Taking the Final Step: Can a Full Multispecies Production Model Tell Us Anything Single-Species Models with Covariates Can't?

We define extended single-species production models (ESSPMs) as single species production models that incorporate time series of principal prey and/or predator species as biological covariates.  One advantage to ESSPMs over single species production models without biological covariates is that these models can be used to test the existence and direction of (one-way) species interactions.  However, estimates of actual interaction strengths are confounded with scaling of the covariates (i.e., catchability).  In addition, biological reference points (BRPs; e.g. maximum sustainable yield, MSY) that can be calculated in ESSPMs may be of limited value compared to those estimates that more directly incorporate species interactions, because of a lack of feedback between prey and predator species in ESSPMs.  We thus developed a full multispecies production model (MSPM) to estimate biological interaction strengths and examine tradeoffs in multispecies MSY.  As a preliminary demonstration, we fit this model to functionally analogous cod and herring species across multiple northern hemisphere large ecosystems.  We compared results from the MSPM with those from ESSPMs that examined the same species groupings.  In most cases, estimated species interactions from the two approaches were consistent regarding significance and direction, although not in all.  In contrast, the two approaches produced different MSY estimates in most of the ecosystems.  At this point, the answer to the question posed in the title is a qualified "yes", but the tradeoffs between modeling approaches to estimating BRPs for cases where biological interactions are known to be important merit further examination before use in a fisheries management context.