8 Mediterranean Pelagic Commercial Fish Species for Distribution Modelling

Published: 18 April 2020| Version 1 | DOI: 10.17632/rtphmvz5fj.1
Dimitrios Effrosynidis


The objective is to create a species distribution model to predict the species probability in each longitude, latitude pair. The fish species Engraulis encrasicolus (729 samples), Sardina pilchardus (596 samples), Sardinella aurita (720 samples), Scomber colias (676 samples), Scomber scombrus (807 samples), Spicara smaris (1225 samples), Thunnus thynnus (1237 samples) and Xiphias gladius (1230 samples) are included. The target for prediction is the overall probability of occurrence. There are 6830 predictors created from 15 environmental variables: temperature, salinity, dissolved oxygen, meridional current, zonal current, chlorophyll, euphotic depth, secchi disk depth, wave height, nitrate, phosphate, distance to coast, distance to major river, bathymetry, and substrate.



Democritus University of Thrace


Fish, Environmental Analysis, Data Science, Machine Learning, Feature Selection, Feature Extraction, Aquatic Species, Distribution System Modeling, Mediterranean Sea