Environmental predictors of phytoplankton chlorophyll-a in Great Lakes coastal wetlands

Abstract

Coastal wetlands of the Laurentian Great Lakes are diverse and productive ecosystems that provide many ecosystem services, but are threatened by anthropogenic factors , including nutrient input, land-use change, invasive species , and climate change . In this study, we examined one component of wetland ecosystem structure – phytoplankton biomass – using the proxy metric of water column chlorophyll- a measured in 514 coastal wetlands across all five Great Lakes as part of the Great Lakes Coastal Wetland Monitoring Program. Mean chlorophyll- a concentrations increased from north-to-south from Lake Superior to Lake Erie , but concentrations varied among sites within lakes. To predict chlorophyll- a concentrations, we developed two random forest models for each lake – one using variables that may directly relate to phytoplankton biomass (“proximate” variables; e.g., dissolved nutrients, temperature, pH) and another using variables with potentially indirect effects on phytoplankton growth (“distal” variables; e.g., land use, fetch). Proximate and distal variable models explained 16–43% and 19–48% of variation in chlorophyll- a , respectively, with models developed for lakes Erie and Michigan having the highest amount of explanatory power and models developed for lakes Ontario, Superior, and Huron having the lowest. Land-use variables were important distal predictors of chlorophyll- a concentrations across all lakes. We found multiple proximate predictors of chlorophyll- a , but there was little consistency among lakes, suggesting that, while chlorophyll- a may be broadly influenced by distal factors such as land use, individual lakes and wetlands have unique characteristics that affect chlorophyll- a concentrations. Our results highlight the importance of responsible land-use planning and watershed-level management for protecting coastal wetlands.