A major goal of ecosystem-based fisheries management is to prevent fishery-induced shifts in community states. This requires an understanding of ecological resilience: the ability of an ecosystem to return to the same state following a perturbation, which can strongly depend on species interactions across trophic levels. We use a structured model of a temperate rocky reef to explore how multi-trophic level fisheries impact ecological resilience. Increasing fishing mortality of prey (urchins) has a minor effect on equilibrium biomass of kelp, urchins, and spiny lobster predators, but increases resilience by reducing the range of predator harvest rates at which alternative stable states are possible. Size-structured predation on urchins acts as the feedback maintaining each state. Our results demonstrate that the resilience of ecosystems strongly depends on the interactive effects of predator and prey harvest in multi-trophic level fisheries, which are common in marine ecosystems but are unaccounted for by traditional management.
Rocky reef multi-trophic level fisheries model: baseline
R script to run the baseline model.
Baseline.R
Rocky reef multi-trophic level fisheries model: equilibria FL x FU
R script to run simulations across values of both lobster and urchin fishing rates
Equilibria_FL_FU.R
Rocky reef multi-trophic level fisheries model: equilibria FL x sigma
R script to run simulations across values of lobster fishing and urchin recruitment facilitation strength
Equilibria_FL_Sigma.R
Rocky reef multi-trophic level fisheries model: equilibria FU x sigma
R script to run simulations across values of urchin fishing and urchin recruitment facilitation strength
Equilibria_FU_Sigma.R
Rocky reef multi-trophic level fisheries model: Hysteresis FL x FU
R script to run simulations testing for hysteresis across values of lobster and urchin fishing
Hysteresis_FL_FU.R
Rocky reef multi-trophic level fisheries model: Hysteresis FL x sigma
R script to run simulations testing for hysteresis across values of lobster fishing and urchin recruitment facilitation
Hysteresis_FL_Sigma.R
Rocky reef multi-trophic level fisheries model: Hysteresis FU x FL
R script to run simulations testing for hysteresis across values of urchin fishing and lobster fishing
Hysteresis_FU_FL.R
Rocky reef multi-trophic level fisheries model: Hysteresis FL x FU- No Size Structure
R script to run simulations testing for hysteresis across levels of lobster and urchin fishing, attack rates of predators on urchins are all equal (ie, no size-structured predation)
Hysteresis_FL_FU_NoSizeStructure.R
Rocky reef multi-trophic level fisheries model: Hysteresis FL x FU- Extra Size Structure
R script to run simulations testing for hysteresis across values of lobster and urchin fishing with predator attack rates on urchins more disparate across sizes (ie, enhanced size-structured predation)
Hysteresis_FL_FU_XtraSizeStructure.R
Rocky reef multi-trophic level fisheries model: Bistability
R script to run simulations examining the full range of parameter space (FL x FU) at sigma= 0, 0.5, and 0.95. Attack rates are set at the baseline level. This provides regions of kelp forest, urchin barren, and bistability.
Stability_FL_FU.R
Rocky reef multi-trophic level fisheries model: Bistability No Size Structure
R script to run simulations examining the full range of parameter space (FL x FU) at sigma= 0, 0.5, and 0.95. Attack rates are set to be equal across urchin sizes. This provides regions of kelp forest, urchin barren, and bistability.
Stability_FL_FU_NoSizeStruct.R
Rocky reef multi-trophic level fisheries model: Bistability Extra Size Structure
R script to run simulations examining the full range of parameter space (FL x FU) at sigma= 0, 0.5, and 0.95. Attack rates are set for enhanced size-structured predation. This provides regions of kelp forest, urchin barren, and bistability.
Stability_FL_FU_XtraSizeStruct.R
Rocky reef multi-trophic level fisheries model: Global sensitivity analysis
R script to run a global sensitivity analysis (GSA) of all 24 parameters in the rocky reef multi-trophic level fisheries model. This GSA uses random forest to quantify parameter importance to the modeled outcome (mean trophic level at equilibrium). It then uses classification and regression trees to determine specific breakpoints in parameter space which drive modeled outcome. We use a uniform distribution for each parameter to set up 4,000 parameter sets which we then simulate the model using. Other distributions can also be used (Gaussian provided very similar results).
GSA_Uniform.R
Rocky reef multi-trophic level fisheries model: baseline & hysteresis with external recruitment
R script to run simulations of the rocky reef multi-trophic level fisheries model now including a constant level of external recruitment into the small urchin compartment. We simulate the baseline time series as well as tests of hysteresis for: FL x FU, FL x sigma, and FU x sigma.
Baseline_Hysteresis_ExtraRecment.R
Rocky reef multi-trophic level fisheries model: Sheephead
R script to run simulations testing the effect of changing predator attack rates and handling times to reflect predation by the fish, California sheephead.
Sheephead.R
Rocky reef multi-trophic level fisheries model: baseline + harvest
R script to run simulations calculating lobster and urchin fishery harvest at each time step. We use these to estimate the fishery harvest level at maximum sustainable yield (F_msy) for lobsters at three levels of urchin harvest.
3stage_WithHarvest.R