Climate warming is have a variety of impacts on marine species (Poloczanska et al 2013) including broad scale changes in their spatial distribution such as poleward range shifts (Last et al 2011). Species distribution models (SDMs) are increasingly being applied in marine science for predicting the present and future geographic range of individual species (Robinson et al 2017), however, the amount of data used to fit SDMs can vary from minimal (e.g., museum collections Saupe et al 2014) to considerable (e.g., annual research vessel surveys Selden et al 2018).
SDMs are purely correlative models that usually neglect spatial population dynamics and assume that species distribution is in equilibrium with the environment (Pagel and Schurr 2012). Many SDMs lack rigorous testing using independent data (for an exception see Champion et al 2018) and there are concerns regarding the lack of appropriate validation procedures (Araújo and Guisan 2006, Hijmans 2012). The majority of SDMs are not tested against independent data partly due to the paucity of fully independent sources of observational data. Consequently, the ability of SDMs to accurately forecast future impacts of climate warming is questionable.
Marine species that are routinely monitored using research vessel surveys, principally commercial fish, offer an opportunity to compare the output of SDMs generated using typical levels (minimal) of observational data with the output of empirically derived distribution maps that are derived from survey data having high spatial and temporal resolution. Systematic divergence between the two model outputs would more clearly indicate the basic nature of caveats that should be applied when SDMs are the used in data-poor situations.
The aim of this project is to evaluate the performance of SDMs in their ability to forecast future distributions based on statistical understanding of habitat suitability. There are currently a wide range of statistical modelling approaches to describing the geographic distribution of fish (Thorson 2019a,b) that will be used to give alternative representations of spatial distributions. Data from annual bottom trawl surveys of European fish will be used, however, there is also scope to consider application of SDMs to non-European fish, e.g., data-poor fish stocks.
This project will require interest and skills in fitting and interpreting a range of different spatial models and, given the extensive literature on the topic, an aptitude for literature research.
Araújo, M.B., and Guisan, A. 2006. J. Biogeogr. 33:1677-1688.
Champion, C., Hobday, A.J., Zhang, X., Pecl, G.T. and Tracey, S.R. 2018. Mar. Freshw. Res. https://doi.org.10.1071/MF17387
Hijmans, R.J. 2012. Ecology. 93:679-688.
Last, P.R., White, W.T., Gledhill, D.C., Hobday, A.J., Brown, R., Edgar, G.J., and Pecl, G. 2011. Global Ecol. Biogeogr. 20:58-72.
Pagel, J., and Schurr, F.M. 2012. Global Ecol. Biogeogr. 21:293-304.
Poloczanska, E.S., et al. 2013. Nature Clim. Change 3: 919-925.
Robinson, N.M., Nelson, W.A., Costello, M.J., Sutherland, J.E., and Lundquist, C.J. 2017. Front. Mar. Sci. 4: Article 421.
Saupe, E.E., Hendricks, J.R., Peterson, A.T., and Lieberman, B.S. 2014. J. Biogeogr. 41:1352-1366.
Selden, R.L., Batt, R.D., Saba. V.S. and Pinsky, M. 2018. Glob. Change Biol. 24:117-131.
Thorson, J.T. 2019a. Fish Fisheries 20:159-173.
Thorson, J.T. 2019b. Fish. Res. 210:143-161.