The decline of great Arctic charr (Salvelinus umbla) in Lake Vättern: implications based on a sizestructured food web model
(English)Manuscript (preprint) (Other academic)
Recent research has illustrated the explanatory power of body size for understanding ecosystem structure and function. Hence, many studies use body size as a tool to parameterize food web models. However, the body size of most species is not a static but dynamical trait that changes during the life time. Many other traits, including diet and vital rates, change in concern with body size. This indicates a need for food web studies to resolve the population structure of a species, rather than describing it by its total abundance or mean adult body size.
We here analyse a multispecies size-structured food web model of the pelagic community of Lake Vättern, Sweden. Here, the stock of great Arctic charr used to be commercially important but have decreased by more than 90% since the 1950s. Causes for the decrease have been debated and suggestions include (i) predation on eggs by introduced signal crayfish, (ii) overexploitation, and (iii) competition between introduced salmon and charr. As a tool for analysing potential impacts on the stock of great Arctic charr and evaluating different management options we present a dynamical food web model where the population of Arctic charr is size structured. The growth equations describe Lotka-Volterra type predator-prey interactions among non-structured species, and consumption-dependent mortality and transition rates of the size-structured Arctic charr. Mortality rates of consumers are allometrically related to body size and trophic interaction coefficients are estimated resulting in realistic abundances of all species. Results from our model analysis points in the direction of salmon stocking being the most important factor affecting the largest size class of great Arctic charr, followed by fishing intensity for the smallest size classes. We believe this multispecies modelling approach is one way of integrating size dependent traits in food web ecology and we discuss future developments of this modelling approach.
IdentifiersURN: urn:nbn:se:liu:diva-77682OAI: oai:DiVA.org:liu-77682DiVA: diva2:528414