2003 Yellowtail Snapper Stock Assessment

This article provides an assessment of the status of yellowtail snapper in the southeast United States through 2003.

Download the Stock Assessment (PDF 1.8 MB)

To view this PDF file, you will need Adobe Reader.
To download Adobe Reader, visit http://get.adobe.com/reader/

A stock assessment of yellowtail snapper,
Ocyurus chrysurus, in the Southeast United States

Robert G. Muller, Michael D. Murphy, Janaka de Silva, and Luiz R. Barbieri
Florida Fish and Wildlife Conservation Commission
Florida Marine Research Institute
St. Petersburg, FL

(revised for Web site)

The status of yellowtail snapper was assessed through the National Marine Fisheries Service's SEDAR process with the Florida Fish and Wildlife Conservation Commission (FWC) taking the lead. The SEDAR process consists of three workshops. The data workshop was held 3-4 March 2003 at FWC's Florida Marine Research Institute in St.Petersburg, and the Stock Assessment Workshop was held 9-13 June 2003 at the same venue. The Peer-Review Workshop was held 28-31 July 2003 in Tampa, Florida.

The following is a summary of the biology, fishery, and assessment of yellowtail snapper with comments about important discussions and conclusions made by the Stock Assessment Workshop Panel and the Peer-Review Panel.

Yellowtail snapper, Ocyurus chrysurus, is a reef fish species that occurs from North Carolina to southern Brazil and is abundant in south Florida. Adult yellowtail snapper typically inhabit sandy areas near offshore reefs at depths of 10 m-70 m (32-230 feet). Yellowtail snapper eat fish, shrimp, and crabs near the bottom but also feed in the water column.

The spawning season in south Florida is during spring and summer with a peak during May-July. Females reach the 50% maturity ogive at 209 mm TL at an average age of 1.7 years. Yellowtail snapper grow quickly initially, but size is a poor indicator of age because of the extensive overlap in ages for a given sized fish. The Data Workshop Panel recommended not pursuing sex-specific differences in growth because, based on the analysis, the available data did not show any obvious differences in size at age between sexes. There were detailed discussions about the potential difficulties in using agestructured assessment approaches when so much variability in length at age was observed. The Stock Assessment Panel finally agreed on the assessment approach after looking at catch-at-age data generated using direct aging of the catch and various pooling strategies for the development of age-length keys. Based on the maximum age of sampled yellowtail snapper (17 years old, confirmed since the Data Workshop) and the established nature of the fishery, the panel recommended using a lower natural mortality rate than suggested at the Data Workshop. A baseline instantaneous rate of 0.2 yr-1 was used with additional runs also at 0.15 yr-1 and 0.25 yr-1. After discussion, the Peer-Review Panel found no reason to change either the baseline rate or the range.

The commercial fishery for yellowtail snapper occurs throughout the tropical, western Atlantic. Average landings from the Caribbean for 1997-2000 have been 3,458 metric tons (mt), and of that total, the United States landings have averaged 747 mt with Puerto Rico and the U.S. Virgin Islands accounting for another 220 mt. The fishery has occurred in the Florida Keys for over a century and mostly uses hook-and-line gear, especially after entangling gear was prohibited in 1990, five years before Florida's Constitutional amendment banning the use of entangling gear from state waters.

For this stock assessment, data from the yellowtail snapper fisheries were divided into two regions: the Atlantic region, which primarily is from Palm Beach County south through Miami-Dade County, and the Keys, which is Monroe County and all counties to the west. The Stock Assessment and Peer-Review panels concurred with the approach to estimate catch-at-age for the MRFSS recreational, headboat, and commercial sectors separately for the Atlantic (Miami-Dade County north) and Keys regions (Monroe County north). Although there are commercial landings data from earlier years, recreational data are only available since 1981; therefore, we have confined the analyses to the years 1981-2001.

Total landings during these years increased from 1,000 mt in 1981 to 1,648 mt in 1993 and then decreased to 802 mt in 2001. Despite Data Workshop panel recommendations that sensitivity analysis include temporal increases in unreported commercial catch, the Stock Assessment Panel did not recommend including this as a sensitivity analysis due to the lack of empirical evidence for changes in reporting. Effort followed a trend similar to that of total landings, increasing to a peak and then decreasing. The number of commercial fishers has decreased from a peak of 8,343 Saltwater Products license (SPL)holders in 1989 to 2,659 SPLs in 2001. Recreational trips declined from 2.3 million trips in 1988 in the Atlantic regions to 1.7 million trips in 2001 and from 1.2 million trips in 1993 in the Keys to 0.4 million trips in 2001. Similarly, the headboat effort was highest in 1981 with 155,000 angler-days in the Atlantic region and generally declined to 63,000 angler-days in 2001 and from 82,000 angler-days in 1989 in the Keys to 45,000 anglerdays in 2001.

The Stock Assessment Panel discussed the estimation of commercial discard rate and discard mortality and agreed to use the preliminary discard data from commercial logbooks instead of the 10% discard mortality rate suggested during the Data Workshop. The Peer-Review Panel noted that the paucity of discard data was unsatisfactory and fishers on the panel indicated that these rates were too high. In the assessment runs, we increased the landings to account for discards. Based on a single year's reef fish logbook data in 2001-2002, commercial discards of yellowtail snapper averaged 16% of the landings, and approximately 28% of those discarded were dead. Recreational discards are estimated directly as Type B2 numbers of fish. With the absence of headboat discard information, the Stock Assessment Panel concluded, after much discussion and examination of the age distribution of the fishery-dependent and fishery-independent samples by region for 1999-2001, that the proportion of fish that would have been discarded by headboat could be assumed to equal the fraction of the catch of the fishery-independent hook-and-line data that was smaller than the legal size limit (305 mm TL)(37% in the Atlantic region and 27% in the Keys region). The panel discussed the 30% discard mortality rate used with the recreational and headboat fisheries and found insufficient evidence to suggest changes to these rates.

Commercial landings in weight were converted to landings in number based on biostatistical sampling of the landings that measured lengths from landings with different gear. Biostatistical samplers visit fish houses, interview fishers, measure fish, and collect hard parts for age determinations. Landings are estimated directly in numbers in the recreational fisheries. Ages were assigned to the lengths based on region, fishery, gear, and year.

The Stock Assessment Panel noted differences in age composition between the Atlantic and Keys for all fishery-dependent and fishery-independent data. This was most evident for maximum age: 7 years in the Atlantic and 17 years in the Keys. The age-length data were sufficient from 1997 onward to derive year-specific age-length keys for both regions. The panel noted that the composite age-length keys could act to obscure yearclass strength information. For earlier years when sample sizes were insufficient, the Stock Assessment Panel recommended combining data from the same region and year but from different gear. When data from alternative gear were not available, the second choice for substitution was to use data from the same region and from different gear in different years. A composite was formed for the years 1980-1986 and 1987-1996. Finally, the panel investigated using age data from 1994 through 2001 to directly age the catch. However, the abrupt change in the younger ages from the composite age-length keys to direct aging led the panel to recommend not using the direct aging method.

We used tuning indices to improve the statistical population models. The two fisheryindependent indices were based on visual surveys conducted by the National Marine Fisheries Service and the University of Miami. These indices were the number of fish less than 197 mm (7¾ inches) per 177 m² that was used for age-1 fish and the number of fish greater than 197 mm per 177 m² that was used for fish age-2 and older. The panel discussed the change in the number of strata used to develop these indices. However, a subsequent conversation with the analyst that developed the indices confirmed that these indices were the most comparable and that the increase in the number of strata was to account for protected areas in the Keys and the partitioning of patch reefs to afford finer resolution. The panel rejected the use of the third fishery-independent index, REEF visual survey. The coarseness of the classification of abundance, i.e. 0, 1-10, 11-100, >100 individuals, was considered to be too great to use the REEF index as a quantitative index for yellowtail snapper abundance.

In addition to the fishery-independent indices, we originally developed five fishery-dependent indices that were standardized with generalized linear models:

  • Commercial kilograms per trip with combined gear (1985-2001)
  • Commercial kilograms per hook-and-line trip from trip tickets (1992-2001)
  • Commercial kilograms per hook-and-line trip from Reef Fish Permit logbooks (1993-2001)
  • MRFSS recreational total number of fish caught per trip (1981-2001)
  • Headboat number of fish landed per trip that was divided into two time periods, 1981-1991 and 1992-2001 because of the aggregate bag limit

After much discussion, the panel agreed that the original CPUE indices for headboat and commercial sectors derived by the analysts before the assessment workshop were valid indices. However, these indices were derived under the philosophy of including many reef trips, only coarsely filtered for yellowtail snapper trips. The panel also felt that another set of valid CPUE indices should be derived based on anglers that were targeting yellowtail snapper. We developed two additional indices: the kilograms per trip, from commercial hook-and-line trips by 107 Reef Fish Permit holders that landed at least 500 kilograms of yellowtail snapper in five of the most recent seven years, and headboat indices, from seven vessels that landed at least 100 yellowtail snapper per year. The Peer-Review Panel pointed out that including interaction terms with year in the indices might not reflect underlying population changes and recommended calculating the indices with just main effects. They also requested an analysis without the commercial index because they thought that perhaps the increase in that index was due to increased efficiency instead of a population increase. The run without the commercial index produced the same trends as before.

Finally, the Stock Assessment Panel noted that the flat CPUE indices with declining landings implied declining effort. Subsequent analyses requested by the panel confirmed declining annual number of angler-days for the headboat sector and declining overall number of trips in the MRFSS recreational and the commercial sectors.

We used two types of models to assess the condition of yellowtail snapper: surplus production and age-structured, statistical models. However, the two surplus production models, ASPIC, a non-equilibrium model, and ASP, an age-structured model, were not stable, and most likely, the instability was due to lack of contrast in the tuning indices or catch rates. The Stock Assessment Panel noted that the generally flat or monotonic CPUE indices could create parameter-estimation-convergence issues with surplus production models.

Both the Stock Assessment and Peer-Review panels agreed with the Data Workshop recommendation that age-structured assessment approaches were appropriate for yellowtail snapper. Year-specific aging information was available for 1994-2001. Age-structured approaches could make use of all available data, lending more confidence in the analyses' predictions of the current status of the stock.

We used two age-structured, statistical models. The first was Integrated Catch-at-Age which used the combined catch-at-age from the three fisheries and tuning indices to estimate the population sizes by age in the most recent year, fishing mortality rates on the earliest fully recruited age of fish, selectivity patterns by age, and catchability coefficients for the tuning indices (76 parameters in this configuration). In the base case run, the full fishing mortality rate in 2001 was 0.21 per year and the spawning biomass in 2001 was 4,943 mt.

The numbers of age-1 fish and the spawning biomass a year earlier were used to estimate the biomass-based management benchmarks given a steepness of 0.8 and alternatives of 0.7 and 0.9. The steepness is merely the proportion of the recruitment at a spawning biomass of 20% of the virgin biomass to the recruitment at the virgin biomass. With the Stock Assessment Panel recommendation of using a steepness value of 0.8, the maximum sustainable yield (MSY) was 941 mt; the F2001/FMSY ratio was 0.62, and the SSB2001/SSBMSY ratio was 1.35, indicating that the stock was not undergoing overfishing and not overfished. The ratios were 0.57 and 1.43 when the analyses were rerun using indices calculated without the interaction terms.

The second age-structured model allows simultaneous estimation of separate fishing mortality rates for the three fisheries. This fishery-specific model estimated the population sizes in the first year (1981), recruitment from a stock-recruit relationship,selectivities by fishery and by two periods corresponding to before and after the 12-inch (305 mm) size limit was implemented in 1983, and catchability coefficients for the tuning indices. This model estimated the sum of the fishing mortality rates on fully recruited fish in 2001 at 0.24 per year and a spawning biomass of 5,200 mt, which is similar to the 0.21 per year and 4,900 mt estimated by ICA. The fishery-specific model estimated a higher MSY of 1,366 mt but only a slightly higher FMSY (0.36 per year as compared to 0.33 per year from ICA). The biomass-based benchmarks were F2001/FMSY = 0.65 and SSB2001/SSBMSY = 1.06. Using the revised indices, the fishing mortality rates on fully recruited fish in 2001 remained 0.24 yr-1, and the estimated spawning biomass increased slightly to 5,300 mt. The revised biomass-based benchmarks were F2001/FMSY = 0.72 and SSB2001/SSBMSY = 0.99, supporting the same conclusion that the stock was neither undergoing overfishing nor overfished.

The retrospective analyses using terminal years of 1998, 1999, 2000, and 2001 did not indicate that the models consistently overestimated or underestimated either the fishing mortality rates in the last year or the spawning biomass.

Landings of yellowtail snapper differ widely by subregion. Yellowtail snapper were rarely landed north of Florida's Palm Beach County on the Atlantic coast. From Palm Beach County south through Miami-Dade County, yellowtail snapper were consistently landed; however, in all three fishing sectors the majority of landings came from the Florida Keys. The fishers from counties north of the Keys on the gulf side also rarely landed yellowtail snapper. This assessment focused on Southeast Florida (Palm Beach through Miami-Dade counties) and the Florida Keys because of the concentration of landings in those two subregions. The geographical distribution of yellowtail snapper landings reflects the distribution of coral reefs in Florida. Effort in terms of fishing trips was proportionately higher in Southeast Florida than in the Keys but there were still more trips in the Keys. Also, in all three sectors, the catch rates were higher in the Florida Keys than in Southeast Florida.

There was high compliance with the 12-inch minimum size (305 mm); only 3% of the commercial landings, 5% of the recreational landings, and 2% of the headboat landings in the Atlantic region were under the limit. In the Keys, the compliance was also high; only 2% of the commercial landings, 4% of the recreational landings, and 3% of the headboat landings were under the limit. While we evaluated the 10-fish aggregate limit by assuming that all of the snappers were yellowtail snapper, most of the recreational anglers caught less than two fish per trip. Only 0.2% of the anglers in the Atlantic region and 1.3% of the anglers in the Keys exceeded 10 fish per trip.

For other information:
Stock assessments for finfish and invertebrate

FWC Facts:
There are 10 to 100 times more animals in seagrass beds than in adjacent sandy areas.

Learn More at AskFWC