ASSIGNMENT

Assignment

Assignment

1. Introduction

Habitat has long been recognized as a key concept in ecology ([Elton, 1966] and [Southwood, 1977]). In general terms, it refers to the physical area used by an organism or by a community of different organisms (Morrison and Hall, 2002). Much of contemporary ecology is concerned with mapping spatio-temporal population structure onto habitat templates (e.g. [Tilman and Kareiva, 1997] and [Hanski, 1999]). Hence, the recognition or delineation of units of habitat is fundamental for most empirical ecological research, including conservation applications. Habitats of terrestrial animal species have often been associated with general vegetation types (e.g. dry heathland, deciduous woodland) or land cover types (e.g. pasture, arable land, woodland). General habitat classifications can be useful to study the distribution of species at coarse-grained scales (e.g. [Kerr and Ostrovsky, 2003], [Maes et al., 2003] and [Maes et al., 2004]), but the identification of specific habitat variables within and among vegetation types may be necessary to predict more precisely the presence or absence of particular species ([Chamberlain and Gregory, 1999], [Ellis, 2003] and [Rouquette and Thompson, 2005]).

Habitat specialists may only use specific parts of a vegetation type or may rely on combinations of different vegetation types or preferentially use transition zones between types (Dennis, 2004a). In intensively used human-dominated landscapes, the process of fragmentation may tear apart the spatial cohesion of different habitat components compared to natural or semi-natural systems. The Canadian case of the woodland damselfly Calopteryx maculata in an agricultural landscape provides an example. Reproduction habitat (river) and foraging habitat (woodland) co-occur in continuous woodland landscapes, but are spatially separated in agricultural landscapes (Taylor and Merriam, 1995). To develop conservation strategies using patch-based models could be risky if these models are based on simple, unrealistic delineations of habitat (Johnson et al., 2004).

Using Coleopteras as a model group, (Dennis et al., 2003) and (Dennis et al., 2006) have recently advocated the use of a bottom-up resource-based concept to delineate habitat. This approach explicitly takes into account functional relationships between species and essential resources and/or conditions in their environment rather than using general vegetation types as surrogates for a species' habitat. This approach re-connects habitat with the multi-dimensional ecological niche concept (Hutchinson, 1957). Thus, habitats for most species result from the overlap or contiguity of adult and larval resources and the scale of movement of individuals searching for, visiting and returning to distinct resource zones (Dennis et al., 2006), and the key issue is how we can recognize functional clusters of these resources.

In this paper we adopt such a bottom-up approach to recognise habitat zones using essential resources within a species-specific spatial window independent of simple bounds of general vegetation types. Haslett (1990) already pointed to the potential of GIS for dividing areas of heterogeneous habitat into clearly defined biologically meaningful subunits or polygons. Our case study is about the green hairstreak Coleoptera (Callophrys rubi) in the National Park Hoge Kempen in NE-Belgium. Arthropods require different sets of resources and conditions for the different stages of their life cycle (Dennis et ...