Distribution and dispersal of species in natural and human-dominated landscapes (Lepidoptera)

The four chapters that compose my dissertation address two ecological issues of importance to biodiversity conservation. The first issue (Chapters 1 and 2) concerns the ecological effects of anthropogenic habitat fragmentation—with a specific focus on the “matrix” (i.e., the human-dominated areas surrounding native fragments). In the first chapter, I sampled moth species richness in the agricultural matrix surrounding forest fragments in southern Costa Rica. I found that agricultural type did not affect moth species richness or composition; however, sites near (<1 km) a large forest fragment were significantly higher in species richness and more similar to forest sites in species composition than sites far (>3.5 km) from the fragment. In the second chapter, I tested whether two distinct matrix types (i.e., conifer forest and willow thicket) differed in their resistances to butterfly movement between patches of meadow habitat in the Rocky Mountains. For the majority of butterflies, conifer was 3–12 times more resistant than willow. These two studies suggest that the surrounding matrix may support substantial native diversity and may influence the “effective isolation” of habitat remnants in fragmented landscapes.; The second issue (Chapters 3 and 4) concerns the utility of indicator taxa as surrogate measures of overall diversity. Well-known indicator taxa often are proposed as efficient ways of identifying conservation priorities, but the correlation between putative indicators and other taxa has been largely untested. In the third chapter, I tested nine taxa as indicators of overall species richness across North America. I found that the three most practical choices for indicator taxa (mammals, birds, and butterflies) are not the most informative indicators of overall species richness, when employed either individually or in combination. In the fourth chapter, I examined whether butterflies were a useful indicator of diversity in a closely-related but relatively poorly-known group, the moths, at a local landscape scale. I found no correlation in diversity between the two taxa. The findings of these two studies suggest that well-known taxa are often not informative surrogate measures of diversity in other groups and that phylogenetic and ecological relatedness are not reliable criteria for selecting appropriate indicator taxa.