16 Habitat Diversity
Description: Species richness was derived from the Northeast Regional Habitat Assessment models for 55 common species sampled by the 2000-2019 spring and fall NEFSC bottom trawl surveys. The joint species distribution model controls for differences in capture efficiency across survey vessels.
Indicator category: Extensive analysis
Found in: State of the Ecosystem - New England (2023), State of the Ecosystem - Mid-Atlantic (2023)
Contributor(s): Chris Haak ChrisHaak@gmail.com
Data steward: Laurel Smith laurel.smith@noaa.gov
Point of contact: Laurel Smith laurel.smith@noaa.gov
Public availability statement: This analysis is based on NEFSC bottom trawl survey data which are publicly available. Please reached out to Laurel Smith with questions.
16.1 Methods
16.1.1 Data sources
Abundance data were extracted from the NEFSC’s SVDBS database using Survdat for 55 fish species regularly sampled on spring and fall NEFSC bottom trawl surveys:
Species included in NRHA Diversity Index:
Common Name | Scientific Name |
---|---|
Acadian Redfish | Sebastes fasciatus |
Alewife | Alosa pseudoharengus |
American Lobster | Homarus americanus |
American Plaice | Hippoglossoides platessoides |
American Shad | Alosa sapidissima |
Atlantic Cod | Gadus morhua |
Altantic Croaker | Micropogonias undulatus |
Atlantic Herring | Clupea harengus |
Atlantic Mackerel | Scomber scombrus |
Barndoor Skate | Dipturus laevis |
Black Sea Bass | Centropristis striata |
Blackbelly Rosefish | Helicolenus dactylopterus |
Blueback Herring | Alosa aestivalis |
Bluefish | Pomatomus saltatrix |
Butterfish | Peprilus triacanthus |
Chain Dogfish | Scyliorhinus retifer |
Clearnose Skate | Rostroraja eglanteria |
Fawn Cusk Eel | Lepophidium profundorum |
Fourbeard Rockling | Enchelyopus cimbrius |
Fourspot Flounder | Hippoglossina oblonga |
Goosefish | Lophius americanus |
Gulf Stream Flounder | Citharichthys arctifrons |
Haddock | Melanogrammus aeglefinus |
Horseshoe Crab | Limulus polyphemus |
Jonah Crab | Cancer borealis |
Little Skate | Leucoraja erinaceus |
Longfin Squid | Doryteuthis (Amerigo) pealeii |
Longhorn Sculpin | Myoxocephalus octodecemspinosus |
Northern Searobin | Prionotus carolinus |
Northern Shortfin Squid | Illex illecebrosus |
Northern Shrimp | Pandalus borealis |
Ocean Pout | Zoarces americanus |
Offshore Hake | Merluccius albidus |
Pollock | Pollachius pollachius |
Red Hake | Urophycis chuss |
Rosette Skate | Leucoraja garmani |
Scup | Stenotomus chrysops |
Sea Raven | Hemitripterus americanus |
Sea Scallop | Placopecten magellanicus |
Silver Hake | Merluccius bilinearis |
Smooth Dogfish | Mustelus canis |
Smooth Skate | Malacoraja senta |
Spiny Dogfish | Squalus acanthias |
Spot | Leiostomus xanthurus |
Spotted Hake | Urophycis regia |
Striped Searobin | Prionotus evolans |
Summer Flounder | Paralichthys dentatus |
Thorny Skate | Amblyraja radiata |
Weakfish | Cynoscion regalis |
White Hake | Urophycis tenuis |
Windowpane Flounder | Scophthalmus aquosus |
Winter Flounder | Pseudopleuronectes americanus |
Winter Skate | Leucoraja ocellata |
Witch Flounder | Glyptocephalus cynoglossus |
Yellowtail Flounder | Myzopsetta ferruginea |
Data were converted to presence/absence for species richness modeling.
16.1.2 Data analysis
16.1.2.1 Species Richness
Estimated species richness is the number of unique species expected to be observed in NEFSC bottom trawl surveys conducted in a given ecological production unit (EPU) and year, based on a fitted joint-species distribution/habitat suitability model (considering only the 55 commonly-occurring species listed above).
16.1.2.2 Model Fitting
A spatiotemporal joint species distribution model was fitted to n=13231 observations of presence/absence in the Spring and Fall NEFSC bottom trawl surveys for the years 2000-2019, using the Community Level Basis Function Model (CBFM) framework with a binomial error distribution and logistic link function. The probability of presence was modeled as a function of environmental predictor variables (using smooth terms), a vessel effect (factor) to account for changes in sampling gear, as well as spatiotemporal (Lat, Lon, Month) and temporal (Year) random effects, which were estimated hierarchically through a set of species-common basis functions. The model thus controls for differences in capture efficiency across survey vessels, permitting predictions on a common scale (here calibrated to the RFV Albatross IV).
16.1.2.3 Environmental Covariates
Covariate values (i.e., environmental parameters) corresponding to the approximate location (and time, when applicable) of each observation (i.e., tow) were extracted from the following sources: Monthly mean surface and bottom temperature, surface and bottom salinity, and sea surface height anomaly were obtained from the GLORYS12V1 reanalysis (Jean-Michel et al. (2021a)), as were annual minimum and maximum surface and bottom temperatures.
Monthly mean underwater optical parameters, including the intensity (photosynthetically active radiation - PAR) and spectral composition (hue angle) of downwelling light at mid-water column, were estimated from remote sensing data, following the methods of Z.-P. Lee et al. (2005) and Z. Lee et al. (2022), respectively.
Hydrodynamic stress near the seabed (95th quantile) was obtained from the USGS Sea Floor Stress and Sediment Mobility database (Dalyander et al. 2012).
Annually-integrated chlorophyll was obtained from the Oceancolour-CCI (version 5) release (https://www.oceancolour.org/).
Bathymetric position index (BPI), benthic structural complexity, and sediment type data were estimated following the methods described at: https://www.conservationgateway.org/ConservationByGeography/NorthAmerica/UnitedStates/edc/reportsdata/marine/namera/namera/Pages/default.aspx/
16.1.2.4 Estimating Richness
Simulating from the fitted model, we generated 100 random draws of “joint” predictions of the species assemblage observed in the survey, taking into account species residual covariances (see Wilkinson et al. (2021) for additional details). We used these to produce estimates of the mean species richness (and corresponding 95% prediction intervals) across all observations within each ecological production unit (EPU) for each modeled year (2000-2019).
16.1.3 Data Processing
The Habitat Diversity indicator was formatted for inclusion in the ecodata
R package with the code found here.
catalog link https://noaa-edab.github.io/catalog/habitat_diversity.html