Plant-herbivore-parasitoid food webs across an edge gradient from natural southern beech to plantation pine forests.

The loss and fragmentation of natural habitats resulting from conversion to production land uses are among the main drivers of biodiversity loss and increase the proportion of habitat edges. These edges drive edge effects, which are ecological and physical changes in response to distance to habitat edges. Habitat edges between natural and production land affect ecological patterns and processes, including the networks of species interactions on either side of the edge. Because food webs depict both the composition of the community and the feeding links, which represent a measure of energy flow, they can reveal the relationships among community diversity, stability and ecosystem functions. Here, we present data on the quantitative food web of a tritrophic system comprising plants, herbivores (Lepidoptera larvae) and parasitoids (Hymenoptera, Diptera, and Nematoda), sampled across habitat edge gradients between New Zealand native southern beech forest (dominated by Fuscospora species) and non-native Pinus radiata plantation forests. Furthermore, we present plant-herbivore-parasitoid interaction data following an experimental reduction of herbivore densities by aerial spraying of bioinsecticide over 20 hectares of plantation forests adjacent to native forests. Overall, across the 24 sampling sites spanning a c.a. 2600-km2 area, we sampled 16,017 (over 680 unique) plant-herbivore and 1127 (over 200 unique) herbivore-parasitoid interactions between over 150, 130, and 65 plant, herbivore, and parasitoid species. The information presented here has potential use in studies across a variety of spatial scales, from local and landscape levels to macroecological research. We expect that further exploration of this dataset can contribute to developing conservation and management strategies for native and plantation forests, as well as to improve our understanding of food web assembly mechanisms in forest habitats. The data are available for reuse under CC BY 4.0 licensing. Please cite this data paper when using the dataset in publications or teaching demonstrations.