Social and structural traits influence species gains while resources influence species losses in a native grassland

Multiple factors influence temporal species turnover, including resource requirements and species traits. The standard model in plant ecology is that adding soil nutrients will result in taller communities, reducing understory light levels and leading to species loss via size-asymmetric competition. However, underlying this model is the notion that competitive dynamics are outcomes of individual species characteristics rather than an emergent trait of the suite of species involved in the interactions. Thus, whether plant social context (identities and interactions of neighbors) impacts competitive outcomes is unclear and potentially overlooked. Using data from a three-year field study manipulating light and soil resources, we asked how resource manipulations, community diversity, or structural (physiological) and social (interactions with neighbors) traits influenced species turnover. We created co-occurrence networks to develop novel metrics that capture the prevalence of positive and negative associations for each of 24 species in a native grassland community. We then estimated temporal beta diversity to partition species turnover into gains and losses, testing whether these compositional changes were impacted by resource manipulations or communities' structural or social traits. We found evidence that resources, structural traits, and social traits all impacted aspects of community assembly. Nutrient addition but not reduced light increased species losses, and communities with either high or low specific leaf area (SLA) and root tissue density (RTD) community-weighted mean (CWM) trait values gained more species. Communities consisting of species forming numerous positive species co-occurrences gained fewer species throughout the study than communities of species forming fewer positive co-occurrences. Thus, a species' tendency to form positive co-occurrences has a functional consequence for community-level compositional stability. Resource addition increased species losses independently of CWM height, suggesting it was not size-asymmetric competition for light that resulted in species loss in our study. Together, these results challenge the notion that nutrient-driven species loss is primarily mediated by size-asymmetric competition, highlighting the role of species' social interactions in governing community change.