Plant diversity and vegetation structure as factors correlated with parasitoid wasp (Hymenoptera: Braconidae and Ichneumonidae) diversity in a degraded dry forest ecosystem in Puerto Rico

Abstract

Tropical dry forests worldwide have a long history of human use, invasion by non-native species and degradation. In Puerto Rico, while overall forest cover has increased due to socioeconomic and land use changes over the past decades, recovery of tropical dry forest areas has been slower. Within the Laguna Cartagena National Wildlife Refuge, a site of critical habitat for migrating waterfowl, recovery of forest cover following agricultural abandonment has occurred both through natural secondary succession (primarily non-native trees) and active restoration (primarily native trees). The next step in this restoration process is the evaluation of the implemented strategies. For this reason, the aim of my research was to evaluate the relationship between the diversity of two families of parasitoid wasps (Braconidae and Ichneumonidae, which are insects that might be sensitive to changes in habitat) with the diversity of plants and plant architecture in three cover types: native trees, non-native trees and grasslands without trees. Three plots were established in each of the three cover types. In each of the plots, I installed one Malaise trap for insect sampling. For the plant sampling, I measured plant diversity and structural variables such as the number of vegetation layers in the canopy and understory, percent canopy cover percent and percent understory plant cover, and tree diameter. Analyses showed differences in parasitoids species composition, abundance and richness/individual among the different canopy cover types. In addition, partition analysis showed that beta diversity (species turnover between canopy cover types) is the most important component for parasitoids regional diversity. I also found differences in plant species composition among the types of coverage as well as average tree diameter. Meanwhile, canopy percent cover, and vegetation percent cover in the understory were also different among treatments. With non-metric multidimensional scaling I found that most parasitoids are associated with forested sites, particularly with the areas having native tree canopy cover. The variables that best explained this result were: number of layers of trees, tree percent cover and tree species richness. These results indicate the importance of tree coverage for the community of parasitoid wasps. This coverage may be providing a variety of shelters and hosts in contrast to grasslands. In addition, these results point to the relevant role of the studied fragment in maintaining the parasitoid wasp community and I recommend an extension of active restoration efforts in the remaining grasslands in order to stimulate the movement of these insects between large areas of the refuge.