Search

Land conversion due to human activities produces distinctive spatialpatterns across the landscape. It remains unclear, however, how particular spatial arrangements of remnant habitat patches influence species persistence. We present a conceptual model of landscape change that focuses explicitly on habitat spatial arrangement. Four sequences,characterized as shrinkage, bisection, fragmentation, and perforation, differ qualitatively in the spatial arrangement of intact habitat and differ quantitatively in boundary length, amount of interior area, and connectivity. We tested the predictions of this conversion sequence model in a "microlandscape" field experiment with insects in a native grassland habitat. The four conversion sequences significantly influenced the species composition of the above-ground insect community, and insect species groups varied in their responses. Four patterns were evident from the field experiment: (1) as habitat decreased, insect density rose sharply on small dispersed patches of the bisection and fragmentation sequences, an effect less evident in the large remnant patch of the shrinkage sequence; (2) species richness declined only in the case of shrinkage, whereas richness increased in the fragmentation sequence; (3) most individual species examined responded similarly to the community as a whole, and (4) large rare species were concentrated in remaining habitat of the shrinkage and fragmentation sequences. Overall, the shrinkage and perforation sequences, which had high connectivity, were similar to one another in species composition, while the bisection and fragmentation sequences, which had low connectivity, were similar to one another. Based on these results, we conclude that a spatial pattern of land conversion that maintains one or more large, closely-spaced remnant patches of native vegetation isessential for species persistence. [ABSTRACT FROM AUTHOR]

The purpose of this data set was to investigate the effects of habitat characteristics and urbanization on grassland butterfly diversity. We sampled butterflies at 66 sites in Boulder County Open Space in the years 1999 and 2000. We tested for effects of habitat characteristics (grassland type and quality) and landscape context (percentage surrounding urbanization) on butterfly diversity and abundance. Although both habitat characteristics did affect butterfly diversity, landscape context did not. The data contain butterfly species diversity and individual species' abundance for the five periods sampled: July 1999, August 1999, June 2000, July 2000, and August 2000. Our samples included 58 species from five butterfly families: Hesperiidae (17 species), Papilionidae (4 species), Pieridae (7 species), Lycaenidae (II species), and Nymphalidae (19 species). These data include endangered butterfly species, as well as surveys from relictual tallgrass prairie. There is some limitation to these data, as our sampling was not exhaustive through the year. That is, many species we observed also occur outside of the period sampled, and some butterfly species may not have been present as adults during the months of June, July, and August, and thus were not sampled. We present these data with the hope that they be combined with future surveys using the methods described here to constitute a long-term data set. These data will undoubtedly be useful in understanding temporal changes in butterfly abundance and diversity, effects of urbanization on different taxa, and determinants of local and large-scale faunal diversity. [ABSTRACT FROM AUTHOR]