A Delta-Lognormal Index of Abundance for Gag Grouper (Mycteroperca microlepis) Using Fish Counts and Habitat Information from a Camera Survey

The NMFS Southeast Fisheries Science Center conducts an annual camera-based survey for reef fish associated with banks and ledges on the continental shelf of the Gulf of Mexico. Indices of abundance are determined using the maximum number of a species in the field of view at a single instant. Habitat information is also obtained from video. Indices of abundance for gag grouper have been generated using a delta-lognormal approach; however previously, these have not used the habitat information. Here, we first use a general linear model in a forward procedure to determine habitat variables retained for the binomial model and lognormal submodel. We then compared the indices of abundance estimated with and without habitat variables. A total of 13 independent variables were examined. Four variables (Year, Depth, relief and proportion of seawhips in the field of view) were retained in the binomial model. The delta lognormal model of positive observations used only year and depth as prediction variables. On average, the coefficient of variation of annual indices of abundance was reduced by 5% using the habitat variables.