Table 1.
Data sources of marine fish predictive modeling articles attained from the Web of Science database (2007–2018) by fish guild.
Fig 1.
Temporal trend in marine fish predictive modeling studies.
Number of published articles from 2007–2018 that predicted the spatial distribution of marine fish based on a search conducted in the Web of Science database.
Fig 2.
Word cloud representing research themes of marine fish predictive modeling studies.
Word size is proportional to its frequency in article abstracts derived from a search of articles from 2007–2018 (n = 225).
Table 2.
Methodology used to predict the distribution of marine fish in studies spanning 2007–2018 (n = 225).
Fig 3.
Frequency of predictor variable categories tested by fish guilds.
For each fish guild (x-axis), predictor variables were summarized into categories of biological (Biol), geographic (Geogr), physical oceanographic (Physical ocean.), physiology-based oceanographic (Physiology), or substrate (n = 224).
Fig 4.
Frequency of predictor variables in marine fish spatial modeling articles.
(Left panel) Frequency of predictors among all marine fish studies; variables in ≥ 3% of studies are shown (n = 224). (Right panel) The most frequent predictors within each fish guild; only predictors in > 15% of studies, or a maximum of 10 variables, are shown (n = 224). * = predictor that distinguished the fish guild from others, as observed in Table 3; BPI = bathymetric position index, dist = distance, DO = dissolved oxygen, HB = hardbottom, Lat/Long = latitude/longitude coordinates, SD = standard deviation, SSH = sea surface height anomaly, SST = sea surface temperature, Temp = temperature, Bott = bottom, Suf = surface.
Fig 5.
Linear discriminant analysis showing differences in predictor variables tested among fish guilds.
The three discriminates (LD1, LD2, LD3) are multivariate combinations of predictor variables tested in 224 marine fish spatial modeling studies. Each fish guild is distinguished by color and shape as depicted in the legend.
Table 3.
Distinguishing predictor variables for each marine fish guild based on a linear discriminant analysis.