PARENT SESSION
Poster Session 10: Restoration and Adaptive Management
Tuesday, August 9, 5:00 PM - 6:30 PM, Exhibit Hall 220 A-E, Level 2, Palais des congrès de Montréal

Modelling target fish communities for lakes undergoing ecosystem recovery.

Jackson, Donald^*,1, Olden, Julian², ¹ University of Toronto, Toronto, ON, Canada² University of Wisconsin-Madison, Madison, WI, USA

ABSTRACT- Determining potential target communities or viable populations in restored habitats is of major importance, but typically done with little or no ability to predict the outcomes. We develop an approach that can be used to predict target communities. Lakes in a region north east of Lake Superior were severely impacted by smelting operations. Lakes varied in their impact with several having extremely acid conditions (e.g. pH 3.5) and no fish species present. As smelting operations have ceased, systems have been recovering both chemically and biologically. Using a series of impacted and unimpacted lakes, we developed artificial neural network models to predict fish species composition based on the lake morphological and chemical conditions. The resulting models have good predictive power based on historical data. We apply these models to the lakes that were severely impacted to assess their potential to support species currently and to predict target communities under various recovery scenarios.

Key words: statistics, restoration, fish, community