Predicting the scale of landscape context effects: How big is a landscape?

点击翻译按钮获取中文摘要

BACKGROUND: Arguably the most important spatial scaling problem in conservation is identification of the distance within which the landscape context affects a given population or community. We know that no pond, forest patch, or protected area is an island. Ecological variables such as population abundance or species richness are strongly influenced by the landscape context, e.g. the surrounding road density, forest cover, or urbanization. For effective conservation action we need to know the distance within which these effects occur, i.e. the ‘scale of effect’ of the landscape. Predicting the scale of effect is also critical to landscape ecological research. If landscape variables are measured at the wrong extent, effects of landscape context can go undetected, and/or relative importance of landscape effects will be incorrectly estimated. Nevertheless, many researchers arbitrarily select a single spatial scale for analysis and others scale the landscape to the movement biology of the organism(s), on the reasonable assumption that the scale of effect depends on the spatial extent over which the organism interacts with its environment. Tests of this assumption so far have been mixed, indicating that other factors come into play. For example, how is the scale of effect influenced by demographic parameters (e.g., reproductive rate), by details of a species’ movement behaviour, and by species interactions? How does the scale change, depending on the landscape predictor (road density, forest cover etc.)? And, does it depend on the response measured, e.g., abundance, species richness? There is currently no comprehensive theory for predicting the scale of effect of the landscape context on population and community responses.OBJECTIVE: Over the next 5 years I aim to build and validate a set of hypotheses for predicting the scale of effect of the landscape context on population and community responses. APPROACH: I propose eight hypotheses, suggesting relationships between the scale of effect and (i) species traits (dispersal, reproduction rate, resource use, traits of natural enemies), (ii) response variables (abundance, occurrence, species richness), and (iii) type of landscape variable (acting through resource availability e.g. habitat amount, or through mortality e.g. road density). We will use individual-based spatial simulation modelling to evaluate the hypotheses and to estimate relative importance of factors influencing the scale of effect. We will then test the hypotheses by first (i) estimating the scale of effect of landscape predictors on multiple individual species and community responses, using multi-scale empirical studies, and then (ii) conducting cross-species (or species group) comparisons to test whether the observed scale of effect changes as predicted from the hypotheses and simulations. Since my goal is to develop a general, validated theory for predicting the scale of effect, we will conduct the empirical tests for several different species groups, including: forest small mammals, pond and forest dwelling amphibians, bark and wood boring beetles, forest floor plants, and alvar plants.SIGNIFICANCE: We know that selecting the ‘right’ spatial scale is hugely important in landscape and conservation ecology, but we generally ‘wing it’ because we do not know the right scale a priori. The proposed research will build validated theory that will allow landscape and conservation ecologists to select the right spatial scale for analyses and application. This will (i) increase inferential strength in landscape ecology studies and (ii) greatly increase effectiveness of landscape-scale conservation actions, ensuring that they are implemented at the correct spatial extent for best success in effecting conservation goals.