Urban-rural gradient studies have observed suppressed recruitment of native woody species in urban forests (e.g. Cadenasso et al. 2007; Aronson et al. 2015). While native tree recruitment limitation in urban forests is apparent, limited research has been conducted on the ecological mechanisms responsible for these changes in recruitment dynamics. One hypothesis is that post-dispersal seed predation may be an important determinant of seedling abundance and diversity in urban forests (Zipperer et al. 2007), however there is limited empirical research to support this. This experiment is the first to investigate urban and rural forest seed predation and tests whether: 1) the rate of seed removal for native and nonnative species is greater in urban than rural forests, 2) seed predators exhibit preference for native tree species. To answer these hypotheses a cafeteria study was designed to measure seed removal rates of six native and two non-native species that represent a range of seed size and dispersal modes. Plots were established in urban and rural forests located in New York City parks and the New York Highlands. Failure-time analyses using the Kaplan-Meier method were conducted to determine if seed removal curves and seed removal rates differ among forest type and species.
Seed removal rates were calculated for four native species (Quercus rubra, Prunus serotina, Acer rubrum, Sassafras albidum, Carya tomentosa, Carya cordiformis) and two non-native species (Acer platanoides, and Ailanthus altissima). Contrary to our hypothesis, removal rates were found to be greater in rural forests (80.3%) when compared to urban forests (67.9%). Irrespective of forest type, seed removal rates were greater for native species (85.6%) than non-native species (60.8%) and were positively correlated with seed mass. When comparing survival curves, seed removal rates were found to be significantly different when comparing forest type (urban vs. rural) (χ = 57.90, p < 0.01) and species origin (native vs. non-native) (χ = 375.29, p < 0.01). The results of this study refute the assumption that seed predation pressure is greater in urban forests. The observed preference for native species by seed predators may, however, allow non-native tree species to evade post-dispersal predation and gain an advantage during early establishment. In turn such differences in seed limitation for native and non-native tree species may significantly impact future forest composition and structure. These results begin to identify ecological mechanisms that may drive forest change in anthropogenic landscapes, advancing ecological theory and urban forest management.