William G. Lyons, Florida Marine Research Institute (FDEP), St. Petersburg, FL.

OBJECTIVES: Knowledge of the relationships between environmental change, habitat change, and the recruitment, growth, and survivorship of constituent fauna is requisite for restoration efforts to be effective in Florida Bay. The work reported here is intended to describe the composition and map the distribution of benthic faunal assemblages in the bay, using mollusks to represent those assemblages, and to assess the impact of salinity fluctuations on those assemblages. The information acquired will enable managers to more accurately predict the consequences of proposed restoration activities that involve alterations of Florida Bay salinity.

Mollusks are important components of the bay fauna, influencing the composition and structure of its sediments and serving as food for other life forms (Lyons, 1995). Even the concept of subenvironments, now adopted by researchers in several disciplines, was promulgated from evidence of dead mollusk shells, whose distributions indicated distinctive assemblages in several regions of the bay (Turney and Perkins, 1972). However, Florida Bay is a dynamic estuary where environmental conditions, especially salinity, fluctuate widely, so faunal composition should shift, at least temporally, in response to environmental perturbations. Evidence of relatively recent faunal shifts has been found in shallow cores from Florida Bay (Wingard et al., 1995a, b), and similar shifts seem evident among data from some of Turney and Perkins' sites, where shells of brackish, marine and even semi-terrestrial species were found mixed in single samples. By restricting observations to living animals, we can produce relatively instantaneous "snapshots" of assemblages, thereby allowing more precise correlations between the assemblages and their ambient environmental parameters.

PROGRESS: Using methods described by Lyons (1995), we sampled mollusks quantitatively at 101 sites distributed throughout the bay in summer 1994. Most of the bay exhibited varying amounts of hypersalinity; bottom salinity ranged upward to 52 %, exceeded 36 % at 75 of the 101 sites and exceeded 30.0 % at 97 of the sites. Discolored waters due to phytoplankton blooms occurred in zones of higher salinity in the central bay. Elsewhere, turbid wind-driven waters bearing silt and organic detritus from breakdown of the seagrass beds swept westward onto and across banks and deeper areas near the western bay boundary.

Ninety-four species and nearly 14,000 specimens of mollusks occurred in the 1994 samples. Most of the species showed discrete distributions. Intersite comparisons of species composition and abundance among the 60+ more common species indicated four groupings of contiguous sites: an east-central and a western group, each with three subgroups; a peripheral Gulf boundary group; and a northern interior brackish-water group. Species richness was generally lower in zones of highest and lowest salinities. Because most species have been found to live in discrete assemblages, the more wide-spread distributions of their dead shells probably result from environmental fluctuations. Live-collected animals of some species have been found distributed in a pattern suggesting that they avoid or cannot tolerate very high salinity; the more widespread occurrence of their dead shells probably indicates more moderate conditions in earlier times. These living assemblages can also be useful in interpreting conditions indicated by fossil evidence. A relatively young fossil molluscan assemblage found at the Bob Allen Keys (Wingard et al., 1995a) was absent when we sampled there during the 1994 high-salinity event, but the assemblage was found at several sites a few kilometers away. Data from those sites can provide insight into prior environmental conditions at the Bob Allen Keys.

The hypersaline conditions of summer 1994 were followed quickly by rains that reduced salinity throughout the bay. The highest salinity measured among 105 sites in August 1995 was 35.5 %; salinity ranged from 28.0-35.5 % at 70 sites, from 20.0-28.0 % at 21 sites, and from 0-20 % at 14 sites. Hypersalinity had disappeared from the bay, and hyposaline conditions prevailed at many sites. The plume of silt and organic detritus had also abated by summer 1995, and sediments began to stabilize at the western banks. Thirty sites selected by stratification among the 1994 groupings were resampled during August 1995, and another site peripheral to the bay was added to investigate faunal transitions. The 31 sites yielded 2,853 specimens in 86 species, including 17 "new" species not recorded in 1994 samples; all of the "new" taxa were uncommon or rare. To examine seasonal influences and to further monitor species richness and abundance, 15 of the 31 sites were resampled in November- December (fall-winter) 1995, yielding 542 specimens in 64 species, including 5 species not recorded in prior samples. Sixteen of the sites were again resampled in early April (spring) 1996, yielding 795 specimens in 78 species, including 8 "new" species. Finally, the 102 sites were all resampled during July-August 1996.

Results that follow are from the 30 sites sampled in the summers of 1994, 1995, and 1996, and from the subset of those sites that was sampled up to five times during slightly more than two years. The cumulative list of species represented by living animals has increased from 94 to 124 since 1994, but all the additional species are uncommon or rare and contribute little to the characteristic fauna, whose composition has remained stable. The most dramatic changes during the period include a general decrease in molluscan abundance and marked faunal changes along the northern rim of the bay where salinity changes have been greatest. Results from fall-winter 1995 and spring 1996 indicate moderate declines in species richness at most eastern bay sites, where salinities have remained around the low 20's. In the western bay, species richness and abundance have increased at the shallow sites where siltation has diminished. Species richness has also increased at southern sites nearest the Keys.

The general decrease in molluscan abundance from 1994 to 1996 is attributed principally to a single species. Abundance in 1994 was much influenced by a mussel, Brachidontes exustus, that occurred at half of all sites and contributed 78% of all specimens. The species occurred throughout the eastern and central bay, usually in low numbers, but it was abundant at a few sites along an arc from Madeira Bay to Pontoon Bank. Mussels were most abundant near Triplet Keys, but its numbers diminished there as they did nearly everywhere in 1995 and 1996. However, a recruitment event near Pass Key during spring-summer 1996 suggests that the population may again be increasing.

Where salinities were highest in the north-central bay, salinities are now moderate but the molluscan fauna has suffered what may be a temporary collapse. Mollusks were rare in Rankin Lake during summers of 1994 and 1995, but a moderately abundant and distinctive fauna was found at both Garfield Bight and Rankin Bight in 1994. That fauna disappeared from Rankin Bight by summer 1995, was gone from Garfield Bight by fall-winter 1995, and did not reappear there in spring 1996. However, Tellina tampaensis and Anomalocardia auberiana, the bivalves that characterized the assemblage at those sites in 1994, reappeared in August 1996. Likewise, Macoma mitchelli, a bivalve that favors mesohaline environments, constituted 94% of all specimens at Joe Bay in 1994 but virtually disappeared in summer and fall 1995, when salinities fell to near zero. However, its abundance began to increase again in spring and summer of 1996, when salinity climbed back into the mesohaline range.

SUMMARY: We have described and mapped distributions of molluscan assemblages throughout the bay during the prior hypersalinity event. Next, we monitored species composition and abundance at 15-30% of the sites for two years to elucidate seasonal faunal shifts and to assess regional trends during the period of transition as bay waters shifted from hypersalinity toward hyposalinity and then back toward a more moderate regime. Finally, we have resampled all of the original sites during summer 1996. After those samples are processed, we will identify and map the constituent assemblages as they occurred under conditions of moderate salinity, and we will describe changes that have occurred during the bay's recovery from the hypersalinity event.

REFERENCES:

Lyons, W. G. 1995. Mapping Florida Bay benthic assemblages: using mollusks to assess faunal change. Pp. 167-169 in: Florida Bay Science Conference: a report by principal investigators, October 17 & 18, 1995. Gainesville, Florida.

Turney, W. J., and B. F. Perkins. 1972. Molluscan distribution in Florida Bay. Sedimenta III. University of Miami, Fisher Island Station, Miami Beach, Florida. 37 pp.

Wingard, G. L., S. E. Ishman, T. M. Cronin, L. E. Edwards, D. A. Willard, and R. B. Halley. 1995a. Preliminary analysis of down-core biotic assemblages: Bob Allen Keys, Everglades National Park, Florida Bay. U. S. Geological Survey Open-File Report 95-628. 35 pp.

Wingard, G. L., T. M. Cronin, D. A. Willard, S. E. Ishman, L. E. Edwards, C. Holmes, and S. D. Weedman. 1995b. Florida Bay ecosystem: measuring historical change. Pp. 118-120 in: Florida Bay Science Conference: a report by principal investigators, October 17 & 18, 1995. Gainesville, Florida.

Joan A. Browder, NOAA, Southeast Fisheries Science Center, Miami, FL; Victor Restrepo and Nelson Ehrhardt, University of Miami, Rosenstiel School of Marine and Atmospheric Sciences, Miami, FL; Michael Robblee, U.S. Geological Survey, Biological Resources Division, Miami, FL; and James Nance, Peter Sheridan and Zoula Zein-Eldin, NOAA, Southeast Fisheries Science Center, Galveston, TX.

The harvest of pink shrimp on spawning grounds in the Dry Tortugas is thought to be dependent upon inshore nursery grounds, including Florida Bay. Landings and catch rates declined substantially in the 1980s, starting a few years prior to observations of habitat decline in the Bay. The pattern of monthly size frequency distributions in landings suggests that, although recruitment to the fishing grounds is continuous throughout the year, until about 1980, two or more major waves of recruitment were detectable, one centered in late fall and the other in early spring. After 1980, only the spring wave was noticeable. Previous studies have suggested that landings and catch rates are positively correlated with indicators of freshwater inputs. Previous studies have suggested that spawning by this species in South Florida apparently is continuous, but spawning maxima coincide with maximum water temperatures on the Tortugas grounds; therefore, variation in spawning does not likely explain the major waves of recritment observed in the fishery. The initial hypothesis of our research is (a) that two or more major nursery grounds contribute to Tortugas landings, (b) that these grounds produce recruits at different times of the year, and (c) that recruitment from these grounds is affected differently by freshwater inputs and, possibly, other environmental variables. The focus our research has been on (1) sharpening our view of recruitment, (2) developing a juvenile abundance index on inshore grounds, and (3) exploring possible relationships of juveniles and recruits to each other and to various environmental and habitat variables that could possibly affect shoreward transport of postlarvae, juvenile growth and survival, or migration to offshore grounds. Such variables include freshwater inputs (rainfall, freshwater inflow, temperature, wind speed, and mean sea level). This work has been conducted by means of laboratory experiments (1st yr only), field studies, statistical analyses, cohort analyses, and simulation modeling. In addition, we have begun developoment of a preliminary shrimp-based ecological index of conditions in Florida Bay for immediate use in the South Florida Ecosystem Restoration Program.

Historic data from Robblee's and Sheridan's multiyear field studies in Western Florida Bay (Johnson Key Basin) provided the basis for developing a multiyear juvenile abundance index time series. Field sampling of juvenile densities was conducted in Western Florida Bay and Whitewater Bay as part of the Florida Bay Program. Data from these ongoing studies will be used to compare the magnitude and seasonal pattern of juvenile densities in the two nursery areas during the same time period. This is the first time the two nursery areas have been sampled during the same time period using the highly efficient throw trap gear developed by Robblee in the 1980s. General additive modeling (GAM) is being used to determine relationships of the juvenile time series with environmental variables. A long-term monthly data time series of recruits to the fishery was produced by means of length-based cohort analysis using fishery landings and catch-per-unit-effort (CPUE) data. When this time series was overlain on time series of environmental variables with an appropriate time lag, certain patterns were suggested: a general concurrence of the temporal pattern with freshwater input variables during the '60s and '70s, a dampened variability in recruitment and some of the freshwater input variables during the '80s, and lack of concurrence with these environmental variables during the '90s. Several approaches to cohort analyses were tested. It was concluded that the recruitment time series might be improved if a better length-age relationship for this species could be developed and if monthly, rather than annual, estimates of the size frequency distribution within commercial catch categories were obtained.

In developing a shrimp-based ecological indicator, the approach was to remove the "rainfall" effect from annual CPUE data and use the residuals of this relationship as an annual abundance index. An index standardized for rainfall variability would more likely reflect the effect of freshwater inputs resulting purely from water management decisions (e.g., releases of freshwater into Everglades National Park from the upstream Water Conservation Areas and Lake Okeechobee). In South Florida, rainfall variability is a major problem in evaluating the environmental effect of changes in water management structures and operations. Principal findings of our analysis were (1) CPUE of both small shrimp (<= 68 count) and large shrimp ( 68 count) correlated well with Royal Palm rainfall, (2) a marked positive trend occurred in residuals of the small shrimp relationship after about 1980, and (3) a markedly negative trend in residuals of the large shrimp relationship was obvious beginning two or three years later. Inquiies with port agents and shrimp industry representatives revealed a major change in marketing practices in the early '80s that led to increased emphasis on landing small shrimp. This may have had an effect on small shrimp CPUE. Indeed, Tortugas shrimp landings contained a larger proportion of small shrimp after about 1980 than in former years. In subsequent analyses, a dummy variable to account for the change was introduced to the model. This appears to have removed the trend in the residuals. Nevertheless, this is a good demonstration of why results of cohort analysis, because they are not dependent on CPUE, may sometimes be better indices of abundance than CPUE.

Laboratory data on survival as a function of temperature and salinity that had been developed by Zein-Eldin in the first study year recently were used to incorporate a temperature- and salinity-based natural mortality function into the growth and survival model. A second natural mortality function, one dependent on temperature, also was introduced. The two were calibrated so that, under sizes and conditions found on the fishing grounds, their sum would roughly equal previous natural mortality estimates. The model eventually will provide a view of how temporal patterns of salinity and temperature in various parts of Florida Bay and adjacent waters might affect the relative magnitude and temporal pattern of recruitment from each area. Further laboratory studies are needed to extend the experiment to lower temperatures and salinities (presently we only have data for temperatures of 25øC to 35ø and salinities of 30 ppt to 50 ppt) and to provide estimates of effects of temperature and salinity on growth (presently we have information for growth rates as a function of temperature only).

Last updated: 2/26/98
by: Monika Gurnée
gurnee@aoml.noaa.gov