Urban gardens provide valuable vegetative complexity for urban biodiversity in harsh, often simplified urban environments. Insects utilizing garden habitats experience landscape complexity at multiple scales according to the relative availability of vegetative resources within and surrounding gardens as well as the dispersal abilities of individual taxa. At the local-scale, differences in the architectural complexity of crop varieties may inform habitat selection of herbivorous insects and their natural enemies. Using a tri-trophic model system of cabbage aphids (A, Brevicoryne brassicae), ladybeetles (L, Hippodamia convergens), and parasitoid wasps (P, Aphidius ervi), I investigated the influence of kale leaf architecture on aphid population regulation in a greenhouse experiment. I used four insect treatments (A, A+L, A+P, A+L+P) and three kale treatments (lacinato, curly, lacinato+curly) for a total of 12 treatments and 15 replicates each. I inoculated kale with aphids and introduced ladybeetles and parasitoid wasps into bug dorms for 24h. I censused aphids on each plant before and after natural enemy introduction. Finally, I calculated a predation effect size (LN(proportion of aphids removed in natural enemy treatments) - LN(proportion of aphids removed in aphid-only treatments)) and used generalized linear mixed models (GLMMs) to assess the explanatory power of plant and insect treatments.
Results/Conclusions
Aphid populations increased on average by 0.7% when alone, and declined by 10% with L, 8.6% with P, and 11.2% with L+P. On average, 6.3% of aphids were removed on curly kale, 7.7% on lacinto, and 7.9% on curly+lacinto. Overall, aphid populations increased when alone on curly kale (9.1%) and declined the most on curly kale with L+P (16.8%). GLMMs indicate plant treatment significantly affected predation effect size (df = 2, F = 9.9165, p = 0.00161). The best model included plant treatment as a fixed factor and replicate as a random factor (AIC = 44.96146, ΔAIC for next best model = 6.10999, df = 2, R2 = 0.5802). Insect treatment was not a significant factor. Curly kale had the highest effect size and lacinato kale had the lowest. These results indicate that aphids and natural enemies may both benefit from the architectural complexity of curly kale (compared to flatter-leaved lacinato), possibly due to increased microhabitat surface area. While simpler crops may attract fewer aphids, they also support less natural enemy foraging. Thus, growing a diversity of crops varying in architectural complexity may buffer the effect of pest accumulation on architecturally complex crops while providing foraging habitat for natural enemies.