LAUSR

Freshwater coastal wetlands provide numerous ecosystem services, including habitat, nutrient uptake, coastal stabilization, and aesthetic value, but the integrity of these ecosystems is threatened by invasion of non-native competitors. Invasive species, such as Phragmites and Typha, are a concern in these wetlands, as they can dominate and outcompete native species. This work sets out to understand the conditions that allow invasive species to dominate. This will allow for better management of landscapes and wetlands. We bring together two datasets to relate landscape conditions to coastal wetland invasion: (1) a spatially explicit map of nutrient inputs (SENSEmap) across the US Great Lakes Basin, and (2) a satellite land use map that includes explicit classifications of wetlands. Using machine learning algorithms, we quantified correlations between wetland plant invasion along the coastline to nutrient loads (both N and P) and other landscape scale variables (hydraulic conductivity, slope, imperviousness, land use, and land cover) across multiple influence zones. We find that high invasion is typically associated with nitrogen loading above 118 kg/ha/year within the watersheds of the invaded wetlands. Forest cover of   2.6% and phosphorus loads < 2.8 kg/ha/year are associated with low invasion. Through N:P ratios, phosphorus was further identified as important. Overall, areas more anthropogenically impacted were more associated with invasion. We conclude that high nitrogen and low forest cover are correlated with invasion. These conclusions will inform management, as well as future efforts to identify linkages between landscapes and coastal invasion.