Natural regeneration in urban forests is limited by early-establishment dynamics: implications for management.

Urban forested natural areas are valuable ecological and social resources, but long-term sustainability of these habitats is challenged by environmental and social factors associated with urban ecosystems. Advances in city-scale assessments of urban forests have increased the resolution of forest community types and conditions, allowing for improved understanding of ecological function, such as natural regeneration, in these urban habitats. By applying metrics of tree regeneration that are commonly used for the management of rural forests, urban ecologists can test the potential for traditional forest management strategies within our cities. In this study, we compare urban and rural oak-hickory forest composition and structure and the capacity for natural regeneration in the New York metropolitan area. Specifically, we use two metrics of advance regeneration that describe the abundance of seedlings and saplings at different size classes to test whether this management for natural regeneration is a viable option. We found differences in recruitment dynamics between urban and rural forests that have implications for the sustainability of these forests and new management strategies. First, after controlling for forest community type, species composition in urban and rural sites was significantly different across multiple strata and within the seed bank. Species-specific capacity for natural regeneration was different in urban and rural sites, signaling the possibility of divergent successional trajectories. Second, while differences in species composition exist, both urban and rural sites were dominated by native species across all forest strata except for urban seed banks. Third, despite finding significantly lower average annual seedling abundance in urban (1.9 seedlings/m2 ) compared to rural (7.1 seedlings/m2 ) sites, we observed greater density of saplings in urban forests, and no significant difference in stocking index between sites. These findings suggest that early-establishment barriers to recruitment are greater in urban forest sites. However, once established, seedling transition into advance regeneration stages may not be different between site types, and advance regeneration may, in some cases, be more viable in urban forested natural areas. These results highlight functional differences between urban and rural forest recruitment dynamics that may impact on the future community composition of oak-hickory forests in the two landscape settings.