This geodatabase includes hotspot maps of 1) nearshore habitat richness, 2) diversity (fish and invertebrates), and 3) biomass (using catch per unit effort of fish and invertebrates), as well as two layers showing the spatial extent of the diversity and biomass hotspot analyses. Full details and methods can be found in the Rubidge et al. 2018 CSAS Research Document 2018/053 available here or at https://waves-vagues.dfo-mpo.gc.ca/Library/40759842.pdf. These data were reviewed as part of a Canadian Science Advisory Secretariat (CSAS) regional peer review process on Nov 1-2, 2017. Habitat Richness Hotspots: Because there are no systematic surveys of nearshore species that span the entire coastline of Northern Shelf Bioregion, the nearshore habitat richness hotspots were developed as a proxy for species diversity in nearshore areas. Habitat richness was calculated from eight habitat features: eelgrass, surfgrass, canopy-forming kelp, estuaries, areas of high rugosity, and hard, mixed, and soft substrate. The number of features within 1 km x 1 km planning units was counted, and hotspots were identified using the Getis-Ord G* tool in ArcGIS. Planning units with Gi_Bin values of 3 (99% confidence) were classified as habitat richness hotspots. Diversity and Biomass Hotspots: Hotspots of fish and invertebrate diversity and biomass were developed as proxies for spatial patterns of productivity in the Northern Shelf Bioregion. Diversity (Shannon diversity) and biomass (kg/hour or count/hook/hour) were calculated from DFO synoptic trawl and outside hard-bottom longline (HBLL) survey catch records. The outside HBLL survey was previously referred to as Pacific Halibut Management Area (PHMA) survey. The synoptic trawl and HBLL surveys have complementary spatial coverage, with the HBLL surveys occurring in more coastal areas (20–260 m) and the synoptic trawl surveys occurring on deeper shelf areas (50–1300 m). Hotspots were identified using the Getis-Ord G* tool in ArcGIS for five separate analyses: fish biomass (trawl), fish diversity (trawl), fish diversity (longline), invertebrate biomass (trawl), and invertebrate diversity (trawl). Using the Minimum Bounding Geometry Tool, convex hull polygons were drawn around groups of hotspot points (Gi_Bin values of 1, 2, or 3; confidence ≥90%) containing 10 or more points. The resulting polygons were then buffered by 1 km and manually edited where needed to exclude any large areas of the polygons that did not include hotspot points.

Description: Biophysical Units: Under the Pacific Marine Ecological Classification System (PMECS; DFO 2016; Rubidge et al. 2016), biophysical units are areas of distinct physiographic and oceanographic conditions and processes that shape species composition at spatial extents of 1000s of km. Geomorphic units: Geomorphic units or geozones are discrete geomorphological structures at the scale of 100s of km that are assumed to have distinctive biological assemblages (e.g., plateaus, ridges, seamounts, canyons). Although the spatial scale of geomorphic units is nested within biophysical units, a single geomorphic unit such as a trough may span more than one biophysical unit. The following 5 layers are included in this geodatabase: 1. Biophysical_Units_L4A - Predicted PMECS Biophysical Units (Level 4A) output from the random forest analysis 2. Biophysical_Units_L4B - Predicted PMECS Biophysical Units (Level 4B) output from the random forest analysis 3. Biophysical_Units_ProbAssign_L4AB - Layer showing the probability that a grid cell was assigned to a given biophysical unit in the final random forest predictive modelling step 4. Cluster_L4AB - Layer showing the output of species assemblage cluster analysis 5. Geomorphic_Units - Geomorphic units for the BC coast that combines geomorphic units produced by Rubidge et al. 2016) and Proudfoot and Robb (2022). Methods: Biophysical Units: Rubidge et al. (2016) used a two-step process to identify biophysical units in British Columbia. First, a cluster analysis based on the similarity of species composition was used to group sites with similar species into distinct biological assemblages. Second, a random forest analysis was used to identify environmental correlates of the biological assemblages identified by the cluster analysis and to predict and assign the biological assemblage present in areas with too few biological data. Two different similarity thresholds were used to identify two levels (4A, 4B) of biophysical units; see Rubidge et al. (2016) for details. Indicator species for each assemblage (biophysical unit) were also identified. Geomorphic units: Rubidge et al. (2016) used the benthic terrain modeller (BTM) tool with broad and fine-scale benthic positioning index (BPI) parameters to define geomorphic units on the continental shelf in the Northern Shelf Bioregion and the continental slope in both the Northern Shelf Bioregion and Southern Shelf Bioregion. In 2022, geomorphic units were produced for the Strait of Georgia and Southern Shelf Bioregions following the same methods as Rubidge et al. (2016) (Proudfoot and Robb 2022). The geomorphic units produced as part of the PMECS process were merged with the geomorphic units produced for the Strait of Georgia and Southern Shelf bioregions to produce a continuous spatial data product representing geomorphic units for the Canadian Pacific continental shelf and slope. After merging, the geomorphic units produced in 2016 were unchanged (i.e., they are consistent with the original geomorphic units described in Rubidge et al. 2016). Data Sources: From Rubidge et al. (2016): Species data was taken from Fisheries and Oceans Canada (DFO) standardized fisheries-independent research surveys: groundfish trawl and long-line (2003-2013), Tanner Crab trawl and trap (2000–2006), and Dungeness Crab trap (2000–2014). Environmental data came from NASA, the Canadian Hydrographic Service, Fisheries and Oceans Canada, Bio-ORACLE, and elsewhere (details in Rubidge et al. 2016). From Proudfoot and Robb (2022): bathymetry data came from Natural Resources Canada (details in Proudfoot and Robb 2022). Uncertainties: The data is intended for use at the bioregional scale, and caution should be used for finer-scale analyses.