Using species distribution models to describe essential fish habitat in Alaska

Abstract

Describing essential habitat is an important step toward understanding and conserving harvested species in ecosystem-based fishery management. Using data from fishery-independent ichthyoplankton, groundfish surveys, and commercial fisheries observer data, we utilized species distribution modeling techniques to predict habitat-based spatial distributions of federally managed species in Alaska. The distribution and abundance maps were used to refine existing essential fish habitat descriptions for the region. In particular, we used maximum entropy and generalized additive modeling to delineate distribution and abundance of early (egg, larval, and pelagic juvenile) and later (settled juvenile and adult) life history stages of groundfishes and crabs across multiple seasons in three large marine ecosystems (Gulf of Alaska, eastern Bering Sea, and Aleutian Islands) and the northern Bering Sea. We present a case study, featuring Kamchatka flounder (Atheresthes evermanni), from the eastern and northern Bering Sea to represent the >400 habitat-based distribution maps generated for more than 80 unique species–region–season–life-stage combinations. The results of these studies will be used to redescribe essential habitat of federally managed fishes and crabs in Alaska.

Spatio-temporal patterns in juvenile habitat for 13 groundfishes in the California Current Ecosystem

“…While not examined here, bottom type, biogenic habitat, temperature, frontal boundaries, and other physical characteristics are known to be important in influencing species distributions, community structure, and the location of juvenile habitats [ 52 , 56 – 58 ]. For demersal fishes early, post-settlement mortality is often high, density-dependent and mediated via competition for shelter [ 59 ].…”

“…Species distribution modes have been increasingly used to define essential fish habitat in regions managed by the U.S. National Marine Fisheries Service [58,78], as mandated by the Magnuson-Stevens Act. For example, Laman et al [58] estimated habitat-based models for Alaskan species. Our analyses increase the information available to fisheries management practices of the 13 stocks we analyzed.…”

Model of trawlable area using benthic terrain and oceanographic variables—Informing survey design and habitat maps in the Gulf of Alaska

“…To date, model output has been applied to inform camera‐based survey of demersal shelf rockfish in this region (Rooper, Boldt, Conrath, & Baker, ) and analyses of species association with structural habitat (Rooper et al, ; Rooper, Boldt, Conrath, & Baker, ). Many of these predictive variables and bathymetric derivatives have also been used in species distribution models using MaxEnt and GAM approaches to predict essential fish habitat (Laman et al, ; Rooney et al, ). These analyses and related products (e.g., maps, variable profiles, and threshold breakpoint estimates) have been reviewed in the context of improving rockfish stock assessments and management recommendations (Alaska Fisheries Science Center Rockfish Working Group annual meetings 2015–2018, Cunningham, Hulson, Lunsford, & Hanselman, ).…”

Bathymetry and Canyons of the Eastern Bering Sea Slope

“…To meet this need, we have been publishing 100 m resolution bathymetry compilations in recent years: AI [3], Cook Inlet (50 m resolution: [16]), Norton Sound [17], and central Gulf of Alaska (GOA, AK, USA) [18]. These bathymetry compilations have been utilized for a variety of fishery research purposes including fish vertical migration [19]; coral and sponge distribution modeling in the AI [20] and GOA [21]; quantifying inshore study sites in the central GOA [22], eastern GOA [23], and bathymetry groundtruthing [24]; bathymetric steering of seafloor current flow [25]; inshore habitat loss [26]; Essential Fish Habitat modeling in the EBS [27], GOA [28,29], and AI [30]; juvenile groundfish habitat suitability models [31]; and capelin (Mallotus villosus) distribution modeling [32].…”