Assembly rules for aggregate-species production models: simulations in support of management strategy evaluation

Abstract

Ecosystem-based fisheries management (EBFM) emphasizes sustainability at multiple levels of organization beyond single target species. Therefore, biological reference points (BRPs) for aggregated groups are required, which optimize yields while preventing overexploitation of individual species. We evaluate the tradeoffs between yield and biodiversity objectives for a wide range of aggregation strategies using multispecies surplus production models and comparing 2 simulated fish communities. We simulated population trajectories with an operating model detailing predation and competitive interactions for all individual species within each community, and with additional stochastic environmental variability for one community. Species trajectories were then aggregated by functional feeding guild, taxonomy, habitat association, size class, and at the entire community level. We estimated production parameters and BRPs (e.g. maximum sustainable yield, MSY) using a simple assessment model applied to each aggregated time series, then we applied the MSY fishing rates to each simulated community as alternative fishing strategies and compared equilibrium biomass and yield under each strategy. We were able to define multi-species reference points to meet both yield and biodiversity objectives across full system, taxonomic, habitat, feeding, and size-based aggregations. Species complexes were best able to meet both objectives when species with broadly similar productivity, environmental sensitivity and species interactions were aggregated into the complex. The impacts of simulated environmental variability on BRPs were substantial for certain species and aggregates, so including the combined impacts of environmental variation and species interactions in precautionary reference points appears critical to EBFM.

Using an aggregate production simulation model with ecological interactions to explore effects of fishing and climate on a fish community

“…Aggregate groups also tend to be more stable than the indi vidual components (Fogarty & Murawski 1998, Auster & Link 2009, so this method can be used initially to determine robust harvest levels to achieve system and aggregate group level BRPs. More complex models which explicitly include multiple species can then be used to examine the implications of those harvest strategies on the individual components of an aggregated group (as was done by Gaichas et al 2012). This 2-stage approach is also recommended due to the observation that aggregated models may have very different characteristics regarding community stability, resilience, and other behaviors of the components modeled, than one in which the components are disaggregated (Pinnegar et al 2005).…”

“…The drawbacks are those usually associated with production models (e.g. missing internal stage or age related dynamics, and ignoring different life history characteristics of individual species within an aggregation, which is explored more thoroughly in Gaichas et al 2012, this Theme Section).…”

System-level optimal yield: increased value, less risk, improved stability, and better fisheries

“…Several instances where such simulation testing has occurred demonstrate improvements across a range of not only yield, but other fisheries objectives (Worm et al 2009;Link et al 2010;Fulton et al 2011;Kaplan and Leonard 2012;Fay et al 2013Fay et al , 2015Smith et al 2015;Jacobsen et al 2017). Additional contrasts have similarly elucidated the performance of OY estimates empirically, with a range of studies having similarly explored this issue (e.g., Ralston and Polovina 1982;Mueter and Megrey 2006;Fogarty et al 2008Fogarty et al , 2012Worm et al 2009;Sparholt and Cook 2010;Gaichas et al 2012;Lucey et al 2012;Rosenberg et al 2014). A few common outcomes emerge from this collective body of work.…”

“…Several theoretical, simulation, and contextual considerations have explored this approach Link 2009, 2012;Worm et al 2009;Bundy et al 2012;Fay et al 2013Fay et al , 2015Fogarty 2014;Gaichas et al 2017;Rosenberg et al 2014;Rindorf et al 2017a) and found the method performs as one would expect from hierarchy theory, namely that resulting model outputs are consistent with the portfolio effect Jin et al 2016). Many works (e.g., Mueter and Megrey 2006;Fogarty et al 2008Fogarty et al , 2012Christensen et al 2009;Gaichas et al 2012;Lucey et al 2012;Rosenberg et al 2014;Rindorf et al 2017aRindorf et al , 2017b have examined these approaches, tested the analytical methods, explored the robustness and vagaries of the results, and generally arrived at the conclusion that at the very least, the analytical methodologies and considerations are not a limitation for the adoption of this systems approach. More so, the results are generally repeatable, robust, and ecologically defensible.…”

“…Thus, given the energetic constraints of both species production and ecological interactions, coupled with some core tenets of portfolio and hierarchy theory, an aggregate approach is actually more theoretically defensible than attempting to simultaneously maximize all single component stocks or species yield in an entire ecosystem. In some instances, particularly such as in data-poor situations, highly diverse systems where species delineations are challenging, or where the risk or history of overfishing is high, the aggregate approach is preferable to accommodate those biological and informational limitations (sensu Gaichas et al 2012;Link et al 2012). The theory, simulations, empirical evidence, and contrast in performance all show that adopting an aggregate OY has notable benefits and in many instances is indeed advisable, if not even preferable.…”