Local habitat complexity and its effects on herbivores and predators in urban agroecosystems

In urban community gardens, cultivated vegetation provides variable levels of habitat complexity, which can suppress pests by promoting predator diversity and improving pest control. In this study, we examine three components of the structural complexity of garden vegetation (cover, diversity, and connectivity) to investigate whether higher garden vegetation complexity leads to fewer herbivores, more predators, and higher predation. We worked in eight community gardens where we quantified vegetation complexity, sampled the arthropod community, and measured predation on cornworm eggs. We found that plots with high vegetation cover supported higher species richness and greater abundance of predatory insects. High vegetation cover also supported a greater abundance and richness of spiders. In contrast, high vegetation diversity was negatively associated with predator abundance. While high predator abundance was positively associated with egg predation, greater predator species richness had a negative impact on egg predation, suggesting that antagonism between predators may limit biological control. Community gardeners may thus manipulate vegetation cover and diversity to promote higher predator abundance and diversity in their plots. However, the species composition of predators and the prevalence of interspecific antagonism may ultimately determine subsequent impacts on biological pest control.

Methods
These datasets describe arthropods, predation levels, and vegetation characteristics sampled and measured in urban community garden sites in the California central coast in August 2019. Arthropods were sampled using a combination of visual surveys of garden plants and pitfall traps placed in sample plots. Predation levels were measured using corn earworm moth eggs as sentinel prey in sentinel pest experiments at each sample plot. Vegetation characteristics were measured in four 1.5 x 1.5 m sample plots per garden site. We used generalized linear models to analyze relationships between arhtropod abundance/richness, predation levels, and vegetation characteristics. More detailed methods are described in our Insects publication of this study. Details about each dataset are found in the README file.

Datasets included:

1) PitfallTrap_1.5.24

Abundance and species richness of herbivores and predators (all predators, ants, and spiders) sampled through pitfall traps
Egg predation (LRR, effect size) measured through sentinel pest experiments
Vegetation cover, diversity, and connectivity measurements for each sample plot

2) VisualSurvey_1.5.24

Abundance and species richness of herbivores (all herbivores, aphids, and whiteflies) and predators (all predators, ants, and spiders) sampled through visual surveys of plants
Egg predation (LRR, effect size) measured through sentinel pest experiments
Vegetation cover, diversity, and connectivity measurements and gardener bed size for each sample plot

Missing value are inducated by NA. See README file for additional details.