Tri-trophic effects of non-native vegetation on insect prey and bird behavior in residential landscapes
Urban-associated changes in the abundance and richness of native organisms have been well documented globally and ecologists have recently pushed for a more mechanistic approach to understanding the drivers behind community patterns. One of the most expansive stressors facing biodiversity today is the conversion of native plant communities into assemblages of non-native species without an evolutionary history to the native fauna. Of particular importance is the reduction of herbivorous insects which occurs because non-native plants have chemical, physical, and phenological defenses for which native insects have few adaptations. The biomass of insects, like Lepidoptera larvae, support large and diverse trophic webs, and are an important component to the diets of insectivorous specialists such as birds. Unfortunately, there are few studies that have determined whether introduced plants are the ecological equivalent of the native species that are displaced.
In this study, we used a resident insectivore, the Carolina Chickadee (Poecile carolinensis), to determine whether tree origin and insect productivity influenced bird breeding behavior in residential neighborhoods. We used a hierarchical framework to model bird survey data, coupled with individual behavioral observation of color-banded birds and insect surveys, to determine local features that influence chickadee occupancy, abundance, nesting and foraging behavior.
Our data suggest that both non-native, and non-native congeneric trees, had a lower probability of hosting caterpillars than native trees (Non-native: β -0.45 ±0.20, X2= 5.4, p<0.05, CI=-0.85, -0.08), although overall insect abundance was similar. Chickadee occupancy was primarily driven by basal area of trees (β 1.36 ±0.58, p<0.05, CI: 0.22-2.50); however, breeding activity was negatively related to total non-native vegetation (β -0.03 ±0.01, p<0.001, CI: -0.05, -0.01). Abundance of chickadees was negatively related to percent pavement at the territory level (β -0.44 ±0.11, p<0.0001, CI: -0.65, -0.23) a proxy for the degree of development in the landscape. Across sites, chickadees preferentially foraged in native trees and avoidance of non-native trees increased as the proportion of non-native trees in the territory increased (β: -0.35 ±0.05, p<1.0-6, CI: -0.45,-0.24). In addition, tree foraging preference was strongly related to a Lepidoptera index related to abundance and diversity (β: 3.80 ±0.37, p<0.0001).
Our multi-trophic approach suggests that non-native plants alter bottom-up resources for consumers in novel ecosystems. Habitat management in residential zones should promote the use of native, and insect-producing, tree species in order for these systems to effectively support biodiversity.