Characteristics of Environmental Data Layers for Use in Species Distribution Modelling in the Newfoundland and Labrador Region

Species distribution models require spatially linked response (e.g., species, habitat type) and environmental (predictor) point data. Often, only limited environmental data types can be collected at the time of sampling, and it may be desirable to capture information both from different data sources...

Full description

Bibliographic Details
Main Author: Lirette, Camille
Format: Dataset
Language:unknown
Published: Mendeley 2019
Subjects:
Online Access:https://dx.doi.org/10.17632/p32f6y5ph5.1
https://data.mendeley.com/datasets/p32f6y5ph5/1
Description
Summary:Species distribution models require spatially linked response (e.g., species, habitat type) and environmental (predictor) point data. Often, only limited environmental data types can be collected at the time of sampling, and it may be desirable to capture information both from different data sources and/or longer term data series in order to predict the distribution of a response variable. In order to link response and predictor variables that are not sampled at the same location or time, geospatial interpolation techniques are applied. Here, we provide a review of 104 environmental variables from each of 8 water column properties: Temperature, Salinity, Current Speed, Maximum Seasonal Mixed Layer Depth, Bottom Shear, Sea Surface Chlorophyll a, Primary Production and Dissolved Inorganic Nutrients for the ‘Newfoundland and Labrador Region’ (a combined spatial extent of DFO’s Placentia Bay-Grand Bank and Newfoundland and Labrador Shelves Large Ocean Management Areas). All of these variables have potential biological relevance to benthic invertebrate species. Original data sources were the Global Ocean Reanalyses and Simulations (GLORYS), the Sea-viewing Wide Field-of-view Sensor (SeaWIFS), and the World Ocean Database 2013 (WOD13). For each variable, the original data characteristics and diagnostics produced from spatial interpolation using ordinary kriging are detailed. Standard error and prediction maps are shown for each variable. Based on these diagnostics, a subset of these variables was subsequently used in species distribution models of corals, sponges, crinoids, ascidians and bryozoans in the Newfoundland and Labrador Region.