Design concepts adopted in long-term forest monitoring programs in Europe—problems for the future?
Introduction
It is widely acknowledged that detection, assessment and monitoring of the effects of environmental changes on natural resources, (e.g. climatic changes, biodiversity loss and air pollution) require long-term integrated programs (Bricker and Ruggiero, 1998, Urquhart et al., 1998, Bull et al., submitted). Once the long-term and integrated features are met, the success of the program is often taken for granted. This belief seems to underlie some important and expensive monitoring programs, like Pan-European intensive monitoring of forest ecosystems (henceforth referred to as Level II program), (e.g. EC and UN/ECE, 2001), with relatively little attention paid to the development of robust design concepts. Unfortunately, long-term commitment to integrated monitoring is necessary but not sufficient to ensure the success of the program, as many other factors and design issues need to be considered, (e.g. Parr et al., 2002, Vos et al., 2000, Olsen et al., 1999).
This paper emphasizes the need for a formal statistical approach in order to provide reliable data to support policy decision making and resource management in a rigorous and defensible way (Urquhart et al., 1998). Although international long-term monitoring programs may always encounter difficulties in ensuring consistency and full harmonization of methods, in the case of Level II program we sustain that basic design issues were not satisfactorily addressed. We argue that this is in itself a likely source of problems for data analysis and may jeopardize future assessment of the program's result. To support our opinion, we concentrate on three major design issues as covered in the Level II program, (e.g. EC and UN/ECE, 2001 and the other reports of the same series): (i) choice of measurements, indicators and indices of ecosystem status and changes (relevance and priority); (ii) sampling strategy (selection of sites in which to perform measurements: preferential, model-based, design-based); and (iii) the sampling tactic (where to locate measurements within the site).
Section snippets
Measurements and indicators
Table 1 shows the measurements and indicator categories used in the Level II program. The indicators are classified according to Vos et al. (2000) in terms of explanatory (independent) and response (dependent) variables, with some variables (termed intermediate) playing both roles, (e.g. soil chemistry). For each indicator, an evaluation is given with respect to priority (based on the mandatory/optional criteria adopted by the EC regulations and on the coverage in terms of number of plots,
Sampling strategy
Although more than 800 Level II monitoring sites are installed and millions of euros are being spent for the program, it is important to recognize and admit that this huge expenditure of resources is incapable of providing quantitative estimates valid on national, subregional or regional (European) scales. In other words, although frequently reported, statements and data about ‘average bulk deposition fluxes as a function of geographic regions’ (EC and UN/ECE, 2000, p. 56, table 5.5) are not
Sampling tactic
The main potential of the Level II program is insights into forest condition, because research at plot level enables various multivariate studies. For such studies, plot-level statistical descriptors are used, and the validity of the results depends on having good estimates of the attributes of interest at plot level. These estimates need to be sufficiently precise and unbiased. While precision is approximately the function of the number of observations, bias is determined more by the way
Conclusions
Long-term integrated monitoring in Europe is an unprecedented chance for forest ecologists, researchers, resource managers and policy makers to collect and evaluate data for implementing sound environmental policy. However, despite the considerable effort that has gone into installing and implementing the program, certain problems have not been sufficiently considered. Many of these problems regard the design basis of the program, and include indicator development as well as sampling strategy
References (40)
Potential and limitations of visual indices of tree condition
Chemosphere
(1998)- et al.
Air pollution and environmental chemistry—What role for tree rings studies?
Dendrochronologia
(2002) Implementation of a national monitoring program
J Environm Manag
(1994)- et al.
Vegetation monitoring in South-Varanger, Norway. Species composition of ground vegetation and its relation to environmental variables and pollution impact
Environ Monit Assess
(1999) Species and area
J Ecol
(1921)- BFH. Manual on methods and criteria for harmonized sampling, assessment, monitoring and analysis of the effects of air...
- et al.
Toward a national program for monitoring environmental resources
Ecol Appl
(1998) - Bull, K.R., Achermann, B., Bashkin, V., Chrast, R., Fenech, G., Forsius, M., Gregor H.D., Guardans, R., Haußmann, T.,...
Sampling techniques
(1977)- et al.
The statistics and biology of the species-area relationships
Am Nat
(1979)
Designing environmental field studies
Ecol Monogr
Short-term changes of response indicators of ecosystem status in broadleaved forests in Tuscany (Central Italy)
Water Air Soil Pollut
On the relation between species and area
Ecology
Quantifying biodiversity: procedures and pitfalls in the measurement and comparison of species richness
Ecol Lett
Theoretical and practical criteria for the selection of ecosystem monitoring plots in Swiss forests
Environ Monit Assess
Cited by (47)
Forest monitoring: Substantiating cause-effect relationships
2019, Science of the Total EnvironmentAnticipating species distributions: Handling sampling effort bias under a Bayesian framework
2017, Science of the Total EnvironmentCitation Excerpt :Given the massive negative economic and ecological effects of invasive species, a robust method for predicting species' distributions is crucial for an early assessment of species invasions and effective application of appropriate management actions (Malanson and Walsh, 2013). Investigating how biodiversity is distributed spatially and temporally across the globe has long been a central theme in ecology (Gaston, 2000) and the methods developed to answer this question have become key tools for biodiversity monitoring (Ferretti and Chiarucci, 2003; Chiarucci et al., 2011). For example, species distribution models (SDMs) have been used to map the current distribution of a single species (Rocchini et al., 2011), model the potential distribution of native and invasive species (Rocchini et al., 2015), investigate the statistical performance of different models to infer the distribution of species under various ecological conditions (Elith and Graham, 2009; Guisan and Zimmermann, 2000), test the transferability in space of modeled distribution patterns (Heikkinen et al., 2012; Randin et al., 2006), predict long term changes to species distributions (Pearman et al., 2008) and make inferences on future biodiversity scenarios (Engler et al., 2009; Pompe et al., 2008), evaluate the potential of satellite imagery bands as predictors of biodiversity patterns (Mathys et al., 2009), analyse spatial autocorrelation in species distributions (Carl and Kühn, 2007; Dormann, 2007), and understand biogeographical patterns (Sax, 2001).
Future impacts of nitrogen deposition and climate change scenarios on forest crown defoliation
2014, Environmental PollutionCitation Excerpt :Modelling of crown defoliation is challenging at the European scale because the tree defoliation integrates a wide range of mutually interacting predictors (de Vries et al., 2000), which, in turn, are spatially varying. The use of crown defoliation as an indicator of forest health has been subjected to much criticism (Innes, 1993; Ferretti and Chiarucci, 2003) and to a high degree of uncertainty due to observer bias and possible methodological differences between the participating country (Ferretti et al., 2007; Eichorn et al., 2010). However, ICP Forests has invested great efforts into harmonization which has led to various manual for the data assessment, training inter-comparison exercises and quality control procedures.
Plot-scale modelling to detect size, extent, and correlates of changes in tree defoliation in French high forests
2014, Forest Ecology and ManagementCalculating landscape diversity with information-theory based indices: A GRASS GIS solution
2013, Ecological Informatics
- 1
Tel.: +39-0577-232872; fax: +39-0577-232860; [email protected]