Assumptions, Evolution, and Application of Conceptual Models of Roanoke Logperch Population Dynamics

Conceptual models of fish dynamics are built upon theory, empirical evidence, and expert judgment, all of which have embedded assumptions. Expert judgment is especially prominent in models for rare species, for which data are often lacking. Herein we draw from our monitoring of endangered Roanoke logperch to illustrate the assumptions, evolution, and application of conceptual models of fish population dynamics. The original, implicit model of logperch dynamics, based largely on expert judgment, assumed stable abundance, deterministic regulation, and a strong connection between abundance and habitat availability. Based on this model, flood-control construction on the Roanoke River was expected to reduce logperch abundance by decreasing availability of silt-free habitat, and this population reduction was assumed detectable. Subsequent data collection and analyses revealed high, stochastic variation in pre-construction abundance, and weak relationships between abundance and habitat. This new knowledge led to refinement of our conceptual model and modification of monitoring and construction procedures. Based on our experiences, we suggest that scientists need to 1) explicate conceptual models and uncertainties, 2) seek opportunities to test key assumptions, 3) expect models to change as more data become available, and 4) work closely with managers to incorporate new knowledge into actions and policies.