CLIMATE VARIABILITY AND LAND-USE CHANGE IN ... - Springer Link

CLIMATE VARIABILITY AND LAND-USE CHANGE IN DANANGOU WATERSHED, CHINA—EXAMPLES OF SMALL-SCALE FARMERS’ ADAPTATION JOHANNA HAGEBACK1 , JENNY SUNDBERG1 , MADELENE OSTWALD2 , DELIANG CHEN1,3 , XIE YUN4 and PER KNUTSSON5 1 Department of Earth Sciences, Göteborg University, Göteborg, Sweden Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, 54-1411, 77 Mass Ave, MA 02139, U.S.A. E-mail: [email protected] 3 Laboratory for Climate Studies at National Climate Center, China Meteorological Administration, Beijing, China 4 Department of Resource and Environment Sciences, Beijing Normal University, Beijing, China 5 Section of Human Ecology, Göteborg University, Göteborg, Sweden

2

Abstract. With global concern on climate change impacts, developing countries are given special attention due their susceptibility. In this paper, change and variability in climate, land use and farmers’ perception, adaptation and response to change are examined in Danangou watershed in the Chinese Loess Plateau. The first focus is to look at how climate data recorded at meteorological stations recently have evolved, and how farmers perceived these changes. Further, we want to see how the farmers respond and adapt to climate variability and what the resulting impact on land use is. Finally, other factors causing change in land use are considered. Local precipitation and temperature instrumental data and interview data from farmers were used. The instrumental data shows that the climate is getting warmer and drier, the latter despite large interannual variability. The trend is seen on the local and regional level. Farmers’ perception of climatic variability corresponds well with the data record. During the last 20 years, the farmers have become less dependent on agriculture by adopting a more diversified livelihood. This adaptation makes them less vulnerable to climate variability. It was found that government policies and reforms had a stronger influence on land use than climate variability. Small-scale farmers should therefore be considered as adaptive to changing situations, planned and non-consciously planned.

1. Introduction A shortage of arable land and a water deficit makes the agriculture in China very dependent on climate and susceptible to climate change (Smit and Cai, 1996). Sixty-four percent of the population of over one billion is rural and 85% of those (686 million) make their living from farming according to census data presented in 2000 (National Bureau of Statistics of China, 2000). The impact of climate change on agriculture has been part of the national key project (1996–2000) “Studies on short-term climate prediction system in China”. Included in this project were studies of strategies for reducing the impact of climate disasters on agriculture (National Climate Centre, 2000). The government in China is realizing the importance of a sustainable agriculture. Integrated control of soil erosion and development of Climatic Change (2005) 72: 189–212 DOI: 10.1007/s10584-005-5384-7

c Springer 2005

190

JOHANNA HAGEBACK ET AL.

sustainable agriculture have been major projects. Over the past 20 years, special attention has been given to the Loess Plateau, with its deep loess and severe soil erosion (Liu, 1999). This region is very important to the agriculture in China, since the soil provides favorable conditions for plant growth and high productivity. The potential to adapt or mitigate the impact of climate change on agriculture has mainly been studied in highly intensive agricultural areas or in large scale, such as the Asia-Pacific Region, Europe and the United States where model simulations are commonly used (Luo and Lin, 1999; Polsky and Easterling, 2001; Olesen and Bindi, 2002). Such large-scale models are inadequate for use on a local level (Reilly and Schimmelpfennig, 1999), and hence more site-specific studies are needed. This study was conducted in the Ansai County where a soil and water conservation station exists. This is one of the 11 case-study sites that scientists have used in a project sponsored by the Chinese government (Liu, 1999). Earlier research in China, where dynamic crop models and statistical models were used, showed that the Loess Plateau would be negatively affected by climate change. Shaanxi is one of seven provinces with high vulnerability and less capability to adapt to climate change (Luo and Lin, 1999). The EROCHINA project focused on the selected Danangou watershed study area (Erochina, 2000/2001). The aim of the project was to develop alternative land use and soil water conservation strategies to increase sustainability by reducing soil and water losses. The farmers in Danangou watershed have been interviewed several times with a focus on land use and various socio-economic aspects (Carlsson, 2000; Chen et al., 2001; Rosmuller, 1999). The relationship between land use and soil conditions in Danangou watershed has been studied (Fu et al., 2000; Qiu et al., 2001; Wang et al., 2001; Fu et al., 2003; Wu et al., 2003). However, no research has focused on climate variability and its relation to the land use in Danangou watershed. There are many uncertainties regarding how climate change and variability will impact agriculture, and how farmers will respond (Reilly and Schimmelpfennig, 1999). A better understanding of the social and economic factors influencing farmers’ perception and their response to a changing climate is needed. Johnston and Chiotti (2000) stress the importance of studying adaptation at the individual farm level. Local level studies of farmers’ perception, response and adaptation to climate variability has been carried out (Ovuka and Lindqvist, 2000; Roncoli et al., 2001; Vedwan and Rhoades, 2001). Many changes can act simultaneously making it hard to distinguish the direct link between change, response and adaptation. Reilly and Schimmelpfennig (2000) define adaptation as a response that improves the outcome, i.e. to increase the level of ones livelihood, which can take place autonomously (without conscious planning) or non-autonomously (planned). This description is used in this paper with the addition that there are often uncertainties and risks involved in the action taken and that the adaptation can be a response to multiple, simultaneous changes within a system. Economic considerations are an important aspect of adaptation (Olesen and Bindi, 2002). Scoones (1998) has developed an analytical tool for rural livelihood where changes play an important

CLIMATE VARIABILITY AND LAND-USE CHANGE

191

role. When changes within a system occur, they can have different effects. They can have restricting or opportune effects, i.e. some will decrease the farmers’ possibilities while others will augment their possibilities. Changes also happen in a different fashion. Shocks occur infrequently and can have large direct impact. Small and regular changes may be predictable and their effect can be cumulative. As a consequence of the above presented relations, the driving hypothesis for this study is that small-scale farmers in a developing country are adapting to climate change. To investigate this hypothesis, several questions are explored. First, what are the local and regional precipitation and temperature trends and variability, and how is climate variability perceived by farmers? Second, what are possible responses and hence adaptations to climate variability, and is there any impact on their farming? Third, are there other changes causing farmers to change their land use? There is a fundamental difference between climate change and climate variability. Climate change constitutes a shift in meteorological conditions that last for a long period of time usually centuries. Climate variability is short-term fluctuations happening from year to year (Burroughs, 2001). In this work both definitions are used: climate change as the conceptual global change that is taking place and climate variability as the process taking place in Ansai with the time-frame considered in this study. The underlying conception is that global climate changes cause local scale impacts and that it is not possible to treat variability separate from climate change (Mearns et al., 1997; Smit et al., 1999, 2000). This region of China is no exception (IPCC, 2001). The data used and the time-frame considered here, however, can only account for a change over a few decades, and could therefore well be variability or a phase of an oscillation (Wang, 2002, personal communication). The focus of this paper is the adaptation of small-scale farmers to change, so it is not important if the change is caused by greenhouse-gas induced global warming or local natural variability. Land use is defined as the way in which humans are using the land, which is an effect of an integrated set of biophysical and socio-economical factors (Verburg and Chen, 2000). Land-use change takes place when the use is altered, partly or totally. The change can be both caused and effected by environmental or socio-economic factors (Krausmann et al., 2003). Land use is, in this case, not totally limited to agricultural land use, even though the area is agricultural. Forestry and grassland are land uses that have become central in this area as well.

2. Study Site China has a complex and diverse topography. The land elevation decreases from west to east (Hsieh, 1973). The 105◦ E longitude boundary between step one and two is a significant climatic dividing line (Domrös and Peng, 1988). It divides the monsoon and non-monsoon zones. The extremely varied landforms have noticeable differences in temperature and precipitation (Domrös and Peng, 1988).

192


In the middle of July, the summer monsoon advances to northern China. In the middle of August, the summer monsoon starts to weaken and withdraw. Thus the summer monsoon in northern China lasts for a month and during this period the region receives most of its annual rainfall. A weak summer monsoon might not reach northern China or might have a shorter duration than normal, which leads to droughts (Ding, 1994). The Danangou watershed (36◦ 53 N; 109◦ 19 E) is located on the middle part of the Loess Plateau in Ansai County in northern Shaanxi Province (Figure 1). Ansai town, the seat of the county, is situated 7 km south of Danangou. Two villages, Danangou and Leipingta, can be found in the watershed, which covers an area of 3.5 km2 . The total population is around 215 people, distributed among 48 households. The main valley of the watershed, with an altitude ranging between 1000 and 1350 m.a.s.l., consists of four valleys where several smaller valleys end. Serious erosion in this area is evident though the ground surface has been incised strongly, developing rills and gullies. Even so, according to the farmers, the living

Figure 1. (a) Location of Shaanxi Province and the Loess Plateau in China. (b) Shaanxi Province, Loess area, Ansai, and location of the six climate stations used for regional precipitation analysis. (c) A topographic map of the study area—Danangou watershed (modified from Chen et al., 2001).


193

standard is better today than in the past. A general statement among the farmers was, “It is getting better and better” (Wang et al., 2001; farmers in Danangou, 2002, personal communication). The climate is semi-arid with an average annual temperature of 8.8 ◦ C and a monthly mean ranging from 22 ◦ C in July to −7 ◦ C in January. The average annual precipitation is 562 mm (max: 645 mm in 1978, and min: 297 mm in 1974) (Wang et al., 2001). The precipitation distribution during a normal year is strongly influenced by the monsoon (Figure 2). The wind-deposited loess soil predominates and is classified as Calcic Cambisol (FAO-UNESCO, 1997). This soil is weakly resistant to erosion (Fu et al., 2003). The land use in the area has been classified as slope farmland, terrace land, orchard, bush land, woodland, traffic land, residential land and sparse wild grassland according to Chen et al. (2001). The crops cultivated in Danangou watershed are: millet, maize, sorghum, potato and buckwheat. Only a small fragment of natural vegetation exists in the area. Several land-use changes have taken place in this watershed in the recent past. The history of the study area can be divided into several periods. 1. Before the establishment of the Peoples Republic of China in 1949: Only a few people lived in the area and landlords, from whom the farmers had to rent, owned most of the land. Large areas were used as farmland because of the low yields.

Figure 2. Yearly distribution of precipitation in Ansai (daily precipitation data from Ansai weather station 1970–2001).

194


2. 1949–1982: The first land reform took place in 1956, when a commune was established. Individual farmer’s land was collected as common land and the community managed and used the land according to the state policy. The change in land use was characterized by conversion from small scale farming to large scale farming. 3. 1982–1999: The second land reform, Household Responsibility System, was introduced in Danangou during 1982. The central government decided to resolve the communes, and the land-use rights were again distributed to individual farmers that could choose what to plant, many times based on market demands. Land-use change was characterized by conversion from large scale farming back to small scale family based farming. Two minor reallocations of land ownership occurred 1992 and 1995 due to the emigration of people (Chen et al., 2001; farmers in Danangou, 2002, personal communication). 4. 1999: As part of the efforts to halt the massive soil erosion from the Loess Plateau, the Cropland Conversion Program was introduced by the Chinese government in Shaanxi Province in 1999. The goal is to minimize the cultivation of crops in slope areas by encouraging farmers to instead plant trees and grasses, causing great land-use change. The farmers in Danangou have started to plant trees on a large scale with food and money as compensation incentives (Rui et al., 2001). Another reform that has affected all sectors in China is the economic reform, Open-Door Policy, adopted in 1978, after which rapid economic development and an increasing globalization has occured (Yeh and Li, 1999). The reform was focused on opening up to the outside market-world under controlled forms, which was done through establish special economic zones. Even though the bulk of this reform had its impact along the coast, its effect has been seen in Shaanxi as well with a gradual change towards a market economy, e.g., through increased imports of non-provincial food, and development of a transportation system leading to accessible off-farm incomes. The force behind the policy was to end years of economic stagnation (Lu and Wang, 2002). Danangou watershed belongs to Zhenwudong Township in Ansai County. This local government covers more than 100 villages, each with their own president. The farmers in the village select the president. The village president is only in charge of village business, such as assisting the local government in collecting taxes. The agriculture tax is dependent on how much land (crops and trees) the farmer has. This tax is paid in money. The farmers also have to sell a specified amount of their harvest to the local government, which pays them less than the market price. The amount is different from year to year. There is no insurance in case of a bad harvest. In bad years the local government has given the farmers some food as compensation (farmers in Danangou, 2002, personal communication).


195

3. Data and Material 3.1.

CLIMATE DATA

Daily precipitation and temperature data was obtained from the Ansai weather station, Shaanxi. The data covers the period from January 1970 to December 2001. The local temperature data contained only a few errors, which were manually corrected by interpolation. A total of 9.3% of the local rainfall data were missing. Data gaps occurred over the entire period, but data were more frequently absent during the summer. The missing data were interpolated linearly. Monthly and yearly means of the daily precipitation and temperature were calculated and trends were investigated. This was done both yearly and seasonally. Winter was defined as December–February, spring as March–May, summer as June– August and fall as September–November (Domrös and Peng, 1988). The regional rainfall data were received from National Climate Center. Six rainfall stations in the region within longitude 106–112◦ E and latitude 33–40◦ N, around Ansai, Shaanxi were chosen (Figure 1). The received data were monthly precipitation during the time period from January 1951 to December 1999. A mean of all six stations was calculated. The same analysis was conducted as for the local rainfall. 3.2.

INTERVIEW DATA

Depending on the scope of the study, the interviewer determines the degree of control of the interview and answers, which can be categorized into four methods (Bernard, 1995). – Informal interviewing lacks any kind of control or structure. It is usually used in the beginning of a field study when the researcher is getting familiar with the surroundings and local population. – Unstructured interviewing is based on a plan with a central purpose, but is kept open for changes and gives the respondents the opportunity to freely answer the questions at their own pace and understanding. – Semi-structured interviewing is more controlled and well prepared since it is based on an interview guide. The interview guide is a set of clear instructions, but the interviewer remains flexible to explore new information. These three methods were used in the beginning of the field work to become familiar with the area by letting the respondents describe issues of relevance for his/her life, such as history, agriculture, climate calendar, education, land use and land owning system. Several visits were made to each respondent. – Structured interviewing is the most controlled method. The respondents are given the exact same questions, e.g., a questionnaire. The idea is that by controlling the information that is given the answers can be systematically

196


compared. For this work, a questionnaire was constructed together with the translator and tested for corrections and rephrasing. During the initial part of the fieldwork, the farmers had difficulties remembering the climate in the past, so the questions comparing today and the past were related to a well known event, i.e. the introduction of the Household Responsibility System in the beginning of the 1980s. Due to the included time-frame, respondents had to be over 30 years old. Further, since women usually move at marriage, the majority of respondents were men. A total of 38 out of 48 farmers participated, 27 men and 11 women were interviewed. 10 households did not participate due to lack of requirement or unavailability. The approximately 40 questions asked focused on climate, climate change and disasters. In-depth interviews with a few informants can be very informative. Berg (1998) defines an informant as a person who has experience and knowledge in the particular subject. A total of 10 unstructured and semi-structured interviews were conducted with several key informants. Topics were history of the village, agriculture and climate calendar, local government system, education, subsidies and land owning system. Informal interviews were conducted frequently throughout the study. Focus group discussions consist of a selected group of people that discuss a particular topic. A group discussion creates interaction between people and leads to verbally expressed thoughts and opinions about the topic (Bernard, 1995; Berg, 1998). Six group discussions, three mens groups and three womens groups, complemented the interviews and questionnaires. An interactive tool, called Climate Game, was invented in the field and used during the group discussions. It consisted of a diagram/table where the x-axis was divided into five, 10-year periods (1962–2012), and the y-axis was divided into four groups: temperature, rain, wind and snow. These four groups where symbolized by drawings along the y-axis. The same drawing was also made on four small separate cards. All interviews were conducted together with an Chinese native interpreter that took active part in the development of questions and phrasing of words and concepts. For more information on the interview methods see Hageback and Sundberg (2002). 4. Results The primary hypothesis is that small-scale farmers adapt to climate change and variability. Thus it first had to be determined if the climate has changed and whether the farmers perceive climate variability. The local precipitation and temperature from the Ansai rainfall analysis shows that 58% of the precipitation falls during summer, i.e. when the monsoon arrives (Table I). The start of the monsoon rain varies from June to September. A decreasing tendency of the rainfall occurs during the period in all seasons except spring. The trend is not significant, since there is a large variability in the amount of precipitation

197


TABLE I Local precipitation and temperature from 1970 to 2001. Based on data from Ansai weather station Yearly

Spring

Summer

Fall

Winter

129 −2.6 0.2 −81.9

92.1 16 44.7 0.3 0.1 9.9

333 58 105 −1.1 0.1 −34.6

139 24 66.0 −1.5 0.2 −47.0

15.9 2.8 13.6 −0.4 0.3 −12.4

−14

11

−10

−34

−78

Precipitation Mean (mm) Percentage of yearly total Standard (mm) Trend (mm/year or season) Correlation Total change calculated from the trend1 (mm/32 years) Total change calculated from the trend (%)

579

Temperature Mean (◦ C) Standard (◦ C) Trend (◦ C/year or season) Correlation Total change calculated from the trend1 (◦ C/32 years)

8.97 0.60 0.03 0.48∗∗ 0.96

10.51 0.87 0.02 0.23 0.64

21.49 0.71 0.01 0.21 0.32

8.71 0.77 0.03 0.39∗ 0.96

−4.81 1.22 0.06 0.47∗∗ 1.92

∗

P < 0.05, ∗∗P < 0.01 Student’s t-test. N = 32. 1 Difference between the trend line value between the first and last year. TABLE II Regional rainfall from 1951 to 1999 (six stations)

Mean (mm) Percentage of yearly total Standard (mm) Trend (mm/year or season) Correlation Total change according to the trend (mm/49 years) Total change according to the trend (%)

Yearly

Spring

Summer

Fall

Winter

457 78.2 −1.8 0.3∗ −86.2

87.6 19 32.0 −0.0 0.0 −1.5

240 53 55.9 −0.7 0.2 −33.8

117 26 43.1 −0.9 0.3∗ −45.6

12.5 2.7 7.7 −0.1 0.1 −2.9

−19

−2

−14

−39

−23

∗

P < 0.05 Student’s t-test. N = 49.

from year to year. The temperature data in Ansai shows an increasing trend, with the increase mostly during winter. In 32 years temperature has risen 1 ◦ C. The regional rainfall analysis shows that more than half of the precipitation in the area falls during the summer months (June–August) (Table II). All seasons show a decreasing trend. The trend is most significant in the fall and during the entire

198


Figure 3. Comparison of local (Ansai) and regional (Shaanxi) precipitation based on mean yearly precipitation.

year, while the other seasons do not show significant trends. The correlation is not strong because of the large variability from year to year. A comparison of local and regional rainfall data indicates that it rains more in Ansai than in the surrounding region (Figure 3). The rainfall trend in Ansai is decreasing except in spring, when a weak increasing trend occurs (Table I). The regional rainfall data indicates a decrease in all seasons. It is well established that rainfall is a local phenomenon with large variability on a regional scale. One approach to describing climate variability is by noting changes in extreme events. The main disasters in this region are drought, flood, hail and frost (indepth interview), including both short-term weather events (e.g., hail) and extreme seasonal conditions (e.g., droughts). During the structured interviews, the farmers were asked to choose which disaster has occurred most frequently and which has caused the most damage to their agriculture 20 years ago and in recent years. Figure 4a displays a large distribution among the most frequent disaster in the 1980s, while Figure 4b shows that 82% of the farmers agree that drought has occurred most frequently in recent years. It is noteworthy that even though hail occurs most frequently in the 1980s, drought is still the disaster that causes the most damage to their agriculture. “Do you feel any changes in the weather now compared to 20 years ago?” was another question. Thirty-four percent thought it has become warmer, 32% thought it has become drier and 24% believed it has become both warmer and drier. Eight percent of the farmers thought that no change had occurred, and the last 2% mentioned seasonal temperature fluctuations. According to the farmers, the temperature change is most obvious during the winter. One farmer noticed


199

Figure 4. Most frequent disasters and disasters causing the most damage according to the asked farmers (%) in (a) 1980s and (b) recent years. 1980s refers to the period after the introduction of the Household Responsibility System in the area (1982).

Figure 5. Farmers’ perception of climate trends from 1962 to 2002 based on the climate game tool with a male group in Leipingta. Photo: Sundberg, 2002.

200


the temperature change during winter as follows: “Before the water was frozen in the bucket that was used to carry water in” (female, 59 years old). Many farmers explained that they used to dress in thick woolen clothes when they were young, but nowadays they only wear thin cotton clothes. The Climate Game that was used in group discussion presented the same results. Figure 5 represents the farmers’ general perception of climate change during this period. They all agree that the temperature has increased continuously during the period while the other factors fluctuate more. “Now the weather is more uncertain and varies more” (male, 70 years old). The second area to examine is possible response and hence adaptation to climate variability that farmers have in relation to land use. Reilly and Schimmelpfennig (2000) describe adaptation as response that improves the outcome. This implies that a change has taken place or is anticipated to take place in a system (IPCC, 2001), e.g., climate system, to which people respond. In China one impact of the Open-Door Policy is the introduction of more frequently used televisions. Farmers can get information from TV about weather predictions and respond. Ninety percent of the farmers in Danangou have TVs. Seventy-nine percent of the farmers get their weather information from TV and 81% of them trust it. All of them say that they use this information. For example, if it is to be cold, they will not open the plastic that covers the crops (see below). If it is supposed to rain, they could cover the vegetables, and if there is supposed to be a rainstorm, 11% said that they could dig a channel that drains the water around the house. Thirty-seven percent of the farmers arranged their work and their planting time depending on the forecast. During the group discussions in Danangou, the farmers were given some scenarios. “What would you do if the weather forecast on TV said that there would be no rain next spring?” The women answered that they would plant crops resistant to drought, while the men would choose a type of crop depending on when the rain comes. “What would you do if there was continuous rain for half a month?” They said that they have no method to protect their crops. A survey of the impact of climate variability on land use leads to several observations. Figure 6 shows a summary of technical improvements and reforms that can be seen as initiators for a land-use change, covering a period from 1970 to 2002. These events are shown in relation to yearly precipitation. The events described have been the most important factors for farm-development according to the farmers themselves. (a) In the 1970s, a new method of planting was introduced. During three years in the 1970s, intercropping was used. Except in 1974, this decade had more rain compared to the previous few years, according to the climate data. (b) In 1982, the main year of the Household Responsibility System in the Danagou area, 18% of the farmers started with off-farm work and 26% with livestock. In 1982–1983, they also started to use the plastic cover on their


201

Figure 6. Precipitation in relation to technical improvements and reforms causing land-use change in Ansai during the last 30 years.

seedlings (used by all farmers today). The years around the period were not unusually dry, hence can not explain the land-use change. (c) In the 1990s, another kind of plastic cover technique was introduced (used by all farmers today), at the same time as a decreasing rainfall trend. According to the data, the last part of the decade was especially dry. (d) During the last years (1997–2002), more than half, 58%, of the farmers changed their main crop: 13 out of 15 livestock keepers stopped raising livestock and more than 40% started with off-farm work. In 2001, 8% started to raise livestock. In 1999, the Cropland Conversion Program was implemented in this area, which also impacted the land use. This period has been very warm and dry, which may be the reason for more off-farm work and the change in main crop. During a large part of this period, relaxation on the household registration system (hukou) and the work unit system (denwei) affected the mobility of particularly the agrarian population in China and hence their possibility to make off-farm income, which in this context is worth mentioning (Chan and Zhang, 1999; Lü and Perry, 1997). As a response to the climate variability and hence adaptation, Table III displays the six main crops in this area. According to research assistants at Ansai Research Station (Hou, 2002, personal communication), these crops are chosen because they are resistant to droughts. Among these crops, buckwheat and corn are most dependent on the rainfall, while soybean and millet are least sensitive. Crops’

202


TABLE III Ranking of the six main crops depending on drought resistance Ranking

Prof. Li

Respondent 42

Respondent 6

1 2 3 4 5 6

Buckwheat Potatoes Corn Soybean Pearl millet Foxtail millet

Soybean Buckwheat, potatoes and sorghum

Corn Potatoes Buckwheat Pearl millet Foxtail millet

Pearl millet Foxtail millet

Ranking of 1 is most rainfall dependant and 6 is least dependant.

dependency on rain was also discussed with scholars at the Institute of Soil and Water Conservation (Li, 2002, personal communication) and two farmers (indepth interviews). They all agree that the timing of the rainfall is more important than the amount of rain. They were also asked to rank the crops according to how dependant crops are on the timing of the rainfall (Table III). The scholar and farmers all agree that millet is least sensitive, while they differed among them on the most rain-dependent crop. The relation between crops and climate was also discussed during the group discussions, where one farmer said: “Our agriculture is based on the climate.” As an adaptive response to climate variability, the men say that they now plant earlier because of warmer weather. They also mentioned millet’s resistance to drought and explained that it used to be their main food until the 1990s, when better transportation in China made rice accessible to the food market in this region. Now they mainly plant vegetables. During the structured interviews, the farmers were asked about their main crop (covering the largest area) now (2002) and 20 years ago. A transition from planting mainly millet to planting different kinds of crops was identified (Figure 7). The farmers were also asked to explain why they had changed their practice. The answers were later categorized into five groups (Figure 8). Economical values, including better price and higher yield, represented 63% (N = 30) of the changes. The farmers have changed to a variety of cash crops, made possible by the introduction of the Household Responsibility System. As Table III shows, these cash crops are more sensitive to rainfall variability. If looking at potatoes and millet, which are the main crops involved in the change (Figure 7), the local market price for potatoes and millet is 0.6 and 2 yuan per kg, respectively, according to farmers in 2002. If looking at the yield statistics of Shaanxi Province (China Statistical Yearbook, 1996–2002), the productivity of millet has decreased drastically from 1094 kg/ha in 1983 to 510 kg/ha in 2002 with great inter-annual variability (standard deviation 431). For potatoes, on the other hand, the productivity has increased from 1709 to 2278 kg/ha, during the same period, with a standard deviations of 247. During the same time (1983–2000), the use of commercial fertilizers has increased from


203

Figure 7. Main crop (crop covering the largest area) in the 1980s and 2002. 1980s refers to the period after the introduction of the Household Responsibility System in the area (1982).

Figure 8. Reasons why farmers have changed main crops during the last 20 years. A total of 30 out of 38 interviewed farmers had changed their crop.

3334 to 10838 ton in Ansai County contributing to the increase productivity (Ansai County Statistical Data, 2000). To get an indication of what the agriculture is most dependent on, the farmers were first asked to choose the most important thing for their main crop (the crop covering the largest area) to grow well. This question was open-ended, and 79% of the respondents answered fertilizer. The questions that followed were close-ended (question with fixed alternatives) and included five factors that the farmers had to rank according to their importance to their current main crop (2002) and that of 20 years prior. Table IV shows that fertilizer was still ranked very high, but rainfall became very important. The same number of farmers ranked rainfall and fertilizer as the most important factor influencing their crops today. One explanation is that when the farmers get an open-ended question, they only think of things that they can influence. As one farmer explained: “I plant

204


TABLE IV Ranking factors influencing agriculture. 1980s refers to the period after the introduction of the Household Responsibility System in the area (1982) Ranking

1980s

2002

1 2 3 4 5

Rainfall Fertilizer Labour Topography

Fertilizer and rainfall Labour Topography Adding water

Ranking of 1 is most important and 5 is least important.

the potatoes, add the fertilizer and then I wait until harvest” (male, 66 years old). Rainfall on the other hand is nothing they can change. They are used to the climate and the lack of rain. The result shows that they are aware of the importance of rainfall to the crops. The reason why adding water is not included in the ranking in the 1980s (Table IV) is because millet was the main crop and was mainly planted on rain-fed hill slope land. The third question to examine is what are other factors affecting land-use change and adaptation by farmers. As seen above, economic and policy driven changes influenced farmers’ landuse practice. An example is when the government employed experts from Shandong Province to introduce the method of using plastic covers over fields to the farmers in the 1980s and 1990s. During the group discussions, the farmers were asked also to comment on the Cropland Conversion Program. “The policy is good, planting trees conserves soil and water. The trees planted last year are all alive, but the government did not give us the food” (male group discussion). They want to have a dialogue with the local government about where and what kind of trees should be planted. A suggestion from the farmers is to plant fruit trees or other kinds of economic trees that can generate income for generations to come. Farmers might not be able to continue planting trees since the practice generates no income (Hou, 2002, personal communication). Some farmers have already resumed cultivation on hill slope land. The consequences of this policy shows how the land area used for cropping, especially hill slope land, has diminished during the last 20 years (Table V). The flat land area has increased, but it is still only a small fragment of the land that can be irrigated. The men prefer planting crops and do off-farm work in their spare time, while the women prefer off-farm work because they feel that farming is the worst kind of job. Currently the women plant while the men do off-farm work. 79% of the interviewed households in the watershed have off-farm work. Examples of off-farm work are


205

TABLE V Change of cultivated land area between 1982 and 2002. 1980s refers to the period after the introduction of the Household Responsibility System in the area (1982) Total cultivated land areaa (m2 /person)

1982 2002 1982–2002 Land area change in percent (%)

Total cultivated land area (m2 /person)

Hillslope

Flat

Total land

Irrigated

Rainfed

Mean total cultivated land area (m2 /person), with variations in brackets

3960 580 −3380 −85

290 410 120 41

4250 990 −3260

180 200 20 11

4060 950 −3110 −77

(4240–4250) ∼ 4245 (990–1150) ∼ 1070 −3175 −75

a In Danangou they measure the land in mu. One mu was originally defined as “the area a cow can cultivate in one day” and equals 666.7 m2 . The population has been approximately 215 since the introduction of the Household Responsibility System until 2002.

transportation, construction work and small businesses. Eighteen percent started with off-farm work in 1982 and 41% started with it after 1997. Everyone is still working with agriculture, but mainly in combination with livestock, planting fruit trees or off-farm jobs. During the structured interviews, the farmers were asked what they would like to work with in the future. Five percent answered planting crops, 16% live stock, 13% planting trees, 21% off-farm work and 24% want to plant crops in combination with the alternatives mentioned above. Thirteen percent have no future plans. It is mainly younger males that do not know what to do in the future: “I can think, but not reach my dreams” (male group discussion). The villagers have ideas for improving their living conditions, but require economic support to do so. The men think that the future of the village depends on the society and how much help they receive from the outside, mainly from the local government. They would prefer help to improve the village instead of working in town. “A job in town would only help us, not the next generation. If we could get help to build a dam and terraces, it will also help our grandchildren” (male group discussion). They think education and off-farm work will be important for the next generation. “I do not want my children to live in Danangou; it is too hard to live here. There is no future in farming” (female group discussions). 5. Discussion 5.1.

CLIMATE AND PERCEPTIONS

The decreasing rainfall trend is weak during the 32-year period (Figure 3). The yearto-year variability is larger and the trend could be due to natural variations. The

206


result from the 20-year period strengthens this interpretation. Earlier studies have shown a 20–30-year oscillation (with no trend) in the rainfall in eastern China based on a 120 year long dataset (Wang, 2002, personal communication). According to Chen et al. (1994) and Qin (2002), the rainfall in this region is decreasing though. This makes it hard to speculate about future rainfall. The increasing temperature trend in Ansai (Table I) agrees well with other research in the region (Chen et al., 1994; Qin, 2002). The increase in temperature will be higher in higher latitudes and more obvious in winter (Wang, 2002, personal communication). Global warming could possibly be seen on the local level. How the warming trend in China might affect the precipitation pattern is unexplored. According to Chen et al. (1994), the thermal difference between the mainland and the ocean in East Asia has been reduced during the last 40 years. This results in a weaker summer and winter monsoon that could contribute to a decrease in rainfall. The farmers think that drought occurs more frequently today than 20 years ago. This agrees with Qian and Zhu’s (2001) results. According to the farmers in Danangou watershed, drought is the disaster that causes most damage to their agriculture. This coincides with research showing that drought disasters cause the largest economic loss among disasters (Qian and Zhu, 2001). The strong agreement among the farmers could be explained by the last three very dry years. The Climate Game was constructed so that a general picture or trend of climate in the past could be received. The main purpose was to present perceptions of climate change. In general the farmers agree that it is raining less now than 20 years ago, but they cannot specify the change since the variability of the rainfall is large. Temperature is increasing continuously according to the farmers. This agrees with both earlier research and the climate data analysis conducted in this study. The analysis of perception of climate has shown that the farmers in Danangou watershed agree on a warming and drying trend. The warming trend is most noticeable during winter. Flooding conditions are less common in recent years, while there has been a persistent drought during the last three years. In this study, nothing was found to contradict earlier research (Messing and Hoang Fagerström, 2001; Hoang Fagerström et al., 2003), in which it was reported that farmers remember details three years back in time. Further back they remember extreme events. To this we can add that the farmers also have a good perception of general climate conditions and long-term climate trends.

5.2.

RESPONSE AND ADAPTATION: CLIMATE, LAND USE AND OTHER FACTORS

The change of main crop (Figure 7) shows that today the farmers plant crops that are less resistant to droughts. One explanation could be that their living situation has changed drastically during the last 20 years, mainly due to economic changes in China, i.e. the Open-Door Policy, which has given farmers the opportunity to buy and sell food to a larger extent, showing the importance of economic considerations


207

as aspects of adaptation (Olesen and Bindi, 2002). Other possibilities include policies affecting land use, i.e. the Household Responsibility System, giving each family a larger possibility to choose a way of farming, and the Cropland Conversion Program, which restricts farmers to farm on certain lands. Further, the farm practice that has continuously changed over the last 29 years has increased the average productivity of the crops such as potatoes, as seen in the statistics. The use of fertilizers has increased over the same period. These results accompany the idea of multiple pressures/changes that has advanced in climate change adaptation research over the last decade (IPCC, 2001; Reilly and Schimmelpfennig, 1999; Smit et al., 1999). Today the farmers have a very limited crop area that is mainly located on the flatland, where it is easier to irrigate (Table V). It could also be a combination of the factors mentioned above. As a consequence of recent development, the farmers have changed from planting food crops to cash crops, i.e. an adaptive response to improve their situation (Reilly and Schimmelpfennig, 2000). Their change of crops shows that a more unreliable climate has not prevented farmers from starting to planting cash crops with low drought resistance, i.e. climate change has not been the driving force of this land-use change. Instead, economic values, including access to markets and higher productivity driven by technical improvements and use of fertilizers, are the driving forces behind the change. Farmers cannot currently earn their living from only cultivating the small fraction of flatland. This is one reason for the transfer from agriculture to off-farm work that has occurred in recent years. Climate variability could be an additional reason, even though it is hard to distinguish the extent of influence. A change of the land use is seldom due to a single weather event, but rather to climate change on a longer time scale. Together with other factors, i.e. policy, technical improvements and market changes, the need to respond can compound. In Danangou, the last 20 years has brought many changes, and it is therefore difficult to say to what extent climate variability has impacted the land-use change that has occurred. Climatologically, the last 20 years has become drier and during the last three years, there has been severe drought. During the same three years, changes in main crop and income sources have made farmers less dependent on agriculture and therefore the climate. Roncoli et al. (2001) indicate that diversification, non-farm income sources and growing drought resistant crops (not the case in Danangou) are some ways of coping with climate variability. Scoones (1998) defines diversification as a livelihood strategy responding to a changing world, which could be advantageously applied in Danangou. Earlier research in China shows that crop diversification is an adaptive response to effects of global warming (Reilly and Schimmelpfennig, 1999). For the farmers to give up most of their cropland in the Cropland Conversion Program, but not receive their compensation, was upsetting. The farmers have realized the importance of planting trees, but suggest instead that trees that can generate income should be planted. Currently, the Cropland Conversion Program has made the farmers find other income sources, and they are no longer

208


solely dependant on agriculture. Off-farm work is the main income for most of the farmers. The climate has varied during the last two decades. Land-use change has occurred but not necessarily with climate as the driving force since economic and policy effects have been seen in Danangou, and the farmers have adapted. One relevant question is whether the farmers have succeeded in their adaptation, i.e. have they increased the level of their livelihood? According to the farmers, “It is getting better and better.” Their standard of living has increased, and one should therefore conclude that they have succeeded. However, if climatic conditions had not become drier and warmer, the farmers might have been doing even better, i.e. positive and negative influences have been working simultaneously, but according to the farmers the overall impact has been positive. However, the outcome of farmers’ adaptation is not always viewed as positive. A negative effect of off-farm work is that many farmers have to leave their village to find jobs, and they do not know how long they will have the job. The off-farm income is not as reliable as that from agriculture. It is noteworthy that the men, who do most of the off-farm work, would prefer planting crops, while the women, who do the planting today, would prefer off-farm work. However, they all think that planting has a very low status. Their plans indicate that they are realistic about their own future, but they still dream of a better future for their children. They all agree that the best future for their children involves getting an education and moving from the village. While presenting their ideas, the farmers also indicated that they were dependent on sources from outside the village. It is mainly the government (local) that is mentioned as the main agent to finance or help the village. This can be a sign that farmers do express and demand more today than previously due to a greater openness, or it can be a sign that farmers are more dependant and less self-sufficient today than before. Hence, the issue is whether small-scale farming is possible or even desired in the future since (a) the yield does not sustain the farmer, (b) the farmer would prefer to do off-farm work if possible, (c) farmers wish their children a future away from farming, and (d) environmental policy is proposing less agricultural activity in the Loess Plateau. Seen in the regional context of temperate Asia, the IPCC Impact Assessment states that “this region appears to be especially sensitive to climate change because of potential increase in the soil moisture deficit” (IPCC, 1998, p. 359) and summarizes three years later that “food insecurity appears to be the primary concern for Asia” (IPCC, 2001, p. 47). One further relevant issue not addressed here is whether the farmers might “over-adapt” to the present situation, in the sense that one or several of the ongoing changes might stop, reverse or simply change, e.g., the present weather condition may be due to a 20–30-year oscillation cycle and could return to a wetter and colder condition within a few years. Or there might be a change in availability of off-farm work or change in markets for agricultural produce now farmed in Danangou. New policies might be introduced. In that respect, the issue of adaptive resilience is of


209

relevance. Are farmers prepared for future adaptations? Is the diversified livelihood that has developed a prerequisite or a burden for possible new changes? To assess the impact of land-use changes in relation to climate variability, it is observed that the farmers in Danangou watershed have adapted their income sources, agricultural methods and type of cultivated land. They are now less dependent on the climate. This adaptation is mainly caused by economic changes, technical improvements and governmental policies. However, it is noteworthy that the farmers are aware of climate variability and their perception indicates the crops’ dependency on the climate. As a farmer said: “The climate is the nutrient of the crop.” The farmers’ dependency on the climate, even though farming is not their sole income source, indicates the importance of being able to predict future climate changes.

6. Conclusions • Local temperature has increased with 1 ◦ C during the last 32 years. The increase is strongest in winter (2 ◦ C). • During the last 32 years, local precipitation has decreased with 14% (2.6 mm/year), while regional precipitation has decreased with 19% (1.8 mm/year) during the last 49 years. The interannual variability is larger. During the last 20 years there has been a significant decreasing trend (95% significance level) in local precipitation (−11.7 mm/year). The last three years have been very dry. • Farmers’ perception of climatic variability corresponds with the climatic data records. The farmers can give detailed descriptions of the weather the last few years. Further back in time they recognize trends. They think it has become warmer, especially during winter. • Reforms and policies have a stronger influence, in terms of the farmers’ choice of cropping system, than climate variability on the land use in Danangou watershed. • During the last 20 years, the farmers have switched from a dependence on agriculture to a more diversified livelihood due to economic changes and governmental reforms. This combination of planned and non-conscious planned adaptation makes the farmers less vulnerable to climate variability and can be seen as a responce to a changing situation, both climatologically and economically. • The findings on adaptation in this study correspond well with results from Burkina Faso (Roncoli et al., 2001) involving development towards a diversified livelihood, while our findings showed that farmers’ perceptions and knowledge of general climatic trends were good compared to that among land users in Western Himalayas where only extremes or maximums in the weather were the perceived (Vedwan and Rhoades, 2001).

210


• To understand local impact of climate variability and/or policies, this type of local study would benefit regional or national multi-sector analyses, as has been proposed by Reilly and Schimmelpfennig (1999), Magistro and Roncoli (2001) and Vedwan and Rhoades (2001). Acknowledgments The researchers would like to thank the Department of Natural Resources and Environmental Science, Beijing Normal University, the National Climate Centre (NCC), the Department of Soil and Water Conservation at Yangling University, the Township of Zhengwudong, and all the farmers in the watershed in China for help and support. The research has been supported by the Swedish International Development and Cooperation Agency through projects SWE-2001-154 and SWE-2002-038 and Minor Fields Study scholarships, and by STINT through an Institutional Grant to Deliang Chen. We thank John Reilly, Ron Prinn and Anne Slinn at the Joint Program at MIT and the anonymous reviewers for useful comments. References Ansai County Statistical Data: 2000, Agricultural Statistics, Ansai Statistical Office. Berg, L. B.: 1998, Qualitative Research Methods for the Social Sciences, Allyn and Bacon, Boston, p. 290. Bernard, H. R.: 1995, Research Methods in Anthropology-Qualitative and Quantitative Approaches, AltaMira Press, London, p. 585. Burroughs, W. J.: 2001, Climate Change: A Multidisciplinary Approach, Cambridge University Press, p. 298. Carlsson, M.: 2000, Farmer’s Knowledge in Land Evaluation—A Case Study of Land Suitability in a Small Catchment Affected by Erosion on the Chinese Loess Plateau, Swedish University of Agricultural Sciences, International Office, Uppsala, p. 53. Chan, K. W. and Zhang, L.: 1999, ‘The Hukou system and rural-urban migration in China: Processes and changes’, China Quart. 160, 818–855. Chen, L., Shao, Y., and Ren, Z.: 1994, ‘Climate change in China during the past 70 years and its relation to the monsoon variations’, in Zepp R. G. (ed.), Climate Biosphere Interaction: Biogenic Emission and Environmental Effects of Climate Change, John Wiley and Sons, New York, pp. 31–49. Chen, L., Wang, J., Fu, B., and Qiu, Y.: 2001, ‘Land-use change in a small catchment of northern Loess Plateau, China’, Agric., Ecosyst. Environ. 86, 163–172. Ding, Y.: 1994, Monsoons Over China, Atmospheric Sciences Library, Kluwer Academic Publishers, The Netherlands, p. 419. Domrös, M. and Peng, G.: 1988, The Climate of China, Springer-Verlag, Berlin, p. 361. EROCHINA: 2000/2001, A Participatory Approach for Soil and Water Conservation Planning, Integrating Soil Erosion Modeling and Land Evaluation, to Improve the Sustainability of Land Use on the Loess Plateau in Northern China—EROCHINA, Soil Erosion Project-Final Report, Contract Number: ERBIC18CT970158, The Netherlands, p. 89. FAO-UNESCO: 1997, Soil Map of the World: VIII, North & Central Asia, Unesco, Paris. Fu, B., Chen, L., Ma, K., Zhou, H., and Wang, J.: 2000, ‘The relationships between land use and soil conditions in the hilly area of the Loess Plateau in northern Shaanxi, China’, Catena 39, 69–78.


211

Fu, B., Wang, J., Chen, L., and Qui, Y.: 2003, ‘The effects of land use on soil moisture variation in the Danangou catchment of the Loess Plateau, China’, Catena 54, 197–213. Hageback, J. and Sundberg, J.: 2002, Climate Variations in Relation to Local Scale Land Use and Farmer’s Perception of Climate in Danangou Watershed on the Loess Plateau, China, Earth Sciences Centre, Göteborg University, B355. Hoang Fagerström, M. H., Messing, I., and Wen, Z. M.: 2003, ‘A participatory approach for integrated conservation planning in a small catchment in Loess Plateau, China. Part 1. Approach and methods’, Catena 54, 255–269. Hsieh, C.: 1973, Atlas of China, McGraw-Hill, New York, p. 282. IPCC: 1998, The Regional Impacts of Climate Change: An Assessment of Vulnerability, Cambridge University Press, New York, p. 517. IPCC: 2001, Technical Summary Climate Change 2001: Impacts, Adaptation and Vulnerability, Cambridge University Press, New York, p. 56. Johnston, T. and Chiotti, Q.: 2000, ‘Climate change and the adaptability of agriculture: A review’, J. Air Waste Manage. Assoc. 50, 563–569. Krausmann, F., Haberl, H., Schulz, N. B., Erb, K.-H., Drage, E., and Gaube, V.: 2003, ‘Land-use change and socio-economic metabolism in Austria. Part 1. Driving forces of land-use change: 1950–1995’, Land Use Policy 20, 1–20. Liu, G.: 1999, ‘Soil conservation and sustainable agriculture on the Loess Plateau: Challenges and prospects’, Ambio 28, 663–668. Lu, M. and Wang, E.: 2002, ‘Forging ahead and falling behind: Changing regional inequalities in post-reform China’, Growth Change 33, 42–71. Lü, X. and Perry, J.: 1997, Danwei the Changing Chinese Workplace in Historical and Comparative Perspective, M.E. Sharpe, Inc., New York, p. 259. Luo, Q. and Lin, E.: 1999, ‘Agricultural vulnerability and adaptation in developing countries: The Asia-Pacific region’, Climatic Change 43, 729–743. Magistro, J. and Roncoli, C.: 2001, ‘Anthropological perspectives and policy implications of climate change research’, Climate Res. 19, 91–96. Mearns, L. O., Rosenzweig, C., and Goldberg, R.: 1997, ‘Mean and variance change in climate scenarios: Methods, agricultural applications, and measures of uncertainty’, Climatic Change 35, 367–396. Messing, I. and Hoang Fagerström, M. H.: 2001, ‘Using farmers’ knowledge for defining criteria for land qualities in biophysical land evaluation’, Land Degrad. Dev. 12, 541–553. National Bureau of Statistics of China: 1996–2002, China Statistical Yearbook China Statistical Press. National Climate Center: 2000, An Assessment System for Impact of Climate Anomaly on Socio-Economic Aspects, National Key Project (1996–2000): Studies on Short-Term Climate Prediction in China, Subproject 3, Beijing, China, p. 28. National Bureau of Statistics of China: 2000, Communiqué on Major Figures of the 2000 Population Census No. 1, http://www.stats.gov.cn/english/newrelease/statisticalreports/200204230084.htm (9 January 2003). Olesen, J. E. and Bindi, M.: 2002, ‘Consequences of climate change for European agricultural productivity, land use and policy’, Eur. J. Agron. 16, 239–262. Ovuka, M. and Lindqvist, S.: 2000, ‘Rainfall variability in Murang’a District, Kenya: Meteorological data and farmers’ perception’, Geogr. Ann. A 82, 107–119. Polsky, C. and Easterling III, W. E.: 2001, ‘Adaptation of climate variability and change in the US Great Plains: A multi-scale analysis of Ricardian climate sensitivities’, Agric., Ecosyst. Environ. 85, 133–144. Qian, W. and Zhu, Y.: 2001, ‘Climate change in China from 1880 to 1998 and its impact on the environmental condition’, Climate Change 50, 419–444.

212


Qin, D. H.: 2002, Assessment on Environmental Change in Western China, Vol. 1, China Sciences Press, Beijing, pp. 29–67. Qiu, Y., Fu, B., Wang, J., and Chen, L.: 2001, ‘Soil moisture variation in relation to topograpghy and land use in a hill slope catchment of the Loess Plateau, China’, J. Hydrol. 240, 243–263. Reilly, J. M. and Schimmelpfennig, D.: 1999, ‘Agricultural impact assessment, vulnerability and the scope for adaptation’, Climatic Change 43, 745–788. Reilly, J. M. and Schimmelpfennig, D.: 2000, ‘Irreversibility, uncertainty, and learning: Portraits of adaptation to long-term climate change’, Climatic Change 45, 253–278. Roncoli, C., Ingram, K., and Kirshen, P.: 2001, ‘The costs and risks of coping with drought: Livelihood impacts and farmers responses in Burkina Faso’, Climate Res. 19, 119–132. Rosmuller, C.: 1999, Socio-Economic Aspects of Catchment Management Planning for Small Watershed on the Chinese Loess Plateau. Erosion and Soil and Water Conservation Group, Wageningen Agricultural University, Wageningen, p. 24. Rui, L., Zhongming, W., Fei, W., Yue, W., and Mingliang, Z.: 2001, Case study on conversion of farmland to forest and grassland in Ansai County, Shaanxi Province, Prepared for the CCICED Western China Forest Grassland Task Force, p. 22. Scoones, I.: 1998, Sustainable Rural Livelihoods a Framework for Analysis, IDS Working Paper 72, p. 22. Smit, B., Burton, I., Klein, R. J. T., and Street, R.: 1999, ‘The science of adaptation: A framework for assessment’, Mitig. Adapt. Strat. Global Change 4, 199–213. Smit, B., Burton, I., Klein, R. J. T., and Wandel, J.: 2000, ‘An anatomy of adaptation to climate change and variability’, Climatic Change 45, 223–251. Smit, B. and Cai, Y.: 1996, ‘Climate change and agriculture in China’, Global Environ. Change 6, 205–214. Wang, J., Fu, B., Qiu, Y., and Chen, L.: 2001, ‘Soil nutrients in relation to land use and landscape position in the semi-arid small catchment on the Loess Plateau in China’, J. Arid Environ. 48, 537–550. Vedwan, N. and Rhoades, R. E.: 2001, ‘Climate change in the Western Himalayas of India: A study of local perception and response’, Climate Res. 19, 109–117. Verburg, P. H. and Chen, Y.: 2000, ‘Multiscale characterization of land-use patterns in China’, Ecosystems 3, 369–385. Wu, Y., Xie, K., Zhang, Q., Zhang, Y., Xie, Y., Zhang, G., Zhang, W., and Ritsema, C. J.: 2003, ‘Crop characteristics and their temporal change on the Loess Plateau of China’, Catena 54, 7–16. Yeh, A. G. and Li, X.: 1999, ‘Economic development and agricultural land loss in the Pearl river delta, China’, Habitat Int. 23, 373–390. Hou, X. L.: May 2002, Research Assistant at Ansai Research Station of Soil and Water Conservation, Ansai, China. Li, D.: April 2002, Institute of Soil and Water Conservation, Yangling, China. Wang, S.: April 2002, Physical Department, Beijing University, Beijing, China. (Received 18 April 2003; in revised form 29 July 2004)