Effects of top predator re-establishment and fishing on a simulated food web : Allometric Trophic Network model for Lake Oulujärvi

Abstract

Fish communities face changes in environmental conditions and fishing that affects the abundances and structures of the populations. Before 1960s there were abundant stocks of both pikeperch (Stizostedion lucioperca) and whitefish (Coregonus lavaretus) in Lake Oulujärvi, but in 1960s–1970s the stock of pikeperch declined to very low levels while whitefish stock remained abundant. Due to massive re-introductions, pikeperch recovered since 1999 and is again common while whitefish stock has declined. To understand the ecosystem-level changes observed along the recovery of the pikeperch stock, we constructed a food web model capturing the two most recent states of pikeperch abundance in Lake Oulujärvi. We used Allometric Trophic Network (ATN) model to simulate the pelagic food web in the presence and absence of pikeperch and in the presence and absence of fishing. To parametrize ATN model based on body masses and food web interactions, we used data collections of fish cohort analyses, fish individuals, fish stomach contents, zooplankton, and phytoplankton in Lake Oulujärvi. Pikeperch decreased the biomasses of its planktivorous prey. Fishing truncated the age distribution of planktivorous fish. Pikeperch and fishing had synergistic negative effects on vendace (Coregonus albula) and smelt (Osmerus eperlanus) percentages of fish biomass, and antagonistic negative effect on whitefish and brown trout (Salmo trutta) percentages of fish biomass. Mysis relicta, Chaoborus flavicans, Leptodora and other predatory zooplankton, and Cyclopoida zooplankton guilds increased with fishing and pikeperch. Fishing, and pikeperch in the presence of fishing, increased biomass of Crustacean zooplankton guild. There were marked differences between the ATN model simulations and empirically observed time series of fish stock abundances suggesting that the observed changes are partially caused by environmental or fishing-related factors that were not included in the model.