Environ Monit Assess (2010) 168:159–171 DOI 10.1007/s10661-009-1100-9
Extensive tree health monitoring networks are useful in revealing the impacts of widespread biotic damage in boreal forests Seppo Nevalainen · Martti Lindgren · Antti Pouttu · Jaakko Heinonen · Marke Hongisto · Seppo Neuvonen
Received: 28 November 2008 / Accepted: 13 July 2009 / Published online: 24 July 2009 © Springer Science + Business Media B.V. 2009
Abstract We surveyed the regional distribution of conifer defoliation in Finland with an extensive monitoring network during 1995–2006 (EU Forest Focus Level I). The average defoliation in the whole Finland was 10.3% in pine and 19.9% in spruce. The sharp changes were often related to abiotic and biotic factors. The mean age of the stand explained more than one half of the between-plot variance in defoliation. In a variance component analysis, the main effect of years was negligible, while most of the random variation was due to plot main effect and plot × year interaction. About one fifth of the defoliation could be attributed to abiotic or biotic damage, and there were strong local correlations, e.g., between the changes in defoliation and degree of pine sawfly (Diprionidae) damage. There were clear temporal and spatial patterns in the incidence of the most important causes [Scots pine: Scleroderris
S. Nevalainen (B) · J. Heinonen · S. Neuvonen Joensuu Research Unit, Finnish Forest Research Institute, P.O. Box 68, 80101, Joensuu, Finland e-mail:
[email protected] M. Lindgren · A. Pouttu Vantaa Research Unit, Finnish Forest Research Institute, P.O. Box 18, 01301, Vantaa, Finland M. Hongisto Air Quality Research, Finnish Meteorological Institute, P.O. Box 503, 00101, Helsinki, Finland
canker (Gremmeniella abietina), pine shoot beetles (Tomicus sp.), and pine sawflies (Diprion pini, Neodiprion sertifer); Norway spruce: rust fungi (primarily Chrysomyxa ledi)]. Our results suggest that extensive monitoring networks can reveal useful information about the widespread outbreaks of pest organisms (insects and fungi) already in their increase phases, giving some time for management decisions. In a changing climate, large-scale, regular monitoring of tree health, including abiotic and biotic causes, is more important than ever before. Keywords Pinus sylvestris · Picea abies · Finland · Defoliation · Forest health monitoring · Forest pests · Fungal diseases
Introduction Concern about large-scale decline in forest vitality both in Europe and North America some decades ago led many countries to initiate extensive national surveys of forest condition (Lorenz 1993; Alexander and Palmer 1999). In Europe, the monitoring of forest condition and the effect of stress factors on ecosystem functioning has been carried out in 38 countries within the EU Forest Focus and UN/ECE–ICP Forests monitoring programs (Lorenz et al. 2007). The monitoring was first carried out under the International Co-operative
160
Programme on the Assessment and Monitoring of Air Pollution Effects on Forests (ICP Forests), which is based on international agreements on the long-range transportation of air pollutants (LRTAP). In EU member states, forest condition monitoring is based on regulations enacted in 1986 and 1994 and on their subsequent modifications. Since 2003, the monitoring program has been carried out under the EU Forest Focus regulation. At the beginning of the forest health monitoring programs, the main interest was directed at the impact of air pollution (e.g., Prinz 1985). In recent years, however, more attention has been paid to other stress factors, e.g., abiotic and biotic (pathogenic fungal and insect) damage and their interactions. Also, stand or tree age is recurrently reported to be the key factor affecting tree/ forest condition (Thomsen and Nellemann 1994; Seidling and Mues 2005), which should be taken into account when analyzing and interpreting the results. A rapid deterioration in forest vitality has been attributed to abiotic or biotic damage in several reports (Keane et al. 1989; Innes and Schwyzer 1994; Sioen and Roskams 2006; Wulff et al. 2006). It has even been proposed that tree health merely reflects the fluctuating effects of biotic or abiotic agents or site conditions (Skelly and Innes 1994). Edgar and Burk (2006) used a simulation study to evaluate the sensitivity of an extensive forest health monitoring network to outbreaks of defoliating insects. They concluded that it is unlikely that outbreaks smaller than 100,000 ha would defoliate on average more than a handful of plots in such a network and that the real utility of the network is in monitoring phenomena that affect very large areas. Relatively few studies have been carried out on the defoliation of conifers and influence of stress factors in the boreal zone. The technical reports and executive summaries published by the ICP Forests Programme between 1999 and 2008 (http://www.icp-forests.org) provide a comprehensive set of literature on this topic (see, e.g., Lorenz et al. 2008). In the boreal zone, forest cover is generally more continuous than in more densely populated temperate areas, and consequently, some of the problems related to the combination of sparse monitoring networks with
Environ Monit Assess (2010) 168:159–171
scattered and fragmented forest cover are alleviated (cf. Ferretti 1997; Seidling and Mues 2005). Significant impacts of air pollution (NOx , SOx , acidic deposition, and tropospheric ozone) have been reported mainly in more southern latitudes, e.g., in parts of central and Eastern Europe, and particularly in the case of deciduous species. The impacts of pollution on conifers may be less significant, at least partly because of the relatively low deposition levels in the boreal regions where conifers predominate. No significant correlation has been found in Finland between the defoliation degree on conifers and sulfur, nitrogen, and heavy metal deposition at the national level (Lindgren et al. 2000). However, correlation between needle discoloration on Norway spruce and modeled sulfur and nitrogen deposition has been reported (Lindgren et al. 2000). Soil acidity parameters did not explain the variation in defoliation, but a relationship was found between the C/N ratio in the organic layer and the defoliation degree on Norway spruce: spruces growing on relatively nutrient-poor sites suffered more from defoliation (Lindgren et al. 2000; Mälkönen et al. 2000). In this article, we focus on tree health as measured by the degree of defoliation (needle loss), including the pathological aspect, and pay special attention to the incidence of biotic factors affecting the tree (Innes 1993). We surveyed the regional distribution of tree health, with an extensive monitoring network, during an extended period (1995–2006) in Finland, with the aim to answer the following questions: (1) How large is the annual variation in defoliation between plots and within plots over years (i.e., plot × year interaction)? (2) What is the relative contribution of tree age to the variation in tree defoliation levels? (3) How much do the biotic damage agents contribute into the variation in the level of defoliation of conifers? (4) How useful/sensitive are extensive tree health monitoring networks to reveal the impacts of widespread biotic damages? In addition, we checked for possible patterns that might be attributable to sulfur and nitrogen depositions (cf. Kandler and Innes 1995).
Environ Monit Assess (2010) 168:159–171
Materials and methods Study material Permanent sample plots (3,009) were established during 1985–1986 in connection with the 8th National Forest Inventory (NFI). A systematic subsample of 450 mineral soil plots was initially selected for extensive forest health monitoring (for details, see, Jukola-Sulonen et al. 1990; Salemaa et al 1991). These plots represented the EU Forest Focus Level I network in Finland. The country was divided into a southern and a northern region (demarcation line 66◦ N). The network in the southern region was based on a 16 × 16-km grid, and in the northern region, on a 24 × 32-km grid. Because of the fixed plot size (300 m2 ) used in Finland during 1986–1994, the number of sample trees on many of the plots was insufficient to fulfill the minimum number of 20 trees per plot in Southern Finland and ten trees per plot in Northern Finland (Commission Regulation 1995 [EC] No. 1398/95). To fulfill the regulation, new trees were added systematically to the network by increasing the radius of the plot in 1995. The increase in radius was dependent on the density of the stand (Lindgren et al. 2000). In addition, some new sample plots were added to the sample in summer 2004 to replace, e.g., cut plots. The new plot was selected from the same permanent 8th NFI cluster. Only results concerning coniferous trees growing on mineral soil sites are reported in this paper. The number of plots and trees varied as follows: 4,385–5,113 Scots pine trees on 362–419 plots per year and 2,644–2,808 Norway spruce trees on 254–290 plots per year.
161
defoliation, which is expressed as the relative leaf or needle loss compared to a reference tree (Lindgren et al. 2000); changes in discoloration will be treated elsewhere. An imaginary or a real, similar-aged tree with full foliation, growing on the same site was used as a reference (JukolaSulonen et al. 1990). Defoliation was assessed on trees growing in the dominating crown layer (dominant and codominant trees with a diameter at breast height of at least 45 mm). Defoliation of Norway spruce was estimated on the upper half of the living crown, and of Scots pine (Pinus sylvestris L.), on the upper two thirds of the living crown in 5% classes (class 0 = 0–1%, 5 = 1–5%, 10 = 6–10%, 15 = 11–15 . . . 95 = 91–95, 99 = 96–99, 100 = dead). A national system for the description of the damage: symptom, apparent severity (degree of damage), and the cause, as well as the age of the damage, was used prior to 2004 (for more details, see, Nevalainen 1999; Nevalainen et al. 2007). The ICP Forests manual of damage causes (referred to as the biotic manual; ICP Forests 2004) was fully adopted in 2005 in Finland. The principles of the national damage survey in Finland before 2004 were to a large extent similar to the current biotic manual (see, Nevalainen et al. 2007). In this paper, abiotic damage refers to damage attributed to climatic (frost, wind, snow) and edaphic (soil) factors (drought, flooding, nutritional disorders), separated from anthropogenic factors (air pollution). The survey was carried out annually, during July–August, by ten to 12 trained observers. About 5% of the sample plots were included in a control survey each year. The control inventory covered all basic measurements and assessments made by the field teams.
Assessment of tree health Environmental variables The most important variables used to describe tree health in Finland were relative leaf and needle loss, i.e., defoliation, extent of discolored leaf and needle mass, and abiotic and biotic damage. These variables were assessed visually according to internationally standardized methods (UN/ECE 1998) and national field guidelines (e.g., Lindgren et al. 2005). Here, we focus on
Because the plots were originally permanent sample plots of the 8th NFI, a large number of stand and environmental variables were readily available. Stand mean age at breast height (the age at breast height of a tree representing the median diameter in the stand) and site type were most often used as explanatory variables. The mean
162
Environ Monit Assess (2010) 168:159–171
stand age was 80.2 years (SD, 44.2, range, 13– 245 years). The original site types (eight classes) were recoded into four classes. The new types (and their proportion in our material, in parentheses) were as follows: (1) grove-like sites, grass– herb forests, and more fertile sites (19.3%); (2) fresh (mesic) sites with Myrtillus understorey vegetation (45.3%); (3) dryish (subxeric) sites with Vaccinium vegetation (28.5%); and (4) dry upland forest (xeric) sites and less fertile sites (6.9%). For the Finnish site type classification, see, e.g., Kujala (1979). The deposition of nitrogen and of sulfur were calculated using the Hilatar simulation model of the Finnish Meteorological Institute (FMI; Hongisto 2003a, b; Hongisto et al. 2003). The model has been validated by model–measurement intercomparisons using monthly, weekly or daily measurement data (concentrations in air and precipitation) of over 90 European EMEP stations (http://www.emep.int), using measurements from field campaigns (Schulz et al. 1999; Sofiev et al. 2001) and model–model intercomparisons (Zlatev et al. 2001). For our study, annual deposition on the field plots was calculated using the FMI operational HIRLAM meteorological forecasts for the year 2005 and European emissions for the year 2002, which, at the time of the simulations, were the latest emission data available in the EMEP emission database (http://www.emep.int). On our plots, the modeled sulfur depositions varied from 126 mg m−2 year−1 in the northernmost Finland to 661 mg m−2 year−1 in the south (mean, 239 mg m−2 year−1 ). The modeled total nitrogen deposition varied from 265 to 799 mg m−2 year−1 (mean, 505 mg m−2 year−1 ). Data analysis The overall contribution of a biotic or abiotic damage d to defoliation was calculated as:
dd − d h × nd ntot
.
(1)
In formula 1, the defoliation values are not adjusted by other factors that may affect defoliation.
The formula was derived by writing the mean defoliation of all the trees in the form: d¯d − d¯h × nd d¯tot = d¯h + (2) ntot where is the sum over all different damage types, d¯tot is the mean defoliation of all the trees, nd is the number of trees damaged by cause d, ntot is the total number of trees, d¯d is the mean defoliation of trees damaged by cause d, and d¯h is the mean defoliation of trees without damage symptoms. Variance component analysis was used to study the relative roles of stand mean age, year main effects, plot main effects, and plot × year interaction on the variation of defoliation. The analysis was performed for untransformed defoliation values using module Varcomp in SPSS Version 15.1 (SPSS, Chicago, IL, USA). Cross-sectional analysis of defoliation was computed by applying linear mixed models to 1-year data at a time. The dependent variable was the square root of the plot mean value. The distribution of the square root-transformed values was close to a normal distribution and the variance did not depend on the level of defoliation. The variance was, however, inversely proportional to the number of trees on the plot. This was taken into account by using weighted analysis with the number of trees on the plot as weights. Spatial correlation was taken into account by means of a spherical distance-dependent covariance function and allowing the nugget effect. The age-adjusted annual average defoliation values and their confidence intervals were computed using the weighted least square means of untransformed plot averages of defoliation at the mean age of the plots. Continuous covariates in the model were age and age2 . The between-year correlations were taken into account by a heterogeneous Toeples covariance structure. Based on Akaikes’s information criteria (AIC), a spatial correlation structure was not included in the model. The trend in defoliation was analysed using Kendall correlation coefficients between time and defoliation as dependent variables. The coefficient
Environ Monit Assess (2010) 168:159–171
was computed for each plot that was visited at least eight times. Spatial correlation was taken into account in the linear mixed models by an exponential distance-dependent covariance function and allowing a nugget effect. Kendall’s partial correlation coefficient and Kendall’s correlation coefficient were used to examine the relationships between the defoliation level and modeled sulfur and nitrogen deposition. Kendall’s correlation coefficient was also used in screening the correlations between the occurrence of diseases and changes in the defoliation level. In all applications of linear mixed model analysis to the spatial data, a model with uncorrelated errors, a model with a spherical covariance function, and a model with an exponential covariance function were compared, and the model resulting in the smallest AIC was selected. Linear mixed model analyses were computed using the SAS Mixed Procedure, Version 8 (SAS Institute, Cary, NC, USA). The theory of linear mixed model analysis is given, for example, in Searle et al. (1992). The covariance functions mentioned are described in detail in the SAS/STAT User’s Guide. The local correlations between the change in defoliation and degree of damage, e.g., the degree of pine sawfly (Diprionidae) damage in 2000, were computed for each plot using tree-wise observations and including all the assessed trees within a range of 200 km around the plot. Spatially smoothed defoliation values for defoliation maps were computed using plot mean values and weights proportional to the product of the number of trees on the plot and the Gaussian weight function exp(−1/1,800 × d2 ) where d is the distance from the center point. The coefficient −1/1,800 results in a practical range of 73 km (the distance at which the weight is 5% of the maximum). In the spatially smoothed trend maps, the trend estimate for a geographic location was computed as the distance-weighted mean of the Kendall correlation coefficients. Coefficients were computed for all the plots that were visited at least eight times. The weight function was the Gaussian function exp(−1/1,800 × d2 ) where d is the distance from the location (in kilometers).
163
Results Changes in defoliation The average defoliation degree of pine and spruce in the whole country was 10.3% and 19.9%, respectively. A high proportion of the trees (pine, 96.3%; spruce, 74.6%) were at most slightly defoliated (needle loss less than 25%). The proportion of slightly defoliated trees increased in pine from 22.1% in 1995 to 31.8% in 2006, and in spruce, from 32% to 43.8%, respectively. However, the proportion of moderately (needle loss 26–60%) or severely (needle loss more than 60%) defoliated trees decreased in both species. These proportions decreased from 4.9% to 3.4% in pine and from 27.8% to 21.8% in spruce between 1995 and 2006. According to variance component analysis, the main effect of years was negligible ( 1 0-
Defoliation class
Discussion The changes in defoliation in Finland during 1995–2006 were relatively small, and the sharp changes were most often related to abiotic and biotic stress factors. This has also been found at the pan-European level (de Vries et al. 2000). The proportion of severely defoliated conifers was relatively low, the number of dead trees was small, and the causes of tree death were primarily storms, snow, and chronic decay fungi. Compared to defoliation for the period 1986–1998 for the whole country (Nevalainen and Heinonen 2000), defoliation of pine has slightly increased, whereas that of spruce was at its highest in approximately 1990 and has remained relatively constant since 1995. The systematic network used in annual forest condition monitoring has been designed to provide national-level information about crown condition and its variation primarily in so-called background areas. The extensive Level I network certainly has some limitations: it is quite sparse and, especially in northern Finland, the low number of spruce plots is problematic (cf. Fig. 4 in Seidling and Mues 2005). The sample is also restricted to dominant and codominant trees and, as a result, some important frequently occurring damaging agents, such as moose damage in young
% % 5% 0% 5% 0% 10 1-1 16-2 21-2 26-6 60 > 1 0-
Defoliation class
pine stands, and some very local but severe damage (storms, insect damage) will not be revealed in our data. On the other hand, the new system of damage cause assessments has been successfully implemented in recent years in Europe and the first evaluations have been started. This harmonized pan-European recording of biotic damage provides a basis for comparisons between countries or ecoregions. Our study showed, for instance, that the occurrence of insect damage in spruce (especially Ips typographus) in Finland is very low compared to the situation in Central Europe (Schelhaas et al. 2003; Wermelinger 2004; Rouault et al. 2006). Edgar and Burk (2006) evaluated the sensitivity of an extensive forest health monitoring network to outbreaks of defoliating insects based on simulations tailored to the conditions of Minnesota, USA. The authors concluded that the real utility of the network is in monitoring phenomena that affect very large areas. Our empirical monitoring results from Finland suggest that extensive monitoring networks can reveal useful information about the widespread dynamic patterns of pest organisms (for example, insect damage). We suggest that the discrepancy between these studies is mainly caused by the following differences: (1) although the density of the monitoring networks in Minnesota and in Finland was
Environ Monit Assess (2010) 168:159–171
roughly the same (Minnesota 1.5 plots/103 km2 ; Finland, Scots pine = 1.2–1.6 plots/103 km2 ), the percentage forest cover is clearly higher in Finland (>60%; Peltola 2005) than in Minnesota (27.5%; Edgar and Burk 2006), leading to more “hits” to forested plots in Finland (cf. also Ferretti 1997); (2) Edgar and Burk (2006) simulated the occurrence of high-intensity defoliation (visible from air; cf. Edgar and Burk 2007) while we used field observations about much lower damage/defoliation levels. This means also that, with extensive monitoring network, it should be possible to detect large-scale outbreaks already in the increase phase, giving some time for management decisions. Wulff et al. (2006), studying the applicability of Swedish National Forest Inventory and National Forest Damage Inventory for estimating Scleroderris canker (G. abietina) outbreaks in pine forests, concluded that, despite of the relatively sparse sample plot density, these monitoring schemes have good potential for estimating characteristics of extensive damage outbreaks. Our results support this conclusion, and we stress the importance of annually conducted large-scale and long-term monitoring of biotic damages. Combined with information about relevant environmental variables (which are different with different damage agents), such monitoring would increase our understanding about the factors influencing the population dynamics of these damage agents. Analysis of this kind of longitudinal, spatial monitoring data is challenging because the observations are correlated over both time and space (cf. Seidling and Mues 2005). In addition, the potential explanatory variables are intercorrelated, and a 12-year monitoring period is too short for testing the effects of, e.g., weather variables. In general, the proportion of systematic components in the variation of defoliation was quite small compared to the unexplained random variation. Exceptions to this rule were most often related to biotic/abiotic causes. The strong south–north climatic gradient in Finland is strongly correlated with a decreasing south–north gradient in nitrogen and sulfur deposition and an increasing gradient in the average pH of bulk precipitation. Thus, it might be impos-
169
sible to separate the effects of weather factors and pollution deposition when studying factors related to tree health. Although defoliation did increase in those areas where the sulfur and nitrogen deposition was the highest, the coincidence provides no information about the causal relationships. The modeled sulfur deposition in Finland is five to ten times lower than that in heavily industrialized Central Europe, and the modeled total nitrogen deposition is, according to the FMI Hilatar model calculations, more than five times higher in Central Europe. The measured wet deposition and modeled total deposition of sulfur have reduced to half of their previous levels in Finland since the beginning of the 1990s. Nitrogen deposition has decreased only slightly (Lindroos et al. 2007). Unfortunately, we were not able to include ozone in our environmental data. In Finland, the FMI carries out air pollution measurements and calculations over background and urban areas, but ozone modules connected to the Silam or Hilatar 3D models are still under construction. The mean accumulated ozone dose (over 40 ppbh) for 6-month periods during 2002–2006 varied from 4.6 to 9.4 ppmh (mean, 7.4 ppmh) at 2 m height in the air quality measurement database of the FMI. However, corrected to the tree crown height (Tuovinen and Simpson 2008), the values do not exceed the current critical level for forest trees (5 ppmh; UN/ECE 2004) at any station. Our results suggest that the main threats to the health and condition of conifers in Finland are abiotic and biotic factors. The stress approach presented by Manion (1981) is still very relevant. Under Finnish conditions, e.g., tree age and unfavorable soil conditions can act as predisposing factors. Insect defoliation, infections by Gremmeniella, extreme drought, mechanical injury, or frost damage may be the most common inciting factors, and bark beetles or root-decay fungi can then act as contributing factors. Climate change and increased frequency of extreme weather events, as well as changes in silvicultural practices, have the potential to increase biotic and abiotic damage (for instance, see, Redfern and Hendry 2002). Large-scale, regular monitoring of tree health, including abiotic and biotic causes, has, therefore, become more important than ever before.
170
References Alexander, S. A., & Palmer, C. J. (1999). Forest health monitoring in the United States: First four years. Environmental Monitoring and Assessment, 55, 267–277. Commission Regulation (EC). (1995). No.1398/95 amending Regulation (EEC) No. 1696/87 (inventories, network, reports). Brussels: 1995, Official Journal of European Communities No. L139/4 of 22 June 1995. 2 p. de Vries, W., Klap, J. M., & Erisman, J. W. (2000). Effects of environmental stress on forest crown condition in Europe. Part I: Hypotheses and approach to the study. Water, Air, and Soil Pollution, 119, 317–333. Edgar, C. B., & Burk, T. E. (2006). A simulation study to assess the sensitivity of a forest health monitoring network to outbreaks of defoliating insects. Environmental Monitoring and Assessment, 122, 289–307. Edgar, C. B., & Burk, T. E. (2007). Demonstration and verification of a model that generates defoliation patterns in forested landscapes. Ecological Modelling, 205, 301–313. Ferretti, M. (1997). Forest health assessment and monitoring—Issues for consideration. Environmental Monitoring and Assessment, 48, 45–72. Hongisto, M. (2003a). Hilatar, a limited area simulation model for acid contaminants. Part I. Model description and verification. Atmospheric Environment, 37, 1535– 1547. Hongisto, M. (2003b). Modelling of the transport of nitrogen and sulphur contaminants to the Baltic Sea Region. FMI Contributions, 40, 188 p. Hongisto, M., Sofiev, M., & Joffre, S. (2003). Hilatar, a limited area simulation model for acid contaminants. Part II. Long-term simulations results. Atmospheric Environment, 37, 1549–1560. ICP Forests. (2004). Assessment of damage causes. Submanual for manual on methods and criteria for harmonized sampling, assessment, monitoring and analysis of the effects of air pollution on forests (4th ed.). UN/ECE, Convention on Long-Range Transboundery Air Pollution, International Co-Operative Programme on Assessment and Monitoring of Air Pollution Effects on Forests. Updated June 2004. Retrieved from http://www.icp-forests.org/bioticdocs/ manual-index.pdf. Innes, J. L. (1993). Methods to estimate forest health. Silva Fennica, 27, 145–157. Innes, J. L., & Schwyzer, A. (1994). Stem damage in Swiss Forests: Incidence, causes and relations to crown transparency. European Journal of Forest Pathology, 24, 20–31. Jukola-Sulonen, E.-L., Mikkola, K., & Salemaa, M. (1990). The vitality of conifers in Finland, 1986–88. In: P. Kauppi, P. Anttila & K. Kenttämies (Eds.), Acidification in Finland (pp. 523–560). Berlin: Springer. Kandler, O., & Innes, J. (1995). Air pollution and forest decline in Central Europe. Environmental Pollution, 90, 171–180. Keane, M., McCarthy, R., & Hogan, J. (1989). Forest health surveys in Ireland: 1987 and 1988 results. Irish Forestry, 46, 59–62.
Environ Monit Assess (2010) 168:159–171 Kujala, V. (1979). Forest types of Finland. In Finnish with English summary. Communicationes Instituti Forestalis Fenniae, 92, 1–45. Lindgren, M., Salemaa, M., & Tamminen, P. (2000). Forest condition in relation to environmental factors. In: E. Mälkönen (Ed.) Forest condition in a changing environment—The Finnish case. Forestry sciences (vol. 65, pp. 142–155). Dordrecht: Kluwer Academic. Lindgren, M., Nevalainen, S., Pouttu, A., Rantanen, H., & Salemaa, M. (2005). Metsäpuiden elinvoimaisuuden arviointi.Maasto-ohje (Field Guide). Metsäntutkimuslaitos. 38 p. Lindroos, A.-J., Derome, J., & Derome, K. (2007). Open area bulk deposition and stand throughfall in Finland during 2001–2004. In: P. Merilä, T. Kilponen & J. Derome (Eds.), Forest condition monitoring in Finland—National report 2002–2005. Working Papers of the Finnish Forest Res. Inst. (vol. 45, pp. 81–92). Lorenz, M. (1993). Die Europäische Waldzustandserfassung. Zeitschrift für Ökologie und Naturschultz, 2, 245–251. Lorenz, M., Fischer, R., Becher, G., Granke, O., Riedel, T., Roskams, P., et al. (2007). Forest condition in Europe. 2007. Technical Report. BFH, Hamburg, (91 pp.), Annexes. Lorenz, M., Fischer, R., Becher, G., Granke, O., Seidling, W., Ferretti, M., et al. (2008). Forest condition in Europe. 2008. Technical reports of ICP Forests. Work report of the Institute for World Forestry 2008/1, (107 pp.), Annexes. Manion, P. D. (1981). Tree disease concepts (pp. 399). Englewood Cliffs: Prentice-Hall. Mälkönen, E., Helmisaari, H.-S., Lindgren, M., & Raitio, H. (2000). Forest condition in Finland—Concluding remarks. In: E. Mälkönen (Ed.) Forest condition in a changing environment—The Finnish case. Forestry sciences (vol. 65, pp. 361–367). Dordrecht: Kluwer Academic. Nevalainen, S. (1999). Nationwide forest damage surveys in Finland. In: B. Forster, M. Knizek & W. Grodzki (Eds.), Methodology of forest insect and disease survey in Central Europe. Second Workshop of the IUFRO Working Party 7.03.10 (pp. 24–29). Sion-Chateauneuf, Switzerland: Swiss Federal Institute for Forest, Snow and Landscape Research. Nevalainen, S., & Heinonen, J. (2000). Dynamics of defoliation, biotic and abiotic damage during 1986–1998. In: E. Mälkönen (Ed.), Forest condition in a changing environment—The Finnish case. Forestry sciences (pp. 133–141). Dordrecht: Kluwer Academic. Nevalainen, S., Lindgren, M., & Pouttu, A. (2007). Biotic and abiotic damage on the Level I network. In: P. Merilä, T. Kilponen & J. Derome (Eds.), Forest condition monitoring in Finland—National Report 2002– 2005. Working Papers of the Finnish Forest Research Institute (vol. 45, pp. 32–40). Peltola, A. (2005). Finnish statistical yearbook of forestry 2005. SVT agriculture, forestry and fishery 2005. Metsäntutkimuslaitos, 45, (424 pp.). Prinz, B. (1985). Symptomatik und mögliche Ursachen der Waldschäden. Forschungsergebisse zur Prob-
Environ Monit Assess (2010) 168:159–171 lematik der Neuartigen Waldschäden. LIS-Berichte, 57, 7–25. Redfern, D., & Hendry, S. (2002). Climate change and damage to trees caused by extremes of temperature. Forestry Commission Bulletin, 125, 29–39. Rouault, G., Candeau, J. N., Lieutuer, F., Martin, J. C., Grégoire, J. C., Nageleisen, L. M., et al. (2006). Effect of drought and heat on forest insect populations in relation to the 2003 drought in Western Europe. Annals of Forest Science, 63, 611–622. Salemaa, M., Jukola-Sulonen, E.-L., & Lindgren, M. (1991). Forest condition in Finland, 1986–1990. Silva Fennica, 25, 147–175. Schelhaas, M. J., Nabuurs, G. J., & Schuck, A. (2003). Natural disturbances in the European forests in the 19th and 20th centuries. Global Change Biology, 9, 1620– 1633. Schulz, M., Cachier, H., Ebinghaus, R., Ferm, M., Hongisto, M., Iverfeldt, A., et al. (1999). Evolution of the aerosol composition in the BASYS Network study and Lagrangian experiments in summer 1997 and winter 1998. Journal of the Aerosol Science, 30, S97–S98. Searle, S. R., Casella, G., & Mcculloch, C. E. (1992). Variance components (pp. 501). New York: Wiley. Seidling, W., & Mues, V. (2005). Statistical and geostatistical modelling of preliminarily adjusted defoliation on an European scale. Environmental Monitoring and Assessment, 101, 223–247. Sioen, G., & Roskams, P. (Eds.) (2006). Crown condition of Quercus robur in Flanders (Belgium). Symposium: Forests in a changin g Environment—Results of 20 years ICP Forests Monitoring, Göttingen, 25.-28.10. 2006. J.D.Sauer’s Verlag. Schriften aus der Forstlichen Fakultät der Universität Göttingen und der Nordwestdeutschen Forstlichen Versuchsanstalt 142, (210– 215 pp.). Skelly, J. M., & Innes, J. L. (1994). Waldsterben in the forests of Central Europe and eastern North America: Fantasy or reality? Plant Disease, 78, 1021–1032.
171 Sofiev, M., Petersen, G., Krüger, O., Schneider, B., Hongisto, M., & Jylhä, K. (2001). Model simulations of the atmospheric trace metals, concentrations and depositions over the Baltic Sea. Atmospheric Environment, 35, 1395–1409. Thomsen, M. G., & Nellemann, C. (1994). Isolation of natural factors affecting crown density and crown color in coniferous forest: Implications for monitoring in forest decline. Ambio, 23, 251–254. Tuovinen, J. P., & Simpson, D. (2008). An aerodynamic correction for the European ozone risk assessment methodology. Atmospheric Environment, 42, 8371– 8381. UN/ECE (1998). Manual on methodologies and criteria for harmonized sampling, assessment, monitoring and analysis of the effects of air pollution on forest (4th ed.). Hamburg/Geneva: Programme Coordinating Centres. UN/ECE (2004). Manual on methodologies and criteria for modelling and mapping critical loads and levels and air pollution effects, risks and trends. Retrieved from http://www.icpmapping.org. Varama, M., & Niemelä, P. (2001). Männiköiden neulastuholaiset (the defoliators in Scots pine forests, in Finnish). Metsätieteen aikakauskirja—Folia Forestalia 2/2001, 275–279. Wermelinger, B. (2004). Ecology and management of the spruce bark beetle Ips typographus—A review of recent research. Forest Ecology and Management, 202, 67–82. Wulff, S., Hansson, P., & Witzell, J. (2006). The applicability of national forest inventories for estimating forest damage outbreaks—Experiences from a Gremmeniella outbreak in Sweden. Canadian Journal of Forest Research, 36, 2605–2613. Zlatev, Z., Bergströn, R., Brandt, J., Hongisto, M., Jonson, J. E., Lagner, J., et al. (2001). Studying sensitivity of air pollution levels caused by variations of different key parameters. TemaNord, 2001, 569.
