Landscape Ecol (2011) 26:1361–1372 DOI 10.1007/s10980-011-9656-5 RESEARCH ARTICLE Large-scale experimental landscapes reveal distinctive effects of patch shape and connectivity on arthropod communities John L. Orrock • Gregory R. Curler • Brent J. Danielson • David R. Coyle Received: 26 October 2010 / Accepted: 2 September 2011 / Published online: 14 September 2011 Ó Springer Science+Business Media B.V. 2011 Abstract The size, shape, and isolation of habitat nectivity (via habitat corridors) independently of area patches can affect organism behavior and population and edge effects. We found that patch shape, rather dynamics, but little is known about the relative role of than connectivity, affected ground-dwelling arthropod shape and connectivity in affecting ecological com- richness and beta diversity (i.e. turnover of genera munities at large spatial scales. Using six sampling among patches). Arthropod communities contained sessions from July 2001 until August 2002, we fewer genera and exhibited less turnover in high-edge collected 33,685 arthropods throughout seven 12-ha connected and high-edge unconnected patches relative experimental landscapes consisting of clear-cut to low-edge unconnected patches of similar area. patches surrounded by a matrix of mature pine forest. Connectivity, rather than patch shape, affected the Patches were explicitly designed to manipulate con- evenness of ground-dwelling arthropod communities; regardless of patch shape, high-edge connected patches had lower evenness than low- or high-edge unconnected patches. Among the most abundant arthropod orders, increased richness in low-edge unconnected patches was largely due to increased Electronic supplementary material The online version of richness of Coleoptera, whereas Hymenoptera played this article (doi:10.1007/s10980-011-9656-5) contains an important role in the lower evenness in connected supplementary material, which is available to authorized users. patches and patterns of turnover. These findings J. L. Orrock (&) suggest that anthropogenic habitat alteration can have Department of Zoology, University of Wisconsin, distinct effects on ground-dwelling arthropod com- Madison, WI 53706, USA munities that arise due to changes in shape and e-mail: [email protected] connectivity. Moreover, this work suggests that cor- G. R. Curler ridors, which are common conservation tools that Department of Entomology and Plant Pathology, change both patch shape and connectivity, can have University of Tennessee, Knoxville, TN 37996, USA multiple effects on arthropod communities via differ- B. J. Danielson ent mechanisms, and each effect may alter compo- Department of Ecology, Evolution, and Organismal nents of community structure. Biology, Iowa State University, Ames, IA 50010, USA Keywords Arthropods Á Coleoptera Á Corridor Á D. R. Coyle D. B. Warnell School of Forestry and Natural Resources, Diversity Á Evenness Á Fragmentation Á Hymenoptera Á University of Georgia, Athens, GA 30602, USA Orthoptera Á Patch shape 123 1362 Landscape Ecol (2011) 26:1361–1372 Introduction 2001; Hunter 2002; Steffan-Dewenter and Tscharntke 2002; Ewers and Didham 2008), highlight the impor- The size, shape, and degree of isolation (i.e. connec- tance of dissecting the potential interactions between tivity) of habitat patches can have important effects on patch shape and connectivity in affecting arthropod the local abundance of organisms (Dunning et al. communities. 1992; Taylor et al. 1993; Collinge and Palmer 2002), Shape and connectivity may each have different interactions among species (Fagan et al. 1999), and the effects on richness, evenness, and beta diversity. Patch structure and composition of ecological communities shape determines the relative relationship between (Turner 1989; Forman 1995). Understanding the role patch perimeter and area, and is thus expected to of patch shape and connectivity on ecological com- mediate the degree to which edge effects alter within- munities is important because anthropogenic habitat patch processes (Fagan et al. 1999; Ries et al. 2004; destruction and alteration change the size, shape and Fletcher et al. 2007). Fletcher et al. (2007) note that connectivity of patches in the landscape (Forman edges act via two mechanisms: changing abiotic and 1995; Dobson et al. 1997; Vitousek et al. 1997; Fischer biotic flows, and by providing organisms with access and Lindenmayer 2007). Although habitat fragmenta- to different resources. Abiotic flows consist of changes tion has documented effects on communities (Harrison in temperature, moisture, and other environmental and Bruna 1999; Fahrig 2003; Fischer and Lindenma- characteristics associated with edges (Chen et al. yer 2007), studies that independently manipulate 1999). Biotic flows include the movement or dispersal shape and connectivity are rarely conducted at large of species across habitat edges, potentially giving rise scales (Harrison and Bruna 1999; Debinski and Holt to ‘‘spillover’’ across habitat boundaries (e.g., Brudvig 2000; Steffan-Dewenter and Tscharntke 2002;O¨ ck- et al. 2009). Biotic flows may also occur due to inger and Smith 2008; Haddad et al. 2011). The lack of foraging movement. For instance, species richness large-scale studies is particularly noteworthy because may also be increased in high-edge patches if organ- patch size, patch shape, and connectivity are of isms are able to gain access to important resources by explicit concern for conservation strategies based on exploiting edge habitats (Dunning et al. 1992; Fletcher corridors (Haddad et al. 2003; Hilty et al. 2006), and et al. 2007). Patch shape can affect immigration for the design of conservation reserves (Janzen 1983). because the amount of patch edge may determine the In this paper, we present a landscape-level exper- effective area available to intercept organisms moving iment that independently manipulates patch shape and through the matrix, i.e. drift-fence effects (Forman connectivity of patches of similar size. Specifically, 1995; Haddad and Baum 1999; Fried et al. 2005). we examine the richness, evenness (i.e. relative equity Patch shape may similarly affect emigration by in species abundance in a community), and turnover of providing a greater amount of area for patch-dwelling genera in space (i.e. beta diversity) of ground-dwelling organisms to enter matrix habitat (Grez and Prado arthropod communities. Although connectivity may 2000; Collinge and Palmer 2002). be important in arthropod communities (Gilbert et al. Based on predictions from island biogeography 1998; Gonzalez et al. 1998), many studies are (MacArthur and Wilson 1967), increased connectivity conducted at relatively small spatial scales (reviewed is expected to increase richness by increasing rates of in Steffan-Dewenter and Tscharntke 2002). Insight colonization from suitable patches. Evenness may also from a few large-scale studies suggests that the effect be increased in more connected patches because of connectivity depends on the size of the patches increased input of species should reduce the likelihood being connected, temporal variation in abiotic condi- that a few species will become numerically dominant. tions, and the species of interest (Collinge 2000; However, these two predictions assume that the O¨ ckinger and Smith 2008). These differing outcomes classical competition-colonization tradeoff is not in of corridor experiments, coupled with the documented operation (Amarasekare et al. 2004). In systems where effects of shape on arthropod communities (Grez and increased connectivity allows superior competitors to Prado 2000; Collinge and Palmer 2002) and the colonize patches where inferior competitors would significant effects of fragmentation on arthropod otherwise persist, richness and evenness would be communities (e.g., Davies and Margules 1998; Did- expected to exhibit a negative relationship with ham et al. 1998; Golden and Crist 2000; Davies et al. connectivity (although this relationship could be 123 Landscape Ecol (2011) 26:1361–1372 1363 unimodal or nonlinear if there are thresholds in patch types provides insight into the relative impor- connectivity required for the persistence of inferior tance of shape and connectivity. When patch shape is competitors). Because well-connected communities most important, low-edge unconnected rectangular are expected to exchange greater numbers of migrants, patches will differ from high-edge unconnected they should also be more similar in their composition. patches, and we expect greater richness and increased As such, increased connectivity should reduce species turnover in high-edge patches because of immigration turnover (i.e. beta diversity). into these patches from forest- and edge-dwelling taxa. We used large-scale experimental landscapes to When connectivity is most important, connected examine whether ground-dwelling arthropod commu- patches will differ from high-edge unconnected nities are affected by patch shape, patch connectivity, winged patches (Damschen et al. 2008) but the nature or both. The seven experimental landscapes (Fig. 1) of this difference depends upon whether a competi- used three patch types of similar area that differed in tion-colonization tradeoff is important. If the tradeoff shape and connectivity: (1) high-edge connected is operating, we predict that connected patches would patches; (2) high-edge unconnected ‘‘winged’’ patches have lower richness and evenness. If a competition- with similar amounts of edge as connected patches but colonization tradeoff