Description: These datasets include a variety of marine managed areas along the U.S. West Coast, which include various state managed areas, Essential Fish Habitat (EFH) designations, Habitat Areas of Particular Concern (HAPC), National Marine Sanctuaries (NMS), and BOEM leasing data.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA), Bureau of Ocean Energy Management (BOEM)
Description: Maritime limits for the United States are measured from the U.S. baseline, recognized as the low-water line along the coast as marked on NOAA's nautical charts in accordance with the articles of the Law of the Sea. The baseline and related maritime limits are reviewed and approved by the interagency U.S. Baseline Committee. The primary purpose of this dataset is to update the official depiction of these maritime limits and boundaries on NOAA's nautical charts. The Office of Coast Survey depicts on its nautical charts the territorial sea (12 nautical miles), contiguous zone (24nm), and exclusive economic zone (200nm, plus maritime boundaries with adjacent/opposite countries). U.S. maritime limits are ambulatory and subject to revision based on accretion or erosion of the charted low water line. To ensure you are up-to-date and for more information about U.S. Maritime Limits and Boundaries, see http://www.nauticalcharts.noaa.gov/csdl/mbound.htm. For the full FGDC metadata record, see http://www.ncddc.noaa.gov/approved_recs/nos_de/ocs/ocs/ocs/MB_ParentDataset.html. Coordinates for the US/Canada international boundary, on land and through the Great Lakes, are managed by the International Boundary Commission. These boundaries are included with this dataset for continuity.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Ocean Service (NOS) Office of Coast Survey (OCS)
Description: Maritime limits for the United States are measured from the U.S. baseline, recognized as the low-water line along the coast as marked on NOAA's nautical charts in accordance with the articles of the Law of the Sea. The baseline and related maritime limits are reviewed and approved by the interagency U.S. Baseline Committee. The primary purpose of this dataset is to update the official depiction of these maritime limits and boundaries on NOAA's nautical charts. The Office of Coast Survey depicts on its nautical charts the territorial sea (12 nautical miles), contiguous zone (24nm), and exclusive economic zone (200nm, plus maritime boundaries with adjacent/opposite countries). U.S. maritime limits are ambulatory and subject to revision based on accretion or erosion of the charted low water line. To ensure you are up-to-date and for more information about U.S. Maritime Limits and Boundaries, see http://www.nauticalcharts.noaa.gov/csdl/mbound.htm.For the full FGDC metadata record, see http://www.ncddc.noaa.gov/approved_recs/nos_de/ocs/ocs/ocs/MB_ParentDataset.html.Coordinates for the US/Canada international boundary, on land and through the Great Lakes, are managed by the International Boundary Commission. These boundaries are included with this dataset for continuity.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA), National Ocean Service (NOS), Office of Coast Survey (OCS)
Description: This data set contains BOEM Planning Area outlines in ESRI shapefile format for the BOEM Alaska Region. The Submerged Lands Act (SLA) boundary, along with the Continental Shelf Boundary (CSB), the Limit of Protraction were used to complete the polygons for the Planning Areas. Because GIS projection and topology functions can change or generalize coordinates, and because shapefiles can not represent true arcs, these GIS files are considered to be approximate and are NOT an OFFICIAL record for the exact block coordinates or areas. The Official Protraction Diagrams (OPDs) and Supplemental Official OCS Block Diagrams (SOBDs) serve as the legal definition for BOEM offshore boundary coordinates and area descriptions. If any discrepancies are found between these shapefiles and the OPDs and SOBDs, it is the OPD and SOBD diagrams which take precidence. Also note that the BOEM cadastre is developed on a UTM projection, and most Planning Areas span multiple UTM zones. This means that area values computed from these shapefiles will not match the official BOEM areas.
Copyright Text: Bureau of Ocean Energy Management (BOEM)
Description: This file represents the active leases for federal OCS oil and gas leases in the Alaska OCS Region.OCS Lease Blocks that are currently leased from the federal government by industry for the purpose of development of traditional oil or gas energy products and may or may not be actively developed or producing. Leases in state waters are not included in this layer. The Outer Continental Shelf Lands Act (OCSLA) (43 U.S.C. 1331-1356a), as amended, authorizes the Secretary of the Interior to issue, on a competitive basis, leases for oil and gas, and sulfur, in submerged lands of the Outer Continental Shelf. The Act authorizes the Secretary to grant rights-of-way and easements through the submerged lands of the OCS.An OCS lease is an agreement that is issued under section 8 or maintained under section 6 of the Act and that authorizes exploration for, and development and production of, minerals on the OCS. The term also means the area covered by that agreement, whichever the context requires. The areas covered by the lease agreement consist of one or more OCS Lease Blocks, or any leasable portion thereof, bid upon as a single administrative unit, that will be part of a single lease.The data represents a close approximation, but might not be the exact coordinates for leases. For the official coordinates please refer to the OPD or SOBD for the lease. ALSO NOTE: older leases were issued on a NAD 27 grid, and their lease coordinates will not change as long as the lease remains active. Although the NAD 27 leases have been projected to NAD 83 for inclusion in this file, when plotted on the current NAD 83 leasing grid, they will not appear to "line up", because the NAD 83 grids represents a complete re-grid and not a re-projection of the old NAD 27 grid. All current leasing, starting with Sale 144 in September 1996, and leases that were issued, were done on the NAD 83 grid.
Copyright Text: BOEM
Leasing GIS
Alaska OCS Region
3801 Centerpoint Drive
Anchorage, AK 99503
Description: This data set contains OCS block outlines in ArcGIS shapefile format for the BOEM Alaska Region. OCS blocks are used to define small geographic areas within an Official Protraction Diagram (OPD) for leasing and administrative purposes. These blocks have been clipped along the Submerged Lands Act (SLA) boundary and along the Continental Shelf Boundaries. Additional details are available from: http://www.boem.gov/BOEM-Newsroom/Library/Publications/1999/99-0006-pdf.aspx Because GIS projection and topology functions can change or generalize coordinates, and because shapefiles can not represent true arcs, these GIS files are considered to be approximate and are NOT an OFFICIAL record for the exact block coordinates or areas. The Official Protraction Diagrams (OPDs) and Supplemental Official Block Diagrams (SOBDs) serve as the legal definition for BOEM offshore boundary coordinates and area descriptions.
Copyright Text: Bureau of Ocean Energy Management (BOEM)
Description: Essential Fish Habitat are those areas that have been identified and described by species and lifestage. Fish require healthy surroundings to survive and reproduce. Essential fish haitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or growh to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
This data layer displays Habitat Areas of Particular Concern (HAPCs), which are smaller habitat areas within EFH that meet at least two of the four considerations: 1) The importance of the ecological function provided by the habitat; 2) The extent to which the habitat is sensitive to human-induced environmental degradation; 3) Whether, and to what extent, development activities are, or will be, stressing the habitat type; 4) The rarity of the habitat type. (Rarity is a mandatory criterion of all Council HAPC proposals.)
Description: Essential Fish Habitat are those areas that have been identified and described by species and lifestage. These layers are grouped by taxonomic order and include EFH designations for Gadiforms (Cods), Decapods (Crabs), Plueronectiformes (Flatfishes), Octopods (Octopus), Salmoniformes (Salmon), Scorpaeniformes (Scorpionfishes and flatheads), Rajiformes (Skates), and the Weathervane Scallop.
Fish require healthy surroundings to survive and reproduce. Essential fish haitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or growh to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Gadiformes (Cods) including: Arctic cod, Pacific cod, Saffron cod, and Walleye pollock. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Decapods (Crabs) including: Arctic crab, Blue king crab, Golden king crab, Red king crab, Snow crab, and Tanner crab. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Pleuronectiformes (Flatfishes) including: Alaska plaice, Arrowtooth Flounder, Dover sole, Flathead sole, Greenland turbot, Kamchatka flounder, Northern rock sole, Rex sole, Southern rock sole, and Yellowfin sole. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Octopods (Octopus) within the Aleutian Islands, Eastern Bering Sea, and the Gulf of Alaska. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Salmoniformes (Salmon) including: Chinook, Chum, Coho, Pink and Sockeye Salmon. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Scorpaeniformes (Scorpionfishes and Flatheads) including: Atka mackerel, Bigmouth sculpin, Black rockfish, Blackspotted rockfish, Dark rockfish, Dusky rockfish, Great sculpin, Greenstriped rockfish, Harlequin rockfish, Longspine thornyhead rockfish, Northern rockfish, Pacific ocean perch, Pygmy rockfish, Quillback rockfish, Redbanded rockfish, Redstriped rockfish, Rosethorn rockfish, Rougheye rockfish, Sablefish, Sharpchin rockfish, Shortraker rockfish, Shortspine thornyhead rockfish, Silvergrey rockfish, Yellow Irish lord, and Yelloweye rockfish.
Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for Rajiformes (skates and rays) including: Alaska skate, Aleutian skate, Bering skate, and Mud skate. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: Essential Fish Habitat (EFH) are those areas that have been identified and described by species and lifestage. The areas displayed represent EFH designations for the Weathervane Scallop. Fish require healthy surroundings to survive and reproduce. Essential fish habitat includes all types of aquatic habitat - wetlands, coral reefs, sea-grasses, rivers - where fish spawn, breed, feed, or grow to maturity. EFH Mapper Data is a combination of three existing data layers: Essential Fish Habitat (EFH), Habitat Areas of Particular Concern (HAPC) and EFH Areas Protected from Fishing (EFHA).
Description: The MPA Inventory is a comprehensive catalog that provides detailed information for existing marine protected areas in the United States. The inventory provides geospatial boundary information (in polygon format) and classification attributes that seek to define the conservation objectives, protection level, governance and related management criteria for all sites in the database. The comprehensive inventory of federal, state and territorial MPA sites provides governments and stakeholders with access to information to make better decisions about the current and future use of place-based conservation. The information also will be used to inform the development of the national system of marine protected areas as required by Executive Order 13158.
Copyright Text: NOAA Marine Protected Areas Center in joint effort with the US Department of the Interior
Description: The MPA Inventory is a comprehensive catalog that provides detailed information for existing marine protected areas in the United States. The inventory provides geospatial boundary information (in polygon format) and classification attributes that seek to define the conservation objectives, protection level, governance and related management criteria for all sites in the database. The comprehensive inventory of federal, state and territorial MPA sites provides governments and stakeholders with access to information to make better decisions about the current and future use of place-based conservation. The information also will be used to inform the development of the national system of marine protected areas as required by Executive Order 13158.
These grouped layers are displayed according to managing agency.
Copyright Text: NOAA Marine Protected Areas Center in joint effort with the US Department of the Interior
Description: The MPA Inventory is a comprehensive catalog that provides detailed information for existing marine protected areas in the United States. The inventory provides geospatial boundary information (in polygon format) and classification attributes that seek to define the conservation objectives, protection level, governance and related management criteria for all sites in the database. The comprehensive inventory of federal, state and territorial MPA sites provides governments and stakeholders with access to information to make better decisions about the current and future use of place-based conservation. The information also will be used to inform the development of the national system of marine protected areas as required by Executive Order 13158.
This layer depicts areas managed by the National Park Service.
Copyright Text: NOAA Marine Protected Areas Center in joint effort with the US Department of the Interior
Description: The MPA Inventory is a comprehensive catalog that provides detailed information for existing marine protected areas in the United States. The inventory provides geospatial boundary information (in polygon format) and classification attributes that seek to define the conservation objectives, protection level, governance and related management criteria for all sites in the database. The comprehensive inventory of federal, state and territorial MPA sites provides governments and stakeholders with access to information to make better decisions about the current and future use of place-based conservation. The information also will be used to inform the development of the national system of marine protected areas as required by Executive Order 13158.
This layer depicts areas managed by the Alaska Department of Fish and Game.
Copyright Text: NOAA Marine Protected Areas Center in joint effort with the US Department of the Interior
Description: The MPA Inventory is a comprehensive catalog that provides detailed information for existing marine protected areas in the United States. The inventory provides geospatial boundary information (in polygon format) and classification attributes that seek to define the conservation objectives, protection level, governance and related management criteria for all sites in the database. The comprehensive inventory of federal, state and territorial MPA sites provides governments and stakeholders with access to information to make better decisions about the current and future use of place-based conservation. The information also will be used to inform the development of the national system of marine protected areas as required by Executive Order 13158.
This layer depicts areas managed by the U.S. Fish and Wildlife Service (USFWS).
Copyright Text: NOAA Marine Protected Areas Center in joint effort with the US Department of the Interior
Description: The MPA Inventory is a comprehensive catalog that provides detailed information for existing marine protected areas in the United States. The inventory provides geospatial boundary information (in polygon format) and classification attributes that seek to define the conservation objectives, protection level, governance and related management criteria for all sites in the database. The comprehensive inventory of federal, state and territorial MPA sites provides governments and stakeholders with access to information to make better decisions about the current and future use of place-based conservation. The information also will be used to inform the development of the national system of marine protected areas as required by Executive Order 13158.
This layer depicts areas jointly managed by the Alaska Department of Fish and Game and the National Oceanic and Atmospheric Administration (NOAA).
Copyright Text: NOAA Marine Protected Areas Center in joint effort with the US Department of the Interior
Description: Surface Locations of Boreholes drilled for exploration or oil and gas production. Dataset is maintained by Bureau of Ocean Energy Management. Please note: older well locations were mapped on a NAD27 grid and have been projected to NAD 83 for inclusion in this file. When plotted on the current NAD 83 grid, they will not appear to "line up", because the NAD 83 grid represents a complete re-grid, and not a re-projection of the old NAD 27 grid. All current wells have been plotted on a NAD 83 grid.
Copyright Text: Bureau of Ocean Energy Management (BOEM)
Description: These data depict the occurrence of submarine cables in and around U.S. navigable waters. The purpose of this data product is to support coastal planning at the regional and national scale. Source geometry and attributes were derived from 2010 NOAA Electronic Navigation Charts and 2009 NOAA Raster Nautical Charts. Polyline features explicitly defined as cables were compiled from the original sources, exclusive of those features noted as 'cable areas'. The scale of the source material was highly variable and discontinuities between multiple sources were resolved with least possible spatial adjustments. The original S-57 data model was modified for readability and performance.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA)
Description: Various shipping zones delineate activities and regulations for marine vessel traffic. Traffic lanes define specific traffic flow, while traffic separation zones assist opposing streams of marine traffic. Precautionary areas represent areas where ships must navigate with caution, and shipping safety fairways designate where artificial structures are prohibited. Recommended Routes are predetermined routes for shipping adopted for reasons of safety. Along certain zones of the East Coast of the United States, ships are required to reduce speeds to 10 knots or less over ground during seasonal periods within designated endangered species areas, such as the North Atlantic Right Whales. Particularly Sensitive Sea Areas need special protection because of their vulnerability to damage by international maritime activities. Areas to be avoided are within defined limits where navigation is particularly hazardous or it is exceptionally important to avoid casualties and should be avoided by all ships or certain classes of ships.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) Office of Coast Survey
Description: This group of layers includes site locations of camera image data collected on a variety of platforms (ROVs, submarine, drop camera) from to 1988 to 2017 as well as predictive habitat grids of deep-sea corals and sponges within the Aleutian Islands, Eastern Bering Sea, and Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) Alaska Fisheries Science Center
Description: This group of layers includes results from a study entitled "Predictive models of coral and sponge distribution, abundance and diversity in bottom trawl surveys of the Aleutian Islands, Alaska". Results include predictions of the best-fitting generalized additive model for upright sponge, coral, Primnoidae and Stylasteridae predicting the probability of presence in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) Alaska Fisheries Science Center
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for upright sponge predicting the probability of presence in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for coral predicting the probability of presence in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for Primnoidae predicting the probability of presence in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for Stylasteridae predicting the probability of presence in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for predicting the abundance of upright sponge (log-transformed catch per unit of effort [CPUE]) in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.This grid displays predictions of the best-fitting generalized additive model for predicting the abundance of coral in the Aleutian Islands bottom trawl surveys.
This grid displays predictions of the best-fitting generalized additive model for predicting the abundance of coral (log-transformed catch per unit of effort [CPUE]) in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for predicting the abundance of Primnoidae (log-transformed catch per unit of effort [CPUE]) in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for predicting the abundance of Stylasteridae (Catch Per Unit Effort [CPUE]) in the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: Ecosystem management requires information to determine and mitigate adverse impacts of fishing on all ecosystem components. Deep-sea coral and sponge ecosystems often co-occur with fishing activities, and there is considerable research documenting the vulnerability and slow recovery of deep-sea coral and sponge communities to damage. The objective of this analysis was to construct models that could predict the distribution, abundance and diversity of deep sea corals and sponges in the Aleutian Islands. Generalized additive models were constructed based on bottom trawl survey data collected from 1991 to 2011 and tested on data from 2012. The results showed that deep-sea coral and sponge distributions were strongly influenced by the maximum tidal currents at bottom trawl locations, possibly indicative of reduced sedimentation or increased food-delivery processes near the seafloor in areas of moderate to high current. Depth and location were also important factors affecting the distribution of deep-sea sponges and corals. The analysis resulted in acceptable models of presence or absence for all taxonomic groups and similar fits when models were applied to test data. The best-fitting models of abundance explained between 20 and 25% of the deviance in the abundance data. Current management protects ~50% of the coral and sponge habitat in the Aleutian Islands at depths to 500 m. The models constructed here will allow managers to evaluate ecological versus economic benefits between protecting coral and sponge habitat and allowing commercial fishing by examining the effect of spatial closures on the amount of coral and sponge habitat that is protected.
This grid displays predictions of the best-fitting generalized additive model for coral diversity predicting the number of families of coral represented in bottom trawl hauls during the Aleutian Islands bottom trawl surveys.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center; Ocean Associates, Inc.; Pacific Marine Environmental Laboratory
Description: This group of layers includes results from a study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey". Results include predictions of sponge, coral, and sea whip habitat within the eastern Bering Sea.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) Alaska Fisheries Science Center
Description: This layer displays the average probability of coral presence from the weighted predictions of the best-fitting generalized additive models of presence or absence from the trawl survey data and camera survey data collected in 2014. The predictions were averaged by weighting with the inverse of the prediction error.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the average probability of sponge presence from the weighted predictions of the best-fitting generalized additive models of presence or absence from the trawl survey data and camera survey data collected in 2014. The predictions were averaged by weighting with the inverse of the prediction error.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the average probability of sea whip presence from the weighted predictions of the best-fitting generalized additive models of presence or absence from the trawl survey data and camera survey data collected in 2014. The predictions were averaged by weighting with the inverse of the prediction error.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Name: Probability of Coral Presence - Camera Survey
Display Field:
Type: Raster Layer
Geometry Type: null
Description: This layer displays the probability of coral presence from the weighted predictions of the best-fitting generalized additive models of presence or absence from the camera survey data collected in 2014.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Name: Probability of Sponge Presence - Camera Survey
Display Field:
Type: Raster Layer
Geometry Type: null
Description: This layer displays the probability of sponge presence from the weighted predictions of the best-fitting generalized additive models of presence or absence from the camera survey data collected in 2014.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Name: Sea Whips - Probability of Presence - Camera
Display Field:
Type: Raster Layer
Geometry Type: null
Description: This layer displays the probability of sea whips presence from the weighted predictions of the best-fitting generalized additive models of presence or absence from the camera survey data collected in 2014.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the predicted fourth-root transformed density (number per square meter) of corals based on generalized additive models of camera survey density data. Predictions are shown for only those grid cells where the average presence-absence model indicated that the invertebrate taxa would be present.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the predicted fourth-root transformed density (number per square meter) of sponges based on generalized additive models of camera survey density data. Predictions are shown for only those grid cells where the average presence-absence model indicated that the invertebrate taxa would be present.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the predicted fourth-root transformed density (number per square meter) of sea whips based on generalized additive models of camera survey density data. Predictions are shown for only those grid cells where the average presence-absence model indicated that the invertebrate taxa would be present.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the predicted heights (centimeters) of corals based on generalized additive models of camera survey density data. Predictions are shown for only those grid cells where the average presence-absence model indicated that the invertebrate taxa would be present.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the predicted heights (centimeters) of sponges based on generalized additive models of camera survey density data. Predictions are shown for only those grid cells where the average presence-absence model indicated that the invertebrate taxa would be present.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This layer displays the predicted heights (centimeters) of sea whips based on generalized additive models of camera survey density data. Predictions are shown for only those grid cells where the average presence-absence model indicated that the invertebrate taxa would be present.
This predictive surface is a component of the results from a NOAA Alaska Fisheries Science Center study entitled "Validation and improvement of species distribution models for structure-forming invertebrates in the eastern Bering Sea with an independent survey".
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: This group of layers includes results from a study entitled "Comparison of modeling methods to predict the spatial distribution of deepsea
coral and sponge in the Gulf of Alaska". Results include predictions of deep-sea sponge, coral, and sea whip habitat within the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) Alaska Fisheries Science Center
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays mean predicted probability of presence of corals in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Name: Probability of Hexactinellid Sponge Presence
Display Field:
Type: Raster Layer
Geometry Type: null
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays mean predicted probability of presence of Hexactinellid sponges in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays mean predicted probability of presence of sea whips in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays the predicted abundance or mean catch per unit effort (CPUE) of corals in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays the predicted abundance or mean catch per unit effort (CPUE) of Demosponges in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays the predicted abundance or mean catch per unit effort (CPUE) of Primnoids in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: Deep-sea coral and sponge ecosystems are widespread throughout most of Alaska's marine waters, and are associated with many different species of fishes and invertebrates. These ecosystems are vulnerable to the effects of commercial fishing activities and climate change. This work compared four commonly used species distribution models (general linear models, generalized additive models, boosted regression trees and random forest models) and an ensemble model to predict the presence or absence and abundance of six groups of benthic invertebrate taxa in the Gulf of Alaska. All four model types performed adequately on training data for predicting presence and absence, with regression forest models having the best overall performance measured by the area under the receiver-operating-curve (AUC). The models also performed well on the test data for presence and absence with average AUCs ranging from 0.66 to 0.82. For the test data, ensemble models performed the best. For abundance data, there was an obvious demarcation in performance between the two regression-based methods (general linear models and generalized additive models), and the tree-based models. The boosted regression tree and random forest models out-performed the other models by a wide margin on both the training and testing data. However, there was a significant drop-off in performance for all models of invertebrate abundance (~50%) when moving from the training data to the testing data. Ensemble model performance was between the tree-based and regression-based methods. The maps of predictions from the models for both presence and abundance agreed very well across model types, with an increase in variability in predictions for the abundance data. We conclude that where data conforms well to the modeled distribution (such as the presence-absence data and binomial distribution in this study), the four types of models will provide similar results, although the regression-type models may be more consistent with biological theory. For data with highly zero-inflated distributions and nonnormal distributions such as the abundance data from this study, the tree-based methods performed better. Ensemble models that averaged predictions across the four model types, performed better than the GLM or GAM models but slightly poorer than the tree-based methods, suggesting ensemble models might be more robust to overfitting than tree methods, while mitigating some of the disadvantages in predictive performance of regression methods.
This grid displays the predicted abundance or mean catch per unit effort (CPUE) of sponges in the Gulf of Alaska.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS) Alaska Fisheries Science Center
Description: These datasets include multibeam echosounder (MBES) survey data from the National Marine Fisheries Service as well as depth contours and a mosaic bathymetry grid created from the GEBCO 2019 gridded bathymetric data set.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS); General Bathymetric Chart of the Oceans (GEBCO)
Name: MBES Sun-Illuminated Bathymetry and Backscatter
Display Field: SurveyID
Type: Feature Layer
Geometry Type: esriGeometryPolygon
Description: This layer depicts footprints of multibeam echosounder (MBES) survey data collected in 2003 by Thales GeoSolutions (Pacific), Inc., contracted by the National Marine Fisheries Service (NMFS), on select sites in the Aleutian Archipelago of Alaska in the North Pacific Ocean and Bering Sea. The data collected provides the first detailed mapping of deep sea coral and sponge habitats for the Aleutian Islands. It also provides estimates of the relative abundance of corals and sponges, their importance to commercially valuable fish and invertebrates, and the degree to which these living substrates have been disturbed, including disturbance by fishing gear. The survey required digital, high-resolution multibeam bathymetry along with calibrated backscatter in all survey areas. The survey consisted of seventeen survey sites, located along the center of the Aleutian Archipelago of Alaska in both the North Pacific Ocean and Bering Sea. Data was also collected over an additional eighteen dive sites while the vessel was transiting between all survey and dive sites.
Copyright Text: National Oceanic and Atmospheric Administration (NOAA) National Marine Fisheries Service (NMFS)
Description: This data layer features contours derived from GEBCO’s gridded bathymetric data set, the GEBCO_2019 grid. It is a global terrain model for ocean and land at 15 arc-second intervals.
Copyright Text: General Bathymetric Chart of the Oceans (GEBCO)
Description: GEBCO’s gridded bathymetric data set, the GEBCO 2019 grid, is a global terrain model for ocean and land at 15 arc-second intervals. Depth is displayed in meters.
Copyright Text: General Bathymetric Chart of the Oceans (GEBCO)