The following methods are excerpted from Rogers et al. (in press):
\nA trip was classified as belonging to a community if it shared the community's gear type and landing port, and the vessel either declared that port as its principal port or landed in that port at least 50% of its trips that year.
Once aggregated into communities, trips were then weighted by a variable (\u201cfisherdays\u201d) indicating labor time expended on each trip: trip length (in days) multiplied by the number of crew on board. Fisherdays indicate how important an area at sea is to a community in terms of how much time they invest in that location.
\nGiven reported trip locations and fisherdays, we then created raster maps using a kernel density method. The resultant maps distribute fisherdays using different size kernels depending upon the fishery/gear-type/length. Nearshore fishing was processed using a smaller kernel (7.5 - 10 km) than offshore fishing (10 - 15 km). We used the area defined by a 90% volume contour (i.e., an area which encompasses 90% of fisherdays) to define the customary fishing grounds or servicesheds for a community.
\nTo compare the relative historical importance of particular species to a community-at-sea, landings data were compiled from vessel trip reports and summed over the available years of data for each community. Price information was extracted from NOAA Fisheries, Fisheries Statistics Division (https://www.st.nmfs.noaa.gov/st1/commercial/landings/annual_landings.html). We used the average price per lb by species, adjusted for inflation (real 2014 prices in US$), over the period for which we had community-level data. State-level prices were used when available, and otherwise regional prices were used.
\nWe assessed a community's exposure to risk based on their historical dependence on species and spatial fishing patterns. A community was more exposed to risk if the species from which it historically earned the most revenue were projected to lose habitat in the locations where the community has traditionally fished. Specifically, risk exposure scores for communities were calculated as:
\n![\"communityAttributesRisk_Formula.png\"](\"https://datadocs.bco-dmo.org/docs/CC_Fishery_Adaptations/data_docs/765477/communityAttributesRisk_Formula.png\")
where S\u209b,\ua700 is the mean projected change in habitat suitability for species s across the serviceshed of community c, and pRev\u209b,\ua700 is the proportion of historical revenues from fishing that the community has derived from species s. Positive risk exposure scores indicated expanding opportunities for communities based on their historical fishing revenue portfolios and projected changes to species habitat at sea, while negative values indicated shrinking opportunities and increased exposure to negative impacts of climate change.
\nThe R file, \"Servicesheds.rData\" (see\u00a0\"Supplemental Documents\" below) is\u00a0a spatial polygon dataframe (SPDF) giving 90% volume contours of fisher-days at sea for 98 communities-at-sea. The polygons outline the at-sea \"servicesheds\" or customary fishing grounds of communities. We use \"serviceshed\"\u00a0to describe the area from which a community has historically received ecosystem services, specifically fish in this case. The file is intended to be read by the program \"R\",\u00a0with data stored in the SPDF object \"Servicesheds\".