How Does Habitat Patch Size Affect Animal Movement? An Experiment With Darkling Beetles

Ecology, Oct, 1999 by Nancy E. McIntyre, John A. Wiens

INTRODUCTION

Environments are heterogeneous in space and time. The patterning of this heterogeneity affects the abundance and distribution of organisms and the array of population, community, and ecosystem patterns that follow from distribution and abundance (Robinson et al. 1992). In a sense, these two statements embody a major focus of landscape ecology (Turner 1989, Ims 1995, Pickett and Cadenasso 1995, Wiens 1995).

A "landscape" may be both an ecological entity and a human construct (Pickett and Cadenasso 1995), viewed in one of (at least) three ways: (1) as a large stretch of land that could potentially contain several local populations of an organism (the traditional definition; Forman and Godron 1986, Forman 1995); (2) as a level in a hierarchy of ecological organization between ecosystem and biosphere (Lidicker 1988, Allen and Hoekstra 1992); and (3) as a template of any size upon which heterogeneity exists. Here, we adopt a "size doesn't matter" stance (With 1994, Burke 1997) and follow the third view of landscape.

Within any landscape, the behaviors of organisms will be influenced by both the heterogeneity or spatial patterning of the mosaic and the scale on which landscape pattern is perceived (Kotliar and Wiens 1990, With 1994, Bissonette et al. 1997). The finest resolution of scale that is perceived by an organism ("grain") is contrasted with the broadest perceptual resolution of scale ("extent") (Turner 1989). Because spatial patterns change with changes in scale, issues of scale are central to ecology and landscape ecology (Meentemeyer and Box 1987, Wiens 1989, Levin 1992, 1993, Dale et al. 1994, Bissonette 1997). Understanding how landscape pattern affects the abundance and distribution of organisms therefore requires an understanding of scale dependency. Given the complexity of organism - environment interactions, achieving such an understanding will require integrating information from both theoretical and empirical studies (Weiner 1995). Because we do not yet have anything resembling a "theory of scaling" (Meentemeyer and Box 1987), however, we must rely on empirical studies to provide baseline information about scaling effects in ecology (Wiens 1995, Bowers 1997, Steinberg and Kareiva 1997).

The linkage between animal behavior and landscape pattern is particularly amenable to empirical analysis. Certain behaviors are strongly influenced by landscape pattern (Wiens and Milne 1989, Crist et al. 1992, With 1994, Ims 1995, Cresswell 1997, McIntyre 1997). Movement behaviors, in particular, represent a record of how an organism responds to spatial pattern (Levin et al. 1984, With 1994). Movement patterns, in turn, may affect the genetic and demographic composition of populations (Levins 1969, Cohen and Levin 1991, McCauley 1995, Cresswell 1997), the spread of diseases and parasites (Holmes 1993), or energy flow and nutrient transfer (Elmes 1991). Thus, understanding how structural features of the landscape influence animal movements may be a key component in comprehending population, community, and ecosystem composition and functioning (Forman 1995, Wiens 1995, 1996). For these reasons, developing an empirical understanding of how animal movement patterns are affected by landscape pattern has been promoted as a research priority for landscape ecology (Wiens 1989, 1990, Kareiva 1990, Ims 1995).

Empirical studies of the relationship between movement and the scale of landscape pattern, however, are scarce. Here, we report the results of an experimental study that was explicitly designed to assess the effects of the grain of landscape pattern on animal movements. We used an experimental model system (EMS; Wiens and Milne 1989, Ims et al. 1993, Wiens et al. 1993b) consisting of darkling beetles (Coleoptera: Tenebrionidae, Eleodes obsoleta Say) moving in microlandscapes in which the proportion of different landscape elements was held constant, but the grain of the pattern was varied. Using an EMS design permits manipulation that would be logistically unfeasible at larger scales. Other studies, particularly in agroecosystems, have documented the relationships between insect movement and patch use or landscape structure (e.g., Kareiva 1985, 1987, Wallin and Ekbom 1988, Wratten and Thomas 1990, Bell 1991, Vermuelen 1994, Riecken and Raths 1996). In this study, however, we have manipulated the grain of habitat patchiness (patch size) while keeping the overall amount of habitat constant, and have examined how these factors influence individual behaviors.

As the scale of patchiness becomes broader (i.e., as patches become larger), the ratio of patch perimeter to patch area decreases, resulting in fewer patch boundary zones. Because beetle movements differ in landscapes containing different habitat types (e.g., grass and sand; Wiens et al. 1985, Wiens and Milne 1989, Crist et al. 1992, Wiens et al. 1997), they are likely to be affected by patch boundaries. From this, we predict that movements should be more constrained and localized in fine-grained landscapes. This is because the frequent boundaries between patches will interrupt the flow of movements in a particular habitat type and cause repeated transitions among movement patterns. In a fine-grained mosaic, an individual's movement behavior must be continually readjusted. Percolation models also predict a nonlinear relationship between habitat patch size and movement patterns (Wiens et al. 1997, With et al. 1997, With and King 1997). Because our experimental design incorporated patches of preferred beetle habitat (grass) embedded in a matrix of less suitable habitat (sand), the results also bear on the issue of how the scale of habitat fragmentation might affect animal movements, compared to the null hypothesis that movements in heterogeneous mosaics are identical to those in homogeneous landscapes (Merriam 1995).