Applying the patch-matrix model to lakes: a connectivity-based conservation framework

CONTEXT: Biodiversity conservation for terrestrial species often emphasizes land protection to help maintain connectivity among habitat patches. However, conservation of aquatic and semi-aquatic species is challenging because aquatic species (e.g., fish) move among lakes using aquatic connections (e.g., streams, wetlands), whereas semi-aquatic species (e.g., amphibians) use both aquatic connections and upland habitats. OBJECTIVES: We applied the patch-matrix model to create an aquatic and semi-aquatic connectivity framework for lakes. We applied our framework using lakes in Michigan, USA to examine (1) the relationship between aquatic and semi-aquatic connectivity for lakes and (2) the extent to which protected areas encompass aquatic and semi-aquatic connectivity among lakes. METHODS: We used principal component analysis to calculate aquatic and semi-aquatic connectivity scores for lakes. We then examined relationships among aquatic and semi-aquatic connectivity scores and existing protected areas (strict and multi-use). RESULTS: Fewer than 3% of lakes had high scores for either aquatic or semi-aquatic connectivity. Connectivity scores were generally higher in Michigan’s Upper Peninsula, which is heavily forested with greater land protection. Although lake protection was overall low (16 and 32% of lake watersheds in Michigan were ≥ 10% protected under strict and multi-use protection, respectively), highly connected lakes were generally more protected than less connected lakes. CONCLUSIONS: We propose using our aquatic and semi-aquatic connectivity framework to (1) identify and prioritize lakes for conservation that are likely to have high biodiversity and conservation value and (2) generate testable hypotheses for studying the integrated terrestrial-aquatic landscape under global change.