Fig 1.
Locations of video samples included in the analyses from the Southeast Reef Fish Survey.
Data is from the southeast United States, 2011–2014 (gray points; N = 4,855 videos in total). Note that points often overlap. Gray isobaths indicate 30-, 50-, and 100-m depths, and arrows indicate the general path of the Gulf Stream.
Table 1.
Sampling information for the 32 most common species seen on video in the Southeast Reef Fish Survey.
Table 2.
Binomial generalized additive model (GAM) results for the significance of position, depth, and substrate in describing the presence or absence of 32 species seen on at least 10 videos collected by the Southeast Reef Fish Survey, 2011–2014.
Fig 2.
Presence and absence of the 16 most common reef fish species seen on videos.
Red points indicate presence and gray points indicate absence from the Southeast Reef Fish Survey, 2011–2014; note that points often overlap. Each background shows the partial effects of position on that species pattern of presence or absence from the generalized additive model; orange is the highest predicted probability of presence and blue is the lowest. The “N” shows the number of videos samples in which the species was present.
Fig 3.
Presence and absence of the next 16 most common reef fish species seen on videos.
Red points indicate presence and gray points indicate absence from the Southeast Reef Fish Survey, 2011–2014; note that points often overlap. Each background shows the partial effects of position on that species pattern of presence or absence from the generalized additive model; orange is the highest predicted probability of presence and blue is the lowest, and background heat maps were removed if the position variable was insignificant. The “N” shows the number of videos samples in which the species was present.
Fig 4.
Depth distribution of reef fish species seen on at least 10 videos.
Bars show the proportion of videos within each depth zone (m) that the species was seen. The thick black line shows the mean generalized additive model fit for each species, and is absent if it was insignificant. Horizontal dashed lines show the overall proportion of videos in which the species was present. Note different y-axes among panels.
Fig 5.
Distribution of reef fish species across substrates that were seen on at least 10 videos.
Bars show the proportion of videos within each substrate type that the species was seen. The thick black line shows the mean generalized additive model fit for each species, and is absent if it was insignificant. Horizontal dashed lines show the overall proportion of videos in which the species was present. Note different y-axes among panels.
Fig 6.
Locations of selected uncommon reef fish species seen on videos.
Note that some points overlap. Gray isobaths indicate 30-, 50-, and 100-m depths.
Fig 7.
Relationships between the predicted number of priority fish species seen on video and predictor variables.
The generalized additive model was built using Southeast Reef Fish Survey video data, 2011–2014. Points (for categorical variables) or solid lines (for smoothed continuous variables) show the mean predicted number of species seen and the dashed lines show 95% confidence intervals.
Fig 8.
Observed and predicted number of priority fish species seen on video.
Spatial predictions for the number of priority fish species on video were made at average values of all other predictor variables in the generalized additive model, and are shown from orange (indicating many species predicted) to blue (indicating low number of species predicted). Open circles show the mean observed number of species on video in 2011–2014 for each 0.25 × 0.25 degree cell.