viçosa, Minas Gerais (MG). NT: no-tillage; DP: disc plow; HH+DP: heavy harrow + disc plow; HH: heavy harrow. Source: Leite et al. (2004).

In Baixa Grande do Ribeiro, in the Cerrado of Piauí (PI), in a Red-Yellow Latosol [Latosol = US: Oxisols; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox] cultivated with soybean – maize, the CQESTR simulator also evaluated the impact of replacing the native Cerrado forest (1988) with a conventional tillage system (1988–1996) and with reduced tillage and no-tillage systems (1996– 2033). CQESTR estimated a decrease in TOC stocks after conversion of native forests into agroecosystems, and this reduction was maintained, even with the subsequent adoption of no-

tillage. The amount of residues added via cover crops (second-crop corn, 4 Mg ha−1) was not sufficient to produce increases in TOC stocks (LEITE et al., 2009) (Figure 9B).

CHAP 9 - FIGURE 9 COT

TOC

Ano

Year

PD

NT

PC

CT

PR

RT

PR1

RT1

PR2

RT2

Início do experimento

Beginning of experiment

Figure 9. CQESTR simulation of the dynamics of total organic carbon (TOC) in a Red-Yellow Argisol [Argisol = US: Ultisol] (Viçosa, Minas Gerais (MG)) (A); and a Red-Yellow Latosol [Latosol = US: Oxisol; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox] (Baixa Grande do Ribeiro, Piauí (PI)) (B). NT: no-tillage; RT1: reduced tillage with disc plow; RT2: reduced tillage with heavy harrow; CT: conventional tillage with disc plow and heavy harrow; RT: reduced tillage with heavy harrow. Source: Leite et al. (2009).

Table 3. Carbon (C-CO2) emission estimated by the CQESTR simulator and from values measured in a RedYellow Argisol [Argisol = US: Ultisol] (AVA) in Viçosa, Minas Gerais (MG) and a Red-Yellow Latosol (LVA) [Latosol = US: Oxisol; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox] (LVA) in Baixa Grande do Ribeiro, Piauí (PI), under various tillage systems. C stock Rate(1) Tillage system

Initial

Final

Sequestration(+) Emission (-) of C

Variation ∆

Mg ha−1

Mg ha−1 year−1 AVA (Viçosa)

CQESTR NT

43.4

34.79

−8.60

−0.53

−0.36

RT1

43.4

28.45

−14.94

−0.93

−0.96

CT

43.4

27.18

−16.21

−1.01

−1.05

RT2

43.4

28.80

−14.59

−0.91

−0.94

NT

44.0

38.54

−5.46

−0.34

−0.17

RT1

44.0

31.23

−12.77

−0.79

−0.82

CT

44.0

30.90

−13.10

−0.82

−0.86

RT2

44.0

31.24

−12.76

−0.79

−0.82

Measured

LVA (Baixa Grande do Ribeiro) CQESTR NT

42.7

36.07

−6.63

−0.51

−0.30

CT

42.7

32.53

−10.17

−0.78

−0.82

RT

42.7

33.75

−8.95

−0.69

−0.72

NT

45.1

40.68

−4.42

−0.34

−0.14

CT

45.1

34.72

−10.38

−0.80

−0.84

RT

45.1

35.76

−9.34

−0.72

−0.75

Measured

Note: NT: no-tillage; RT1: reduced tillage with plow; CT: conventional tillage; RT2: reduced tillage with harrow; RT: reduced tillage with harrow. (1)

A contribution of 6 Mg ha−1 from cover crop (0.17 Mg ha−1 year−1 and 0.20 Mg ha−1 year−1 for AVA and LVA, respectively, considering 45% as C) (SALTON et al., 2005) was assumed for the no-tillage system. C emissions (0.045 Mg ha−1 year−1 and 0.031 Mg ha−1 year−1, for conventional tillage and reduced tillage, respectively) were included to represent additional fuel use (DIECKOW et al., 2004). Source: Leite et al. (2009).

For the studies in Viçosa and Baixa Grande do Ribeiro, CQESTR estimated the emission of C (C-CO2) into the atmosphere. In Viçosa, emissions were 0.36 Mg C ha−1 year−1 and 1.05 Mg C ha−1 year−1 under no-tillage and conventional tillage, respectively. The estimate using measured values also indicated C emission for all systems, although of smaller magnitude (LEITE et al., 2009) (Table 3). These estimates differ from those reported in most studies where there is C sequestration under no-tillage. The difference can be attributed to the method used to calculate the estimates. In this study, C stocks at the beginning of the experiment, i.e., at time zero, were used as reference in order to verify the real contribution of no-tillage, whereas in other studies the reference values were those observed in conventional tillage. If that had been the case, there would be sequestration of C in the no-tillage system, on the order of 0.47 Mg ha−1 year−1 (simulated) and 0.48 Mg ha−1 year−1 (measured), which would be consistent with the values observed by other authors for the same type of soil (0.52 Mg ha−1 year−1) (LOVATO et al., 2004). In Baixa Grande do Ribeiro, emissions of C were also observed. The values ranged from 0.30 Mg ha−1 year−1 (no-tillage) to 0.82 Mg ha−1 year−1 (conventional tillage) estimated by CQESTR, and from 0.14 Mg ha−1 year−1 (NT) to

0.84 Mg ha−1 year−1 (CT) calculated using measured values (Table 3). However, taking TOC stocks in conventional tillage as reference, the no-tillage system presented a C sequestration rate of 0.18 Mg ha−1 year−1 to 0.38 Mg ha−1 year−1 for measured and simulated values, respectively.

Comparison of simulation of carbon dynamics by Century in CQESTR in Latosols [US: Oxisols] under conventional tillage and no-tillage In an analysis of CQESTR and Century simulators, Leite and Doraiswamy (2007) observed, in a Red-Yellow Latosol (LVA) [Latosol = US: Oxisol; Red-Yellow Latosol = US: Rhodic/Xanthic Haplustox] in Baixa Grande do Ribeiro, Piauí (PI), that Century presented a better correlation (R2 = 0.96) between simulated and measured values than CQESTR (R2 = 0.88) (Figure 10), which can be explained by its more robust structure, with several residue pools (surface and soil, subdivided into structural and metabolic) and soil organic matter pools (active, slow and passive) with different residence times. Decomposition rates for these pools are obtained through several multiplicative functions of soil temperature and moisture, lignin-to-nitrogen ratio and clay content, which allows for greater control of processes involved in carbon dynamics. In the CQESTR simulator, there is a lesser number of residue pools (surface and soil) and soil organic matter is unicompartmental, suggesting a greater simplification of the model’s equations and assumptions. In addition, the texture factor is limited to a texture class descriptor (clay = −1 and sand = 1). Soils with the same texture class descriptor, however, may have different clay contents, and, as a result, will have different cation exchange capacities, which will influence carbon stabilization differently. Therefore, this is a limitation to the model. Despite the absence of important mechanisms, however, differences in TOC stocks simulated by CQESTR in relation to those simulated by Century and to those measured by analytical techniques were of low magnitude, in all tillage systems. In the no-tillage system, differences between simulated and measured were, on average, 2.9% for CQESTR and −10.5% for Century. In remaining systems, differences were 1.9% and −4.7% (for conventional tillage) and 3.9% and 5.3% for (for reduced tillage) for Century and CQESTR, respectively (Figure 11). CQESTR’s greater simplicity and differences between simulated and measured TOC stocks (

Carbon stocks and greenhouse gas emissions in ...

Carbon stocks and greenhouse gas emissions in ...

Suggest Documents

Quantifying soil carbon stocks and greenhouse gas fluxes in the ...

Quantifying soil carbon stocks and greenhouse gas fluxes in ... - SciELO

Soil greenhouse gas emissions and carbon budgeting in ... - CiteSeerX

Greenhouse Gas Emissions Inventory

Pricing Greenhouse Gas Emissions

Greenhouse Gas Emissions Report

Carbon stock, greenhouse gas emissions and water balance ... - Helda

Review of carbon flux estimates and other greenhouse gas emissions ...

Greenhouse gas emissions limited by low nitrogen and carbon ... - PeerJ

Greenhouse gas emissions and carbon sequestration by ... - Locus UFV

2016 UK greenhouse gas emissions

Comparing Greenhouse Gas Emissions across

Greenhouse gas emissions regulations - InfluenceMap

Greenhouse gas emissions regulations - ShareAction

greenhouse gas emissions from cattle

2016 UK greenhouse gas emissions

Estimated reduction in greenhouse gas emissions ...

Estimating Greenhouse Gas Emissions in Agriculture

Carbon soundings: greenhouse gas emissions of the UK music industry

Greenhouse Gas Emissions Reductions in France - ClimateXChange

Reducing Greenhouse Gas Emissions in the ... - WineEcoReports

Consistency in Greenhouse Gas Emissions ...

Greenhouse Gas Emissions in Montgomery County, Pennsylvania ...

Estimating Greenhouse Gas Emissions in Agriculture