Predicting and mapping Laurentian Great Lakes reefs in support of fisheries management and conservation

Reefs represent vitally important benthic habitat for numerous aquatic species. Despite their importance, rocky reef habitat mapping in freshwater environments is lacking, particularly across large spatial extents. We develop predictive reef models within nearshore zones (depth ≤ 30 m) of the Laurentian Great Lakes using known reef locations and depth, bathymetric position index, and distance to hard substrate as predictors at a 30 m resolution. We estimate reef area in each lake and characterize reefs according to size, shape complexity, and distances to nearest neighboring reef and offshore habitat (depth > 30 m). This modeling approach is extended in two case studies creating detailed reef maps with high resolution 1‒2 m benthic data. Results demonstrate the rarity of this habitat type, with total reef area estimated at 0.8% of benthic area across the Laurentian Great Lakes and ranging from 0.2 to 1.1% among lakes. Reef measures indicate that Lake Superior reefs are typically larger, have more complex shapes, are more distant from neighboring reefs, and are situated closer to offshore habitat than reefs in other Great Lakes. Independent validation of reef predictions with historic spawning observations for two reef-spawning fishes, lake whitefish ( Coregonus clupeaformis ) and lake trout ( Salvelinus namaycush ), indicates congruence between predicted reefs and observed spawning locations. Case studies demonstrate the applicability of this approach in developing high resolution reef maps with regionally available data. Collectively, these results can aid fisheries management and conservation as well as reef restoration efforts involving habitat degradation, invasive species, and changing climate.