2005 Snook Stock Assessment Update - Final

To provide a historical context, we present the commercial snook landings from 1918 until 1957 when the Legislature declared snook a gamefish. Recreational landings expressed in pounds in recent years exceed the earlier commercial landings.

Median total-harvest estimates for snook in 2004 were similar to those from recent years (2001 - 2003) on the Atlantic coast but on the gulf coast, the landings were higher and second only to the 1997 estimates. Anglers catch and release such a large number of snook that approximately 35% of the estimated 2004 total statewide harvest was attributed to deaths that occurred after snook were caught and released alive.

In 2004, based on MRFSS interviews with anglers who indicated the type of fish they were targeting, snook were the fifth most targeted species on the Atlantic coast and fourth most targeted on the gulf coast. Several measures of effort, such as the estimated number of directed snook-fishing trips, species preference, and sales of snook stamps, indicate that the fishing effort directed at snook is high and increasing even though snook only occur in southern Florida.

Until recently we were unaware that MRFSS added variables to their dataset in 1991 that allowed identifying additional angler interviews from a specific trip. Previously, MRFSS recommended calculating catch rates using only trips with a single angler. We think that the new method of combining all the interviews from a single trip provides better insights even though the time series is shorter.

The details of calculating catch rates differed from earlier assessments because we only included terms that reduce the mean deviance in generalized linear models by more than 1% and we are using Poisson distributions for the number of fish per trip instead of assuming that catch rates are normally distributed and using the log-likelihoods to determine their significance (Type III sum of squares). The terms that significantly reduced the deviance in the MRFSS catch rates included year, two-month wave, mode, county, time fished, and avidity.

On the Atlantic coast, MRFSS total-catch rates have increased since 1997 and the median standardized total-catch rate in 2004 was the highest in the fourteen year series. The catch rates on the gulf coast have been increasing and the median total-catch rate in 2004 was also the highest value.

We also calculated catch rates from the Everglade National Park creel survey and, for the first time, we developed an index from the Park’s fishing guide survey. These two indices were applied to the gulf coast analyses. These indices showed similar increases in total-catch rates in recent years as did the MRFSS index. We were unsuccessful in developing a similar index for the Atlantic coast using Biscayne National Park’s creel survey because there were too few trips that reported catching or targeting snook.

The fishery-independent haul seine catch rates in numbers of snook per set increased in 2003 and 2004 on the Atlantic coast after decreasing over the 1999– 2002 period. On the gulf coast, the 2003 value was the highest followed by a return in 2004 to the lower values seen in the late 1990s. We used age-length keys by coast from the fishery independent samples to develop indices for ages 1 and 2. We only used those two ages because older fish had some individuals that exceeded the 650 mm (standard length) limit on culling animals for age determination.

A nagging question in assessments for gamefish has been “What are the sizes of fish that anglers release?” In the last snook stock assessment (2002), we assigned sizes to these released fish using two methods: the same sizes as the fish that anglers kept and the sizes from fishery independent sampling. However, FWC began a program in 2001 with willing anglers to measure every fish that they caught and FWC also began a fishery-independent catch-and-release fishing program in 2002 that also recorded the size of every fish caught. These two programs reported sizes for 14,286 snook. As expected, on the Atlantic coast 62% of the fish were undersized (less than 26 in) and 11% were oversized (greater than 34 in) while on the Gulf coast 80% of the fish caught were undersized and only 2% were oversized. Although these programs only had three years’ of data, we used the average lengths from the three years’ of data to assign lengths for released fish in the earlier years.

We continue to use Integrated Catch-at-Age (ICA) to estimate population sizes and fishing mortality rates. However, instead of weighting all ages and years equally as we had done in previous stock assessments, we downweighted the age-1 estimates because only a few turn up as undersized fish that were released and died subsequently and we also downweighted the numbers of fish in the last age class (plus groups) because they were composites of ages. We also weighted the catches-at-age by year based on the number of snook that were aged in a given year.

In addition to ICA, we added a program from the National Marine Fisheries Service’s Stock Assessment Toolbox called Age-Structured Assessment Program (ASAP) to estimate population sizes and fishing mortality rates. This method is more flexible than ICA in terms of numbers of indices, years for estimating selectivity, differential weighting of population parameters, and other technical details.

This is the second assessment since the slot-size limit was implemented in 1999. Fishing mortality rates on the Atlantic coast were similar to those estimated in earlier assessments but diverged after 1991 with the current values being higher.

The 2004 estimate of average fishing mortality for the reference age (Age-7) on the Atlantic coast was the same with either analytical method at 0.24 per year, approximately the average after 1997. On the gulf coast, the fishing mortality rates were much lower after the slot limit was implemented and lower than the rates estimated in earlier assessments. In 2004, the fishing mortality for the same reference age was 0.36 per year down from the 1997 high value of 0.72 per year when estimated from ICA and 0.48 per year in 2004 down from 1.01 per year in 1997 when estimated from ASAP. The spawning biomass of snook continued to decrease on the Atlantic coast after the mid-1990 highs and generally increase on the Gulf coast.

The Commission’s management objective for snook is to maintain the spawning potential ratio (SPR) at or above 40%. The transitional SPR (tSPR) values did not achieve the Commission’s objective on either coast. The transitional SPR in 2004 was estimated at 25% on the Atlantic coast by ICA and 26% by ASAP and tSPR values have been at this level since 1998. These values were lower than tSPR values estimated in earlier assessments reflecting the higher estimated fishing mortality rates. The tSPR levels on the gulf coast have been increasing since 1998 and the level in 2004 was estimated as 32% by ICA and 26% by ASAP. The static SPR values indicate that if fishing mortality rates continue at 2004 levels, snook will remain overfished on both coasts.

Anglers had high compliance with the 26–34 in slot limit that was implemented in January 1999; based upon the lengths recorded by both MRFSS and ENP samplers on average for 2002-04, approximately 88% of the fish harvested on the Atlantic coast and 90% of the fish on the Gulf coast were within the allowable slot limit. The proportion of fish larger than 34 in was greater on the Atlantic coast (8%) than on the Gulf coast (2%). During 2002–2004, one out of 1,585 trips on the Atlantic coast kept more than two fish per angler while eight out of 2,609 trips on the Gulf coast kept more than one fish per angler.

The change in measuring total length from relaxed to pinched has the potential to increase the harvest of snook in the lower end of the slot. If every angler currently measured the relaxed total length and switched to pinched total length, then the harvest could increase by 3% on the Atlantic coast and 22% on the gulf coast.