Species interactions form the backbone of ecological communities, but the intensification of agricultural landscapes has had massive impacts on global biodiversity, consequently affecting species interactions and the ecological processes they maintain. Yet, agricultural systems hinge on ecosystem processes driven by species interactions. Every year $39.4 billion USD is spent globally on preventing plant-herbivore interactions, and 35% of crops worldwide depend on plant-pollinator interactions. Within a community, all of the species interactions form an interaction network, the structure of which may determine the functioning of said ecosystem processes, as well as biodiversity maintenance and community resistance to invasion and extinction. Despite the economic and ecological utility of understanding the impact of agricultural intensification on species interaction networks, empirical evidence on the impacts of agricultural intensification on network structure is scarce. Of the available data, most studies focus on one network type at a time, for example, mutualistic plant-pollinator networks or antagonistic plant-herbivore networks. However, in real systems, multiple networks connect to form so-called ‘hybrid’ networks. Because different types of networks tend to display different patterns of network structure, they may have distinct responses to agricultural intensification, ultimately driving changes in overall hybrid network structure. Using empirical, hybrid plant-pollinator-plant-herbivore networks from 16 sites along an agricultural intensification gradient, I compare how changes in mutualistic versus antagonistic networks contribute to changes in overall hybrid network structure along the intensification gradient.
Results/Conclusions
We find that both network interaction richness and evenness decline along the agricultural intensification gradient in both antagonistic and mutualistic networks, leading to similar changes in overall hybrid network structure. However, preliminary results suggest that interaction richness and evenness decline at different rates along the gradient in mutualistic and antagonistic networks, suggesting that different network types make disparate contributions to changes in hybrid network structure. We also find a decline in network specialization along the gradient, and that plant species have significantly different interaction characteristics depending on whether the hybrid, antagonistic, or mutualistic network is being considered. These results highlight the necessity of empirical data on hybrid networks to paint a realistic and complete picture of the impact of agricultural intensification and other anthropogenic disturbances on species interaction network structure. A more complete understanding of the response of networks to agricultural intensification is critical for accurately predicting and effectively preventing biodiversity loss and change in agricultural landscapes.