Skip to main content

Data from: Cascading ecological effects of landscape moderated arthropod diversity

Cite this dataset

Smith, Jeffrey R.; Schmitz, Oswald J. (2015). Data from: Cascading ecological effects of landscape moderated arthropod diversity [Dataset]. Dryad.


Understanding the mechanisms regulating the diversity and distribution of arthropods is essential to understanding food web interactions and ecosystem functioning. Local arthropod diversity is known to be linked to features of surrounding landscapes, including the area of human-developed land. Yet, how such landscape moderation of diversity affects processes within local sites remains understudied. We report on a study that 1) measured the impacts of human development surrounding old-field habitats of arthropods on arthropod food web structure within those habitats and 2) determined if these shifts were associated with cascading impacts on the plant community. We sampled the arthropod community in 16 old-fields that span an urban-rural gradient throughout southern New England, USA. In each field, we also established paired mesocosms enclosing vegetation, one of which allowed arthropod herbivory while the other excluded such interactions, to isolate impacts of arthropod herbivory on three functional groups of plants: grasses, goldenrod and non-goldenrod forbs. Biomass of both herbivorous and predatory arthropods were positively related to the proportion of natural area surrounding a field early in the growing season (June). This relationship persisted later into the season for predatory arthropods (through July), but not for herbivorous arthropods. We found no evidence that the biomass of predators was related to the abundance of herbivorous arthropods in a field; or that biomass of herbivores was correlated to change in plant biomass between the two types of mesocosms. We did, however, find that in fields with low predator abundance there was greater herbivory on grasses (nutritious host), but that in high predator fields goldenrod was increasingly impacted (safe host), as is predicted by past work in old-field ecosystems. The findings support the generalizability of landscape moderated biodiversity to non-agricultural systems and suggests that observed shifts in food webs have implications for community and ecosystem dynamics.

Usage notes


Rhode Island
United States
New England