Research Document - 2013/131
Recovery Potential Assessment for the American Eel (Anguilla rostrata) for eastern Canada: recovery potential assessment population modelling
By Jennifer A.M. Young and Marten A. Koops
Abstract
The Committee on the Status of Endangered Wildlife in Canada (COSEWIC) has assessed American Eel (Anguilla rostrata) as Threatened in Canada, due to dramatic declines over a significant portion of its distribution (COSEWIC 2012). In support of a recovery potential assessment (RPA) for American Eel, we present population modelling of this panmictic species to assess population sensitivity to perturbation, and to compare possible dynamics of American Eel under various hypotheses regarding the distribution of eel larvae (e.g., maternal effects versus random distribution). Special attention is paid to transient dynamics (short term fluctuations in abundance as dynamics approach long term stability). The species’ range was subdivided into RPA geographical zones. Population growth tended to be proportionally more sensitive to changes in natural mortality than in any other life history parameters, especially during early life stages (leptocephali, glass, elver), regardless of larval distribution assumptions. Sensitivity across zones was very dependent on larval distribution. Larval distribution also strongly influenced the type and duration of transient (short term, non-constant) dynamics observed in population projections. Scenarios with random distribution of larvae converged quickly to stable distributions, while strong maternal effects caused transient dynamics lasting 100s and 1,000s of years. These dynamics were often counterintuitive; apparent long-term growth and decline were both observed in populations with stationary vital rates (stable long-term abundance). Expected stable stage distributions also varied dramatically with larval distribution; if larvae were distributed by a “water attraction” hypothesis, 87% of spawners came from US zones, while eels were more distributed among zones with a “nearest neighbour” hypothesis. Stochastic simulations revealed large variation in possible trajectories and suggest that environmental variability may mask changes resulting from management or transient dynamics.
- Date modified: