Diversity and conservation of insects in urban gardens: Theoretical and applied implications

Urban community gardens located in the Bronx and East Harlem in New York City were sampled for insects from 2003 through 2006. During this time, 73 insect taxa (predominantly insect families), 24 butterfly species and 54 bee species were identified in 19 gardens. Insect richness was most influenced by local variables such as garden area and floral area. In addition, for all three measures of insect richness there was a consistent but weak effect of the proportion of green space in the surrounding landscape. A manipulative experiment was carried out whereby ∼70 native wildflowers were added to a subset of urban gardens. Perhaps due to the limited size of the wildflowers additions, there was no effect on bee or butterfly abundance or richness. In addition, although some butterflies nectared from the added native wildflowers, many others predominantly utilized exotic (non-native) ornamental flowers already present in gardens. Similarly, bee abundance in urban gardens was greater on non-native crops relative to native wildflowers. In a mark-recapture experiment, the majority of P. rapae butterflies were found to remain in gardens less than two days and four individuals moved a mean distance of 1033m across heavily developed roadways. Mean recruitment to gardens was two butterflies per sampling date and recruitment was positively correlated with the floral abundance of urban gardens but not the proportion of surrounding green space. In a separate mark-recapture experiment of Bombus impatiens workers, many marked individuals were resighted (44%) although no movements were detected among gardens, suggesting site fidelity of workers to florally rich urban habitat patches. Recruitment of B. impatiens was positively correlated with the proportion of surrounding green space perhaps due to more nesting sites for colonies. After taking into account detection probability, several butterfly and bee species were predicted to use all urban garden sites, highlighting the spatial scope of some urban insect populations. For butterflies, the probability of encountering most species on any sampling date was low (\u3c0.15), indicating that urban gardens are utilized sporadically throughout a season. Across years, random site use was observed for most species, indicating fluctuations in insect presence in urban gardens