Habitat suitability modeling based on a spatiotemporal model: an example for cusk in the Gulf of Maine

Abstract

24Habitat use and distribution is a critical aspect in the management and conservation of a species, shortcomings of conventional HSIs: 1) the abundance indices from survey catch data typically 58 incorporated in these models do not account for changes in catchability over a time series; and 2) 59 the commonly used abundance indices, and therefore HSIs, are unable to incorporate surveys 60 from multiple gear types which sample different segments of the population and likely cover 61 different types of habitat. These issues need to be addressed in order to produce an unbiased 62 evaluation of spatio-temporal changes in habitat quality for a species over its distributional 63 range. Conventional HSIs use available data from sampled locations, hereinafter referred to as sample- 78based HSIs, which are often restricted to the locations of occurrence and typically processed to 79 assume that the samples are representative (i.e., the species is effectively sampled) and are 80 comparable through time (i.e., no changes in sampling distribution and efficiency). Therefore, 81 the sample-based HSIs might not be appropriate in at least the following two situations: 1) the 82 survey misses a significant portion or type of the species' habitat; and 2) sampling efficiency 83 (i.e., catchability) changes over space and/or through time due to density-dependent processes. 84Density-dependent habitat selection is a likely process for species in decline (MacCall 1990). 85When a species population is high, individuals move into previously marginal habitat because 86 high quality habitat is saturated; thus, the overall suitability of all occupied habitat declines on 87 average (MacCall 1990 Cusk (Brosme brosme) in the Gulf of Maine is one species where assessment is difficult using 104 conventional HSIs. It is a data-limited species, with low abundance and low catchability. central GOM to better sample species that primarily reside in complex habitat (Hoey et al. 2013). 146Six survey strata were selected for the LLS from ten offshore and four inshore strata from the The study area was divided into 5,710 cells (0.05º x 0.05º) for predicting grid-based densities, The second stage of the model approximates positive catches (c): 199The probability density function Gamma (c, x, y) is evaluated at c given a gamma distribution, knots that are generated based on the proportional density of survey data over the defined 208 domain (i.e., the 0.05º x 0.05º grid; Thorson et al. 2015). The spatial (ω) and spato-temporal (ε) 209 random effects were used in both spring and fall density estimates. 211Encounter probability p and positive catch rates λ are approximated using linear predictors 212( Thorson et al. 2015): 215where ‫‬ and ߣ are the expected probabilities of an occupied habitat and positive catches given 216 occupied habitat for sample i at a given location; ݀ ் () is the average reference density 217(encounters/positive catch rates) in year ܶ () ; ܳ is catchability for each survey; w i is the area 218 swept for sample i;...

Modeling the impacts of climate change on thorny skate (Amblyraja radiata) on the Northeast US shelf using trawl and longline surveys

“…The longline survey has very limited spatio‐temporal coverage compared to the bottom trawl survey; however, the longline model is still useful to help consider the effects of the biases of the trawl survey and the low catchability of the thorny skate (Runnebaum et al., 2018). Despite the limited spatio‐temporal coverage, the sampled thermal range of survey is unlikely to play a role in the longline results.…”

“…Runnebaum et al. (2018) developed a model standardizing catch between longline and trawl surveys for the data‐limited demersal fish, cusk ( Brosme brosme ). They did this by standardizing area swept in each survey, modeling cusk density for each method, and using that modeled density as an input for calculating a habitat suitability index.…”

“…They did this by standardizing area swept in each survey, modeling cusk density for each method, and using that modeled density as an input for calculating a habitat suitability index. The resulting habitat suitability indices improved on modeling each method separately, albeit with high uncertainty due to multiplication of errors, simplification of survey details, and necessary assumptions (Runnebaum et al., 2018). This approach may supply a roadmap for the next steps in better quantifying how thorny skate will be affected by climate change; however, combining this approach with climate models amplifies uncertainty even further (Hawkins & Sutton, 2009).…”

Examination into a Vessel Effect for a Multi-Vessel Industry-Based Sea Scallop Dredge Survey

“…Advantages for this approach include minimizing variance associated with using multiple vessels by maintaining a consistent sampling platform and having a vessel designed to satisfy data collection needs (Helser et al 2004). Although there are benefits to the deployment of dedicated research vessels, an increasing number of resource surveys are being conducted using a cooperative approach, where commercial fishing vessels are chartered as research platforms (Hanna 1995, Helser et al 2004, Runnebaum et al 2018, Thorson & Ward 2014, Wurtzell et al 2016. This type of approach can decrease research costs and increase stakeholder buy-in to survey, management, and assessment results (Hartley & Robertson 2006, Johnson & van Densen 2007, Wurtzell et al 2016).…”

Quota allocation for stocks that span multiple management zones: analysis with a vector autoregressive spatiotemporal model

“…Recent development of a vector autoregressive spatiotemporal (VAST) model (Thorson & Barnett, 2017) provides a tool to estimate biomass in defined spatial areas and could be useful in quota allocation (Thorson, 2019). In the Northwest Atlantic Ocean, VAST has been used to standardise indices for the northern shrimp ( Pandalus borealis Krøyer) assessment (Cao et al, 2017); to combine data from multiple fishery‐independent surveys to predict density estimates for cusk ( Brosme brosme (Ascanius)) for use in habitat suitability indices (Runnebaum et al, 2018), as well as to predict hotspots of cusk bycatch in the American lobster ( Homarus americanus Milne‐Edwards) fishery (Runnebaum et al, 2020); to test the incorporation of environmental covariates in a length‐structured assessment of the American lobster (Hodgdon et al, 2020); and to examine distribution shifts for cod (Guan et al, 2017) and summer flounder ( Paralichthys dentatus (L.)) in the region (Perretti & Thorson, 2019).…”