Joan A. Browder, NOAA, National Marine Fisheries Service, Miami, FL; Nelson M. Ehrhardt, Victor R. Restrepo, Univ. Miami Rosenstiel School of Marine & Atmospheric Science; Peter Sheridan, Z.P. Zein-Eldin, James Nance, NOAA, National Marine Fisheries Service, Galveston, TX; Cheryl Woodley, NOAA, National Marine Fisheries Service, Charleston, SC; Michael Robblee, National Biological Service, South Florida/Caribbean Field Unit, Miami, FL.

The Tortugas fishery for pink shrimp (Penaeus duorarum) experienced a severe decline in catch rates from the mid 1980s through the early 1990s. The decline appeared related to reduced recruitment, because lower catch rates coincided with decreased, rather than increased, effort, and the size distribution of the shrimp did not change. Components of the fishery have followed different time trajectories. Recruits to the fishery during the 6 months from July through December declined from 1960 through 1993, but recruits during the 6 months from January through June were more stable. This suggested there might be two cohorts in the fishery and that the decline mainly involved one cohort.

A multi-investigator study was initiated in 1994 to explore the decline in the fishery and to examine it in relation to conditions on known nursery grounds, particularly Florida Bay. The decline in pink shrimp recruitment was only one of several recent signs of deterioration in the ecological health of Florida Bay. This investigation addresses questions in the Florida Bay Interagency Science Plan concerning recruitment to coastal fisheries as related to nursery ground conditions.

Pink shrimp is viewed as a potential ecological indicator species in the adaptive environmental management process of the South Florida restoration effort. Shrimp landings are positively correlated with indices of freshwater runoff, and loss of freshwater inflow is a major hypothesis for the bay's decline. Water management actions in the early 1980s are thought to have had a major influence on freshwater flow to Florida Bay and other estuaries of Everglades National Park. In particular, after floods in August and September of 1981, regulation water stages were lowered in the South Dade Conveyance system (L-31N, L-31W, and C-111 canals). Further lowering of regulation stages took place in early 1984. This has diverted water away from Everglades National Park and Florida Bay. Pink shrimp may also act as an indicator of the health of seagrass beds. Seagrass beds have been shown to be favorable habitat for pink shrimp. A die-off of seagrass beds is one aspect of the declining ecological health of Florida Bay.

Learning more about pink shrimp ecology is the key to using this species as an ecological indicator. Three critical questions helped focus study design: 1) What are the functional relationships between pink shrimp growth and survival and salinity and temperature on their nursery grounds?, 2) Where are the nursery grounds of each of the cohorts in the fishery and what time of the year are they used?, and 3) Is there more than one physiological phenotype in the fishery, responding differently to freshwater inflow?

The first question is fundamental. A positive correlative relationship between Tortugas shrimp landings, adjusted for effort, and freshwater inflow to the coast has been established by previous work of some of these investigators. For almost a decade, the National Marine Fisheries Service has used a model based on freshwater inflow indices to accurately forecast Tortugas landings. The mechanism underlying the relationship is not understood because the geographical distribution of pink shrimp suggests this species tolerates higher salinities than either of the other two penaeids in U.S. waters. Therefore, one might think that freshwater inflow would be less important to pink shrimp than to the other species.

The second question addresses another paradox. Whitewater Bay, a water body in which salinities are highly variable and seasonally brackish, was considered an example of important pink shrimp nursery habitat in studies conducted during the 1950s and 1960s. In fact, a bait shrimp fishery once operated in passes to Whitewater Bay and Lake Ingraham. Yet recent studies of pink shrimp on their nursery grounds have focused on Florida Bay and have found the highest densities in the western part, where salinities are relatively stable and approach seawater strength. Salinity is an important variable influencing physiological processes in estuarine organisms. How does the same species use nursery grounds having such different salinity regimes?

It is has been shown for some other organisms that genetic variability allows a species to conform to a wide range of conditions without necessarily requiring adaptation of the same individuals to the entire range. The third question, therefore, is: Does more than one physiological phenotype contribute to Tortugas pink shrimp landings? If so, where are the nursery grounds and what time of the year are they used? Two or more cohorts could, of course, originate from the same grounds and be exposed to different conditions because they utilize the grounds at different times of the year. But only one annual maximum in juvenile densities has been observed in western Florida Bay, which suggests that another location is a more likely nursery ground for the second cohort.

One objective of the FY94 study was to better define the within-year cohorts in the fishery and relate them back to cohorts on nursery grounds that might have contributed to the fishery. Virtual Population Analysis (VPA) was used to estimate abundance, by size and age groups, from landings data. A computer model was developed to simulate shrimp growth as a function of temperature and determine the month and day of maximum shrimp catches (e.g., Nov., April, etc.), by size, resulting from any month and day (e.g., mid June or mid Nov.) of maximum abundance of juveniles on the nursery grounds.

A second objective in the first year was to explore the full range of environmental variables that might be influencing recruitment. Exploratory analyses of archived resource survey and environmental monitoring data were undertaken to identify environmental variables statistically related to juvenile pink shrimp densities on the nursery grounds and recruitment to the fishery computed with VPAs.

A time series of modal sea surface temperatures for the bay was developed for use in the simulation model. These data were computed from 5-day composite SST fields prepared from 4-km (to a side) resolution SST fields prepared by the University of Miami from NOAA satellite AVHRR data.

To support expansion of the simulation model to include effects of salinity, as well as temperature, laboratory experiments were conducted to measure survival and growth of young pink shrimp under different temperature and salinity combinations. Another aspect of Yr-1 work was to lay the groundwork for examining genetic variation in the pink shrimp contributing to the Tortugas fishery.

Yr-1 analyses supported our initial hypothesis that two or more major within-year age cohorts recruit to the Tortugas shrimp fishery. The analysis confirmed our cursory observation that the decrease in landings and CPUE observed from the mid 1980s through the early 1990s was due almost entirely to the decline in the fall cohort. This suggests that some change in conditions has severely affected the cohort that historically contributed most strongly to the fishery. The within-year cohorts in the fishery may come from different nursery grounds and may survive and grow best under different conditions.

Our simulation model showed that differences in growth rate during the first 30 days on the grounds resulted in substantive differences in the number of days between peak densities on the grounds and peak recruitment in the fishery. Results showed that juveniles exposed to the temperatures prevailing in the bay starting in July, in September, or in November grew at markedly different rates.

After correction of some translation errors in the historic salinity file, a strong relationship was found between Tortugas landings and salinities in Florida Bay. This result brings us closer to defining and understanding the mechanism(s) underlying the correlative relationship observed between pink shrimp landings and indices of freshwater inflow.

Yr-1 exploratory analyses of juvenile pink shrimp densities in relation to environmental variables identified significant relationships with Key West mean sea levels, wind speed, rainfall (at the Royal Palm Ranger Station), and water releases into Everglades National Park through the S-12 structures. A model based on these variables is a good estimator of the temporal pattern of juvenile densities. Rainfall and water releases through the S-12 structures are indicative of freshwater inflow to the coast.

Yr-1 laboratory experiments with pink shrimp collected from western Florida Bay indicated reduced survival at temperatures exceeding 30°C, often found in summer, and salinities of 45 ppt or greater. These salinities are common and persistent in northcentral Florida Bay basins.

Summarization of 8 yrs of remotely sensed sea surface temperature data from Florida Bay suggests that the bay had more days of temperatures greater than 30°C from 1988 through 1992 than it had from 1985 through 1987. We found that routine methods used to compute sea surface temperature from satellite imagery are inadequate for providing precise estimates of temperatures greater than 30.5°C.

Genetics work conducted during Yr-1 resulted in the design and optimization of DNA extraction protocols to produce sufficient quantities of DNA for polymerase chain reaction (PCR) techniques and cloning projects. A species-specific genetic marker that distinguished pink shrimp from brown and white shrimp (Penaeus aztecus and P. setifera) was identified in mtDNA. Preliminary analyses have been conducted to detect mtDNA regions of variability that might be useful in addressing population level questions in pink shrimp. Simultaneously with the mtDNA work, we started preparing a "shotgun" genomic library to screen for microsatellites, which have been powerful population-level markers in marine mammals and teleosts.

Our Yr-1 work showed, for the first time, a correlation of juvenile densities with freshwater inflow. Previous work had suggested an effect of freshwater inflow on landings, but this new analyses suggests an effect of freshwater inflow on early survival, as reflected in juvenile densities on the nursery grounds. Also for the first time, a significant relationship was found between Tortugas landings and salinities in Florida Bay. This helps to support a causal relationship in the correlation of landings with indices of freshwater inflow.

Our laboratory results concerning tolerances to temperature and salinity suggest that parts of Florida Bay may be excluded as pink shrimp nursery grounds in most years. Temperatures greater than 30°C often are found in Florida Bay in the summer, and salinities greater than 45 ppt are persistent in some parts of Florida Bay through all but the wettest years.

Another significant Yr-1 finding is the suggested linkage of densities with oceanic water levels and, through wind stress, Ekman transport. Both might relate to transport processes carrying pink shrimp from their offshore spawning grounds to their inshore nursery grounds. The explanatory strength of these variables suggest that we should determine whether substantive changes in oceanic patterns occurred within the period from the mid 1980s to the early 1990s that could have affect larval transport.

On the other hand, the relationship with sea level may reflect some seasonal or long term influence of sea level on shrimp habitat. Juvenile shrimp appear to occur at highest densities on banks and flats. These areas are, to varying degrees, either exposed or subjected to extremely shallow water during lower stages of the tide; and the mean tide varies seasonally, causing the sea level at low tide to be lower during some times of the year. These sea level variations may affect shrimp growth and survival.

Although shrimp landings were extremely low from about July, 1986-June, 1987, through July, 1992-June, 1993, they increased greatly in 1993-1994 and 1994-1995. The increase corresponds to the higher rainfall of the last few years, appearing to further confirm a causal relationship between catch rates and freshwater inputs. This supports the focus of our Yr-2 work.

Recently, we initiated a field study to compare pink shrimp densities in Whitewater Bay with densities being measured in western Florida Bay in a related project. The same gear and procedures will be used. This will be the first time that the throw trap, a more efficient gear than that used previously in Whitewater Bay, will be used concurrently in both western Florida Bay and Whitewater Bay.

Critical questions our continuing research addresses are: 1) What nursery grounds provided the main support for the fishery before its productive value declined?, 2) What nursery ground supported the fishery during the 9-yr period of exceptionally low catches, 3) Which environmental variables are most important in determining pink shrimp recruitment, and 4) What are the mechanisms for their effects? The fact that catch rates in the fishery have improved in the last two years may help us answer these questions.

Herrnkind, William F., Florida State University, Tallahassee, FL 32306; Mark Butler J. IV, Old Dominion University, Norfolk, VA 23529; John H. Hunt , Florida Department of Environmental Protection, Florida Marine Research Institute, Marathon, FL 33050. .

Florida Bay and the shallow waters surrounding the Florida Keys are the primary Florida nursery for the Caribbean spiny lobster (Panulirus argus). However, environmental conditions are deteriorating in the Florida Keys and a cascade of ecological disturbances have recently plagued the region, thus revealing the coupled dynamics of this tropical marine ecosystem. Our studies of spiny lobster recruitment, set against this backdrop of environmental change, demonstrate how changes in the coastal environment can affect key fishery species, sometimes in unexpected ways. For example, salinity, temperature, and nutrient loads have increased in portions of Florida Bay. Changes in water quality have presumably contributed to the loss of thousands of hectares of seagrass in Florida Bay, and together have sparked the development of cyanobacteria blooms that have become widespread in the bay since 1991. An unexpected consequence of the cyanobacteria blooms was the decimation of the sponge community throughout much of central Florida Bay. The ramifications of these environmental changes for the lobster population in south Florida are complex. Hardbottom habitat replete with macroalgae, sponges, solution holes and other structures are thought to be prime nursery habitat for lobsters, but recent results suggest that seagrass may be more important in this regard than we previously thought. Although the implications of seagrass loss for lobster recruitment might therefore seem obvious, the actual repercussions are probably minimal because the areas impacted thus far are largely isolated from postlarval supply or have salinity and temperature regimes that are intolerable to settling postlarvae. In fact, abnormal salinity or temperature in much of northeastern Florida Bay diminish the recruitment potential of lobsters. The most serious threat to lobster recruitment to date is the widespread mortality of sponges. Juvenile spiny lobsters rely on sponges for shelter, so the rapid loss of sponges at sites exposed to cyanobacteria blooms has resulted in similar declines in local lobster abundance and shifts in shelter use by the remaining individuals. We recently began investigating the ramifications of ecosystem change, particularly the loss of sponges, on lobster recruitment on a regional scale by coupling large-scale surveys with a spatially explicit, individual-based model of the system. The initial predictions from the model are that lobster recruitment will decline 2 - 19%, depending on the availability of alternative shelters, which corresponds with a field survey based estimate of a 10% loss in new recruits.

William G. Lyons, Florida Department of Environmental Protection, Florida Marine Research Institute.

Objectives: Acquire information on the composition of faunal assemblages and map their distributions in Florida Bay, using mollusks as indicator organisms.

Mollusks are important components of the Florida Bay fauna. More than 200 species of mollusks live in brackish, estuarine, and marine waters of Florida Bay (Tabb and Manning, 1961; Tabb et al., 1962; Turney & Perkins, 1972), and their shells in bay sediments constitute more than 75% of all particles whose sizes exceed 1/8 mm (125 microns) (Ginsburg, 1972), prompting Turney (1972:15) to declare the fauna of the bay to be dominantly molluscan. The concept of Florida Bay "sub-environments" used today was promulgated on evidence from distributions of dead (empty) mollusk shells (Turney, 1972; Turney & Perkins, 1972). This report documents assemblages of live mollusks in the bay during summer 1994 and compares their distributions to the sub-environments described from dead shells several decades ago. Results will provide a baseline for measuring changes due to environmental perturbation or to efforts to mitigate such disturbance.

Methods: To avoid potential biases of a priori stratification, the initial search for distributional pattern applied equal sampling effort throughout the bay using an iteration of the U.S. Environmental Protection Agency hexagonal grid (White et al., 1992) that subdivided the study area into approximately 300 units (area of each unit 6 km2). By choosing every third unit, a sampling pattern was obtained that covered the bay evenly with 101 noncontiguous units. Global positioning system (GPS) technology was used to locate units and to identify sampling sites, which were selected within the prevalent environment of each unit. Water temperature, salinity, conductivity, pH, and dissolved oxygen were measured at each site; surface and bottom values were recorded when depths exceeded 1 m. Bottom sediments (mud, clay, sand, peat, etc.) and vegetation were also noted.

Fifteen 6-inch (15.24-cm) diameter cores were taken at each site. Cores were washed through two sieves (mesh sizes 3 mm [upper] and 1 mm [lower]), and sieved residues of three consecutive cores were combined to produce five samples of each size fraction at each site. A core sampled a surface area of 183 cm2 (0.018 m2); surface area of samples combined from three cores was 0.054 m2, and five three-core samples represented an area of 0.27 m2.

Large-fraction (>3-mm) samples were sorted into five categories (mollusks; annelid worms; arthropods; echinoderms; other phyla); 10% of the samples were re-sorted to ascertain rates of error. Sample residue, principally empty shells, was air-dried and stored for use in other comparisons. Live-collected mollusks were identified to the lowest taxonomic unit (usually species) and counted. Resultant data sets were analyzed using the Community Analysis System (CAS) program. Coefficients of intersite similarity [abundance data transformed, logn (x+1)] were used to identify faunistically similar groups (assemblages) whose distributions were mapped using GIS techniques.

Results: Sampling in 1994 was conducted during 1-23 June (71 sites); 7-14 July (2 sites); 3-17 August (28 sites); and 31 August (2 sites resampled; original samples poorly preserved). Fifteen cores were taken at each of 101 sites (total cores 1515) and combined in sets of three (1515/3 = 505 samples). Surface area sampled at each site was 0.27 m2; total surface area sampled was 27.27 m2.

Salinity ranged from <10 ppt. in Joe Bay to 50-52 ppt. at three sites near Rankin Key. Sites where salinity exceeded 40 ppt. were common in the central bay. Salinities in the eastern bay were markedly lower (31-35 ppt), even in early and middle June before the full onset of summer rains.

Live-collected mollusks consisted of 94 species, 1,436 species lots and 13,774 specimens. Numbers of species at sites ranged from 1 to 23 (mean = 7.3), but 4 to 9 species were taken at nearly two-thirds (63%) of the sites. Sites with higher species richness (>9 species) generally were located west of a line extending from Flamingo to Lignumvitae Key, but high species richness also occurred at Buttonwood Sound, Blackwater Sound, and at several sites along the Intracoastal Waterway north of the more easterly Florida Keys. Sites with fewest (1-3) species were usually located in the upper bay near and east of Rankin Key or in basins of the eastern bay.

Molluscan abundance varied greatly among the sites; numbers of specimens per site ranged from 1 to 6406, and maximum density was more than 23,700/m2. Five sites where density exceeded 1,000 mollusks/m2 were dominated by a mussel, Brachidontes exustus. Those sites tended to have below-average species richness (<7) but also included the site of highest species richness, Pontoon Bank. Mussels occurred in low density (<200/m2) at most sites in the eastern half of the bay, but the species displayed great increases in density (1,000-23,725/m2) at several sites along a track extending from western Madeira Bay southwestward through Whipray Basin and Twin Key Basin to Pontoon Bank. This species was virtually absent from the northwestern part of the bay; a single juvenile was found at 1 of 29 sites north and west of Pontoon Bank.

Most species occurred in zones of semi-contiguous distribution, indicating affinity for particular subenvironments. Coefficients of intersite similarity indicate a Gulf group (7 sites; 33 species), a Lake group (2 sites; 3 species), and a Bay group (92 sites). The Bay group includes an Eastern component (45 species; 29 sites), a Central component (35 species; 18 sites), a Western component (57 species; 33 sites, with a southern and two northern subgroups), and 15 sites either transitional between larger components or too depauperate to classify. The Central component generally tracked the distribution of highest salinity and the path of concurrent blooms of microalgae. Most Central sites were species-poor but several were specimen-rich, reflecting the high densities of Brachidontes exustus.

Site groupings of assemblages of live mollusks in 1994 generally confirmed the locations of the Interior, Transitional and Gulf sub-environments proposed by Turney & Perkins (1972), but the Northern and Atlantic sub-environments they defined were not distinguished from other groupings in 1994, probably because of the influence of widespread hypersalinity.

Continuation of Work: Thirty sites, proportionally allocated among 7 groups and subgroups of 1994, were resampled in August 1995 to test consistency of groupings. Salinity was measured again at all 101 sites for comparison with 1994 values. Fifteen of the sites will be resampled several more times during the year to ascertain the influence of seasonal variation in species abundance on site groupings perceived in summer mapping efforts.

References:

Ginsburg, R. N. 1972. Introduction to Recent sedimentation. Pp. 4-11 in R. N. Ginsburg, ed. South Florida carbonate sediments. Sedimenta II. University of Miami, Fisher Island Station, Miami Beach, Florida.

Tabb, D. C., D. L. Dubrow, and R. B. Manning. 1962. The ecology of northern Florida Bay and adjacent estuaries. State of Florida Board of Conservation, Technical Series No. 39:1-79.

Tabb, D. C., and R. B. Manning. 1961. A check list of the flora and fauna of northern Florida Bay and adjacent brackish waters of the Florida mainland collected during the period July 1957 through September 1960. Bulletin of Marine Science of the Gulf and Caribbean 11(4):552-649.

Turney, W. J. 1972. [Florida Bay] Molluscan fauna. Pp. 14-16 in R. N. Ginsburg, ed. South Florida carbonate sediments. Sedimenta II. University of Miami, Fisher Island Station, Miami Beach, 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.

White, D., A. J. Kimmerling, and W. S. Overton. 1992. Cartographic and geometric components of a global sampling design for environmental monitoring. Cartogr. Geograph. Inf. Syst. 19(1):5-22.

Michael B. Robblee, National Biological Service, South Florida/Caribbean Field Laboratory, Florida International University, Miami, FL.

In the fall of 1987, a widespread, rapid die-off of turtle grass, Thalassia testudinum, began in Florida Bay. Die-off occurred in areas of dense seagrass cover and principally in and around Rankin Lake, Rabbit Key Basin and Johnson Key Basin in western Florida Bay. Increasingly extensive and persistent turbidity and algal blooms, apparently linked to the loss of seagrass cover, have been associated with active seagrass die-off sites since 1988 and have characterized western and central Florida Bay since 1991. Recolonization of impacted grass bed habitats by shoal grass, Halodule wrightii, is occurring.

Loss of seagrass habitat on the scale observed in Florida Bay is unprecedented in tropical seagrass systems and hypothesized to threaten the Bay's water quality, sportfishery, and nursery function. In the short-term, grass canopy loss and declining environmental conditions may lead to shifts in species composition and reduced abundance of grass canopy dependent organisms. Over the long-term increasing seagrass habitat heterogeneity may lead to enhanced nursery function and an improve sportfishery.

A detailed quantitative database is available from Johnson Key Basin from October 1984 to April 1987 prior to seagrass die-off. Limited additional data is available from between May 1989 and August 1991, a period following the onset of seagrass die-off in the Bay but prior to the extensive and persistent plankton blooms which have characterized it since 1991. For caridean shrimps, fishes and pink shrimp this database documents population and community dynamics and spatial relationships with seagrass habitat. Therefore, it provides an excellent baseline against which to observe the response of characteristic seagrass associated species in Florida Bay to grass canopy loss, seagrass community change, and changing environmental conditions following seagrass die-off. The purpose of this project is to duplicate over the period, October 1994 to April 1997, the experimental design and sampling protocols employed previously in Johnson Key Basin prior to seagrass die-off in order to address the following objectives: 1) to document changes in seagrass community structure and habitat complexity following seagrass die-off; 2) to document changes in species composition, abundance, and seasonality of caridean shrimps and fishes with changes in seagrass habitat; 3) to document temporal and spatial abundance and size frequency distribution of the pink shrimp, Penaeus duorarum, in relation to changes in seagrass habitat; and 4) to evaluate quantitative relationships between animal abundance and species composition and grass bed micro-structure and habitat complexity.

In 1984 thirty stations were established in Johnson Key Basin. Stations were located generally with no a priori consideration of the seagrass habitat present. The stations were evenly stratified among the principal seagrass macro-habitat types present in Florida Bay: bank, basin, and near-key. Nine of these thirty stations, 3 within each macro-habitat type, were repetitively sampled on a six-week interval between October 1984 and April 1987 in order to address questions of timing. These nine stations were also sampled between August and December on a six-week interval between 1989 and 1991. All thirty stations were sampled four times (January 1985, May 1985, May 1989, and January 1990) in order to address animal versus habitat questions.

Quantitative animal samples were collected using a throw trap. The throw trap consisted of an open­ended 1 m2 aluminum box, 45 cm deep, with panels of nylon netting (0.16 mm stretch mesh DELTA netting) attached on parallel edges at the top of the throw trap. Each panel of netting was large enough to cover the top of the throw trap when it was used in water deeper than 45 cm. At each station four replicate throw trap samples were located along a 20 m transect, one each in each 5 m segment. After the trap was dropped in place, it was cleared of animals with three separate passes of a 1 m wide frame sweep net of mesh size similar to the panels. It has been estimated that three sweeps collect at least 95% of target species present in the trap. SCUBA was used while clearing the trap in deep water. All fishes, caridean and penaeid shrimps were removed from each throw trap, identified, counted and sized as appropriate in the laboratory. When all thirty stations were sampled the connection between these seagrass associated animals and grass bed structure was made by associating each throw trap collection with estimates of grass bed micro-structure including: seagrass standing crop and blade density; algal biomass; sediment texture, depth, organic content and compaction; root and rhizome biomass; and water depth.

During the current effort six-week interval sampling has been ongoing since October 1994 (9 collections have been made) and the thirty stations have been sampled in January and May 1995. Dr. Mike Marshall of Mote Marine Laboratory is processing samples and identifying the organisms recovered. To date processing efforts have emphasized back-logged samples originally collected between 1989 and 1990 and recent samples collected during January and February 1990. Results reported here focus on a comparison among the thirty stations sampled in January/February of 1985, 1990 and 1995.

Not all of the thirty stations were sampled in each of the three years. Extreme low water in Florida Bay characteristic of January and February precluded sampling high bank stations successfully in all three years. Because of this results reported here are based on the twenty-four stations (5 bank, 9 basin and 10 near-key) sampled in each year. At these sites distinct changes in seagrass habitat have occurred since 1985. By 1995 the standing crop of Thalassia had declined by 82% as compared to 1995. Similarly, Halodule had declined by 53% and Syringodium has completely disappeared. These changes resulted in a marked shift in seagrass dominance among the 24 stations. In 1985 Thalassia was the dominant grass at 17 of 24 stations in Johnson Key Basin; Halodule dominated at 5 stations, all of them near-key habitats. By 1995 Thalassia dominated at only 9 stations while Halodule had expanded its presence into deeper water and dominate at 11 of 24 stations. Further, by 1995 Syringodium, never common, had disappeared and a new habitat, bare sediment, not present in 1985, dominated at 4 stations.

The abundance of seagrass associated caridean shrimps, fishes and pink shrimp was lower in 1995 when compared to either 1985 or 1990. Habitat change due to seagrass die-off in 1990 was localized within Johnson Key Basin and apparently affected only 6 of the 24 stations unlike 1995 when all but 6 stations evidence significant habitat change. Species composition differences were evident in 1995 when compared to 1985 and 1990. The killifishes, Lucania parva, and Floridichthys carpio, and the toadfish, Opsanus beta, were present in significantly lower numbers than in previous years. In contrast, the code goby, Gobiosoma robustum, and the bay anchovy, Anchoa mitchelli were found in greater numbers. The killifishes were community dominants within the Johnson Key Basin grass bed prior to widespread habitat change and virtually absent in 1995. Among the caridean shrimps, Alpheus sp. have increased in abundance, however, this result may be a sampling artifact due to the loss of the grass canopy. Evidence was not found supporting the hypothesis that increasing coverage by Halodule would translate to increasing recruitment of the pink shrimp, Penaeus duorarum.

At this point the project will continue with six-week interval sampling through April 1997. On completion, effects of seagrass die-off on seasonal timing can be addressed in addition to animal/habitat relationships. Additionally, data collected in this project will continue to contribute to ongoing statistical and population dynamics models of the pink shrimp in relation to the Tortugas fishery and to the expansion of the ATLSS fish and invertebrate model into Florida Bay.

John M. Stevely, Florida Sea Grant Extension Program, 1303 17th St. W., Palmetto, FL 34221-2998, (941) 722-4524, fax (941) 742-5998; Donald E. Sweat, Florida Sea Grant Extension Program, 36702 Highway 52, Dade City, FL 33525-5198, (904) 521-4288, fax (904) 523-1921.

The work described here was initiated in response to concerns regarding ecological and fishery impacts resulting from increased sponge harvesting effort in the late 1980s and early 1990s. The objective of the initial phase of the work was to document and quantify the contribution of commercial sponges (sponges of the genera Hippospongia and Spongia) to total sponge community biomass.

During 1991 and 1992 a total of 15 areas were sampled (five areas north of Long Key, four areas within Everglades National park, two areas west of Everglades National Park and four areas north of Marathon). The total area surveyed was 34,620 m2. Sampling methodology consisted of counting all sponges found within sixteen 100-m x 2-m transects at each area. During this phase of the study specific numerical abundance was recorded only for commercial species (Hippospongia lachne, Spongia barbara, Spongia graminea) and largest most common species (Spheciospongia vesparia, Ircinia campana, Ircinia strobilina, and Ircinia spp.). All other sponges were lumped into a miscellaneous unidentified category. In addition to numerical counts, data on volumetric biomass of the different sponge species and sampling categories were collected. This methodology consisted of estimating sponge specimen volume by measuring the volume of water displaced when the sponge was placed in a bucket fitted with an overflow spout.

The mean abundance for all sponges was 7,250/hectare and for commercial sponges was 106/hectare. The mean volumetric biomass of all sponges was 364 ml/m2. These data represent the most compressive baseline information available on sponge community biomass in portions of the area subsequently affected by sponge mortalities. All methods employed to estimate sponge biomass indicated that the contribution of commercial sponge biomass to the total sponge community biomass was relatively small (1.4% based on numerical counts, 2.4% based on volumetric estimates). Two species of sponges Spheciospongia vesparia (loggerhead sponge) and Ircinia campana (vase sponge) represented 68% of the total sponge community biomass based on volumetric estimates.

The effects of sponge mortalities, apparently caused by cyanobacterial blooms, became apparent in late 1992. Consequently, a second phase of work was begun to focus on documenting the effects of the sponge mortalities on sponge populations and monitoring the recovery or lack of recovery of sponge populations in following years. Fiscal and time constraints limited follow-up work to two areas (one off of Long Key and one of off Marathon) in 1993 and 1994, In 1995 a third area in Everglades National Park (west of Arsenicker Keys) was added to the sampling regime.

Results of the 1993 field work documented highly significant declines in sponge abundance, with a reduction of up to 90% of the sponge community volumetric biomass. Based on sampling data and general field reconnaissance, the severity of the mortality varied significantly over the entire area affected by the cyanobacterial blooms. Data indicated that the sponge Cinachyra sp. was the most resistent to the sponge mortalities; sponges of the genera Spongia, Hippospongia and Ircinia appeared to be among the most susceptible.

The loggerhead sponge (Spheciospongia vesparia) appeared to more resistent than many species, but was completely eliminated throughout extensive areas.

As work progressed in 1994 and 1995 we began to develop a more comprehensive listing of the species previously lumped in the miscellaneous unidentified category. The most recent work now includes counting 20 sponge species. Therefore, if future work is funded, a much more complete analysis of sponge communities in the sampling areas will be possible. Analysis of the 1994 and 1995 data has not yet been completed. A preliminary perusal of the data indicates some initial evidence of recovery in the Marathon area, and possibly increased abundance of some sponge species compared to pre-mortality conditions. Due to the manner in which of the data were collected in 1991 and 1992 it may not be possible to conclusively document which species are proving to be the most rapid recolonizers. Data collected at the third monitoring area in Everglades National Park, beginning in 1995, provides a dramatic contrast with the other two areas sampled in 1993 and 1994. Data from this area indicates an even more dramatic impact, and may suggest the area has been subjected to additional mortality events.

The work conducted to date has been supported by the Florida Department of Environmental Regulation Florida Marine Research Institute, Florida Sea Grant College Program, and Keys National Marine Sanctuary Program. The 1995 work was conducted in collaboration with Dr. Shirley Pomponi, Harbor Branch Oceanographic Institution, and the FDEP contract supporting this work will be completed by November, 1995.

Currently, no funding has been obtained for future work. It is abundantly clear that many years will be required for sponge populations to recover to pre-mortality conditions. Therefore, future long term funding is needed to continue and build upon the work that has been accomplished. Without such a long term commitment to document the recovery of sponge populations, efforts to assess alterations of hardbottom communities, monitor ecological conditions, and model Florida Bay food webs will be severely hampered.

Last updated: 07/16/98
by: Monika Gurnée
gurnee@aoml.noaa.gov