Abstract
Littoral habitat is critical for lake biota but is adversely affected by residential shoreland development through the loss and reduced structural complexity of lakeshore vegetation. There currently exists no assessment methodology for evaluating littoral habitat condition of individual lakes in northeastern US. We addressed this assessment need by creating multi-metric indices of littoral habitat condition that focus on lakeshore residential development as the primary stressor. We did this by using habitat metrics derived primarily from National Lake Assessment (NLA) Physical Habitat (PHAB) survey field observations to create Linear Discriminant Analysis (LDA) models that assign lakeshore stations into littoral habitat condition categories. Lake PHAB survey data were used from New England NLA surveys as well as state-level surveys completed in Maine, New Hampshire, and Vermont. Prediction success rates in New England models averaged 83%. The Maine LDA models, which used finer scale survey methods, had an average prediction success rate of 89%. We used 95% bootstrapped confidence intervals to make assessment designations of natural (meeting reference quality), diminished (not meeting reference quality), or intermediate (existing between natural and diminished) littoral habitat condition for each lake. Our results show that efficacious single-lake littoral habitat assessments may be completed within the framework of NLA PHAB methodology, but confidence in assessment results, and therefore better-informed management decisions, can be improved with finer-scale observation data.