In this paper, we employ the VAST (vector autoregressive spatio-temporal) modelling approach to predict spatio-temporal changes in macrourid bycatch in the Ross Sea region (RSR) toothfish fishery. We re-estimated the VAST models for Macrourus whitsoni and M. caml presented to WG-SAM-2022, using catch rate data collected by scientific observers up to the 2021 fishing season. Then, we employed the fitted VAST models to estimate the spatial patterns of relative abundance of M. whitsoni and M. caml in the RSR and, hence, the relative proportion of each species in the different management areas of the RSR.
Preliminary analyses confirmed that it is appropriate to employ the results of the VAST models for M. whitsoni and M. caml to set bycatch limits for Macrourus spp. in the RSR. Specifically, the three other macrourid species of the RSR (M. holotrachys, M. carinatus and Coryphaenoides armatus) are very seldom recorded by scientific observers, and preliminary analyses of the data revealed that there is no spatial separation between these three species and M. whitsoni and M. caml. The VAST models for M. whitsoni and M. caml predicted that the spatial patterns of relative abundance of the two species in the RSR were quite similar. However, M. caml were predicted to be more numerous than M. whitsoni in the different management areas of the RSR, particularly in the N70 (north of 70°S) management area.
Given that M. caml is a higher proportion of the Macrourus abundance in each management area of the RSR, the different productivities of M. whitsoni and M. caml should be considered when setting bycatch limits in the management areas of the RSR. Noting that there may have been changes in bycatch reporting in the RSR, we also recommend that future studies investigate how any changes in bycatch reporting might impact our understanding of the spatial abundance patterns of M. whitsoni and M. caml in the RSR.