Nativity and seed dispersal mode influence species’ responses to habitat connectivity and urban environments

Abstract

Aim: Urbanization alters local environmental conditions and the ability of species to disperse between remnant habitat patches within the urban matrix. Nonetheless, despite the ongoing growth of urban areas worldwide, few studies have investigated the relative importance of dispersal and local environmental conditions for influencing species composition within urban and suburban landscapes. Here, we explore this question using spatial patterns of plant species composition. Location: The Research Triangle area, which includes the cities of Raleigh, Durham, Chapel Hill and Cary, in central North Carolina, USA. Time period: 2012-2014.Major taxa studied: Vascular plants. Methods:We sampled riparian forest plant communities along an urban-to-rural gradient and used redundancy analysis to identify predictors of species composition patterns for groups of species categorized by nativity and seed dispersal mode. We first compared the ability of different models of habitat connectivity (least-cost paths that avoided urban land cover versus Euclidean and along-stream distance) to explain spatial patterns of species composition. We then partitioned the variation in species composition explained by habitat connectivity models, local environmental conditions and measures of urbanization in the surrounding landscape.Results: We found that several groups of native species were best explained by least-cost path models that avoided urban development, suggesting that urbanization impedes dispersal within this landscape, particularly for short-dispersed species. Environmental variables related to urbanization (e.g., temperature, stream incision) were important predictors of species composition for many species groups, but measures of urbanization in the surrounding landscape were more important for exotic than for native species.Main conclusions: Our results demonstrate that urbanization influences plant species composition via its effects on both habitat connectivity and environmental conditions. However, the strength of these effects varies somewhat predictably across seed dispersal modes and between native and exotic species. These results highlight the importance of landscape-scale planning for urban conservation. K E Y W O R D S dispersal syndrome, habitat connectivity, metacommunity, Moran's eigenvector maps, plants, riparian, urbanization

Testing the effects of four urbanization filters on forest plant taxonomic, functional, and phylogenetic diversity

“…); indeed, in a related study (Lopez et al. ), we found a time lag between reductions in habitat connectivity due to urban development and corresponding changes in species composition. Habitat transformation is also more likely to have overall negative effects on biodiversity in habitat types that are systematically replaced or transformed by urban development, such as some types of wetlands (Williams et al.…”

“…Despite the direct negative effect of habitat transformation on native and overall species richness, we found that the total effects of habitat transformation, including its indirect effects via environmental variables and fragmentation, were positive. These positive effects may be short-term changes that could reverse with the onset of extinction debts in the future (Hahs et al 2009); indeed, in a related study (Lopez et al 2018), we found a time lag between reductions in habitat connectivity due to urban development and corresponding changes in species composition. Habitat transformation is also more likely to have overall negative effects on biodiversity in habitat types that are systematically replaced or transformed by urban development, such as some types of wetlands (Williams et al 2009), than in the relatively intact riparian forests we studied.…”

“…Dispersal modes were compiled from multiple published sources in a previous study (Lopez et al. ). Continuous trait data (specific leaf area, seed mass, and leaf nitrogen content) came primarily from the TRY database (Kattge et al.…”

“…Wetland indicator status information came from the Army Corps of Engineers' Wetland Plant List (Lichvar et al 2014), supplemented by habitat information from the Flora of the Southern and Mid-Atlantic States (Weakley 2010), which also provided information on leaf type and growth form. Dispersal modes were compiled from multiple published sources in a previous study (Lopez et al 2018). Continuous trait data (specific leaf area, seed mass, and leaf nitrogen content) came primarily from the TRY database (Kattge et al 2011), supplemented by a data set of seed mass and leaf nitrogen for Eastern North American tree species (Coyle et al 2014) and from other sources for a few species with high cover that were missing from TRY (Appendix S1).…”

Conceptualizing social-ecological drivers of change in urban forest patches

“…Insights gained from urban forest patches at the landscape scale (Wu 2014) have contributed to an increased understanding of plant functional groups and forest edges in cities (Godefroid and Koedam 2003a) and biodiversity conservation (Sukopp and Weiler 1988;Löfvenhaft et al 2002;Godefroid and Koedam 2003b). Others have examined patterns of change in forest patch size, shape, and number (Zipperer et al 2012), and have emphasized the importance of the urban landscape matrix to native and non-native species dispersal (Lopez et al 2018). Recent city or metropolitan region-wide approaches have yielded new insights such as indicating that land cover, land use histories, and pre-urban vegetation have lasting effects on ecosystem structure and function (e.g., Fahey et al 2012;Carter et al 2015;Fahey and Casali 2017).…”