Riparian vegetation performs a suite of important ecosystem functions, but it has been heavily impacted by human activities. For example, in California, extensive landscape modification has caused the loss of approximately 95% of the original riparian habitat. Riparian habitat on small lowland streams, which once permeated California’s Central Valley, has been particularly impacted, first by agriculture, and more recently, by rapid urbanization. Today, there is arguably no plant community associated with small lowland streams in this region that is not impacted by one of these land uses, but the extant plant communities and how they differ between these two land uses is poorly understood. The ground layer (<1 m height) is of particular interest because it may contribute disproportionately to the maintenance of regional biodiversity, be sensitive to disturbance, and influence long-term compositional shifts by affecting tree seedling recruitment. We compared vegetation cover, species composition, and diversity indices of the riparian ground layer of 40 urban and 40 agricultural small lowland stream reaches in California’s Central Valley and examined associations with bank slope, stream flow, and overstory cover within each land use.
Agricultural and urban riparian zones had similar amounts of live ground cover (p=0.867), but their compositions differed. Agricultural sites had lower alpha diversity (p=0.02), whilst urban sites had lower beta diversity (p<0.001). Common non-native, invasive, and upland-affiliated species were well-represented but distinct to each land use. The communities also shared a suite of common native wetland-affiliated species, suggesting some resilience to both landscape modifications. Several observations suggest that different types of human influence affected communities. In urban sites, cover and diversity were positively associated with stream flow, lower overstory cover, and gentler bank slopes, whereas these relationships were not apparent in agricultural sites. Also, where live cover was low, agricultural sites tended to have more bare ground (p=0.02), whereas urban sites had more litter and other non-living cover (p<0.001). We hypothesize that these contrasting trends were the result of direct management disturbances in the agricultural system, including channel shaping and vegetation clearing, overriding influences of other ecological drivers. Further, higher beta diversity in agricultural sites may reflect heterogeneous management activities among farms. In contrast, urban riparian zones were less directly managed and plant communities were indirectly influenced by urbanization through its effects on stream flow, bank stability, and overstory cover.