Data up to 2012. Landings of gag from the southeastern US represent approximately 20% of total landings in FAO AREA 31. NMFS landings show a decreasing trend for 2013-2016. The catch-age model included data from four fleets that caught gag in southeastern U.S. waters: commercial handlines (hook-and-line), commercial diving, recreational headboats, and general recreational. The model was fitted to data on annual landings (in gutted weight for commercial fleets, in numbers for recreational fleets); annual dead discards (in numbers) from all fleets but diving; annual length compositions of landings; annual age compositions of landings; and three fishery dependent indices of abundance (commercial handlines, headboat, and general recreational). As in SEDAR10, discard mortality rates (proportions) of 0.4 for commercial handlines and 0.25 for recreational fleets were applied to the total discards. Ages used in composition data were 1−12+, a modification from SEDAR-10, which used ages 0−20+. The reason for this modification was that very few fish were observed at age 0 or older than age 12, and too many zeros in multinomial likelihoods can be problematic. Data used in the model are described further in §2 of this report.

• A primary weakness of this assessment is the lack of fishery independent information. In general, fishery dependent indices of abundance may not track actual abundance well, because of factors such as hyperdepletion or hyperstability. Furthermore, this issue can be exacerbated by management measures, such as closures due to quota regulations. During a closure, CPUE goes to zero and hence provides no information on relative abundance. In months outside of the closure, fishing effort may become concentrated with changes in catchability, which would affect the relationship between CPUE and abundance. As such management measures become more common in the southeast U.S., the continued utility of SEDAR 10 Update 24 Assessment Update Report April 2014 South Atlantic Gag fishery dependent indices in SEDAR stock assessments will be questionable. This situation amplifies the importance of fishery independent sampling. • The assessment accounted for the protogyny of gag implicitly by measuring spawning stock as the sum of male and female mature biomass, as recommended by Brooks et al. (2008). Accounting for protogynous sex change is important for stock assessments (Alonzo et al. 2008; Shepherd et al. 2013), and the approach taken here has the advantage of being tractable. However, it ignores possible dynamics of sexual transition, which may be quite complex (e.g., density dependent, mating-system dependent, occurring at local spatial scales). In addition, a protogynous life history accompanied by size- or age-selective harvest places disproportionate fishing pressure on males. This situation creates the possibility for population growth to become limited by the proportion of males. When this occurs, accounting for male (sperm) limitation may be important to the stock assessment (Alonzo and Mangel 2004; Brooks et al. 2008); however, in practice there is typically little or no information available to quantify sperm limitation. In this assessment, the proportion of adult fish that are male drops below 10% in most years, and is below 10% in recent years since the mid-1980s (Table 10). The equilibrium proportion of adult fish that are male at MSY is near 20% (in numbers), but again, this estimate does not explicitly account for the dynamics of sperm limitation. • Most assessed stocks in the southeast U.S. have shown histories of heavy exploitation. High rates of fishing mortality can lead to adaptive responses in life-history characteristics, such as growth and maturity schedules. Such adaptations can affect expected yield and stock recovery, and thus resource managers might wish to consider possible evolutionary effects of fishing in their management plans (Dunlop et al. 2009; Enberg et al. 2009).