Change-in-ratio methods for estimating recreational exploitation rate and absolute abundance of Gulf of Mexico red snapper

End Date: 12/05/16


  1. To identify areas of high recreational removals within Gulf of Mexico federal waters as the basis for defining study areas.
  2. To identify potential collaborators working in these areas and evaluate if the Change-in-Ratio (CIR) method is appropriate and feasible for these regions.
  3. To evaluate the index-removal (IR) method, as well as tagging studies, as potential additions to the CIR method by using Monte Carlo simulation to assess precision and by consulting with other researchers in the Gulf of Mexico to assess logistical feasibility. 
  4. To determine which estimates of recreational removals in each region are most accurate or least biased (MRIP vs State surveys) and what modifications, if any, might be needed to obtain information such as areas fished.
  5. To conduct pilot studies comparing catch rates on artificial reef  to catch rates in different habitats (natural reefs, low relief areas) to determine if some areas can be neglected from consideration. 


Following the identification of potential collaborators located in areas of high recreational removals, we will evaluate the efficacy of the CIR method, which has shown potential applicability for red snapper on artificial reefs within the AARZ. Using this method, an unbiased estimate of exploitation can be obtained via (formula will not translate here)  where p1  and  p2 represent the proportion of legal-size fish in the pre- and post-season survey, respectively (Seber 1982, Pollock and Hoenig 1998). Preliminary results have shown that the surveys may be several months apart as long as growth, natural mortality, and gear selectivity are all incorporated into the model. Abundance can be estimated using the CIR method by dividing the removals in the area by the estimated exploitation rate. If exact removals for the area are not known, then relative abundance can be calculated using the fraction of removals that occur in the area. In addition to the CIR method, we will evaluate the use of the index-removal (IR) method, as well as if IR models that incorporate changing catchability (Ihde et al. 2008) are appropriate for Gulf of Mexico red snapper. If so, IR can be combined with CIR in a likelihood framework to obtain more precise abundance estimates (Chen et al. 1998).

Spatially-explicit removals must be quantified so the total removals in the area are known. We will evaluate sources of removals data (e.g. state data, MRIP) to ensure that the best available information is used for these analyses. If the spatial scale of the landings data is not fine enough to determine the removals for the area, or if we believe the data require validation, we will consider implementing interview questions at port intercepts during the federal recreational fishing season to determine the fraction of harvest occurring in the area of interest.

Based on the work we have done in the AARZ, we are confident that this method can be scaled up to encompass all major hotspots in the Gulf of Mexico. In order to identify red snapper hotspots over specific habitat types in regions that we are unfamiliar with, we will rely on interviews with anglers, other researchers and other stakeholders as well as the recent work by Karnauskas et al. (in review), which has succeeded in spatially mapping the relative abundance of red snapper across different habitats and structures in the Gulf of Mexico. To quantify if differences in catch rates exist between the different habitats identified, we will conduct pilot studies using the vertical longline methods in the pre and post survey our preliminary findings are based on. 


Gulf of Mexico red snapper are a valuable natural resource, and as such there is a clear need to independently and directly evaluate the abundance of this stock. The objective described in this request for proposals is “to assess the population of red snapper on artificial reefs and other structures, and to provide a Gulf-wide estimate (with estimates also produced for natural habitats) of absolute abundance of fish Age-2 and older in the U.S. Gulf of Mexico (by age or age-groups).” The primary approach we propose for estimating population size is the change-in-ratio (CIR) method. 

 The CIR method can be used to estimate exploitation and abundance of legal- and sublegal-sized fish in areas where removals are high. CIR is based on the idea that selective harvest of legal-size individuals shifts the population ratio of legal to sub-legal fish. This method requires fishery-independent surveys before and after the fishing season. Estimates of exploitation rate can be calculated without knowing removals. An estimate of absolute abundance requires knowing the absolute removals in an area; however, an estimate of relative abundance simply requires knowing the fraction of removals (taken in the entire US Gulf of Mexico) that occurs in an area. 

 We have done preliminary work in the Alabama Artificial Reef Zone (AARZ) using the University of South Alabama (USA) paired VL and remotely-operated vehicle (ROV) surveys before and after the recreational fishing season from 2011 – 2014, and have successfully applied the CIR method to the red snapper population. Logistically, estimating abundance on natural reefs is the same as for artificial reefs. We therefore propose that this method is feasible and effective for other areas of dense removals within the Gulf of Mexico. This method could also be used on low relief bottom provided the density of fish there is sufficiently high to attract substantial fishing pressure to change the ratio of legal to sublegal fish. If abundance is too low for the method to work, then neglecting to study low fish density areas will cause abundance to be somewhat underestimated. Preliminary vertical longline (VL) surveys will be used to assess the relative abundance of red snapper on artificial reef, natural reef and low relief areas.

Not only will this approach provide an alternate method of abundance estimation, but it also sets the stage for area-based management. The CIR method allows for direct comparison of exploitation and abundance between areas of high removal in the Gulf of Mexico. Given the significant spatial heterogeneity in the distribution of red snapper within the Gulf of Mexico, the use of CIR or other removal-based methods provides for evaluation of hotspots across the entire Gulf of Mexico. This method establishes a baseline for further research possibilities involving spatial management and/or assessing if red snapper hotspots have large catches due to heavy exploitation versus high abundance.

 Another estimator that could be used in addition to CIR is the index-removal (IR) method. Index-removal uses fishery-independent surveys to monitor the change in the catch rate of legal-size fish between two surveys. Exploitation and abundance can be calculated using removals data similarly to CIR. Alternatively, if the total number of artificial structures are known in a particular hotspot, an experimental IR could be set up similar to what’s been done in the AARZ (see Rationale), where abundance is estimated for several artificial reef sites, and the mean abundance is expanded out to the entire area. 

We recognize the value of including recreational fishers, both for their knowledge of locations where red snapper are abundant and for their cooperation during the recreational fishing season. During Phase I, we will incorporate recreational fishers’ (as well as state and scientific stakeholders’) knowledge of red snapper locations in order to verify and improve our sampling frame, and will invoke their feedback to finalize study design elements. During Phase II, we will use data from recreational fishers to verify fishing locations obtained from creel sampling programs to obtain accurate spatially-explicit estimates of red snapper removals.

We will work with our NMFS external partners to ensure that the project results are expressed in a way that is similar to the NMFS assessment. This will allow for meaningful comparison between our abundance estimates and what the assessment predicts. Our estimates will be on a finer spatial scale than the two-area model that the assessment currently uses, and we anticipate that incorporating our results from red snapper hotspots would benefit the assessment and management process.