The influence of anthropogenic shoreline changes on the littoral abundance of fish species in German lowland lakes varying in depth as determined by boosted regression trees
Residential development on lake shores is regularly associated with the conversion of natural littoral habitats to riprap, sheet piles, beaches, parks, or marinas. The subsequent loss of littoral vegetation induces a decline of structural diversity and impacts littoral fish communities. These impacts may be shaped by lake morphology. Using boosted regression trees (BRT) to relate fish abundance data from 57 north-east German lowland lakes to various factors characterizing trophic state, lake morphology, and shoreline development, we investigated the response of 11 fish species to shoreline development. The analyses revealed that mean depth followed by trophic level and shoreline length (SL) contributed most in explaining littoral fish abundance. BRT models built for deep and shallow lakes separately confirmed that primarily trophic level and SL influenced fish abundance but that littoral vegetation was relatively more important in deep compared to shallow lakes, indicating that the effects of shoreline development may be more pronounced in deep lakes where the littoral makes up a smaller proportion of the lake area as compared to shallow lakes. The BRT further demonstrated species-specific responses to shoreline degradation, indicating that the reliability of ecological quality assessments of lakes can be improved by applying separate metrics for individual species.