The Effects of Hydrology on Fishes of the Florida Bay Mangroves Zone

Jerome J. Lorenz, National Audubon Society, Tavernier FL.

The mangrove-dominated zone that forms the interface between the Everglades and Florida Bay and Biscayne Bay is an important area for wildlife in south Florida. This polyhaline dwarf red mangrove (Rhizophora mangle) forest is a primary habitat of the endangered American Crocodile and West Indian Manatee and is heavily used as a feeding ground by wading birds. Historically, this ecotone may have been an important nursery ground for game fishes. Prior to the C&SF project, freshwater from the Everglades reached the Florida Bay mangroves via Taylor Slough. In recent decades, the hydrology of the mangrove zone has been altered by the construction and operation of a system of canals and levees. Outstanding among these is the C-111 canal, which diverts water from Taylor Slough to an area just east of US Highway 1, thereby disrupting the natural processes in both areas. Further complicating the freshwater flow pattern is the US Highway 1 levee which has blocked all sheet flow from the Everglades to Biscayne Bay, however large pulses of freshwater reach Biscayne Bay during storm events via the C-111 canal.

There is strong evidence that, over the past 50 years, the flora and invertebrate fauna of the mangrove zone has changed from predominantly fresh water species to salt tolerant species, indicating that the marine environment has recently intruded into this area. During this same period, the number of wading birds nesting on Florida Bay keys has declined precipitously. Since the primary foraging area for these birds is the ecotonal area, it is possible that these physical and ecological changes have adversely affected the quality of the areas as foraging grounds for wading birds. The basic hypothesis of this study is that prey base fish populations have ultimately been impacted by these changes in the hydrological regime over the last several decades, and that these changes in prey base may be implicated in the decline of wildlife throughout the Florida Bay region.

To accomplish our goal of linking manmade changes in water delivery to resident species abundance, we sampled fish populations in the mangrove zone of Biscayne and Florida Bays from August 1990 through present at four to six week intervals. A 9 m2 drop trap designed specifically for sampling fish in this environment was used to collect fish samples. In conjunction with these samples, a detailed analysis of hydrologic conditions was made through the use of continuous recorders. Sites were selected based on the proximity and hydrologic influence of Taylor Slough. Three sites north of Florida Bay were selected based on a decreasing west to east gradient of the influence of the Slough. TR (Taylor River site located in eastern Taylor Slough) received the most fresh water, HC (Highway Creek site just east of US1) received the least and JB (Joe Bay site located northeast of Joe Bay) was intermediate. The first samples were collected at JB and HC in August 1990. Collections began at TR in December 1991. A fourth site, BS (Barnes Sound site located southwest of Card Sound Bridge) was sampled starting in February 1992. Historically, BS received fresh water surface flow from the Everglades, but the construction of US1 and Card Sound Roads had impounded the area, effectively cutting off all sheet flow connection to the Everglades.

A wide variety of hydrologic conditions occurred over the six years of data collection ranging from drought conditions in 1990 to extremely high water events in 1994 and 1995. Combined with the spatial variability in fresh water flow to the sites, this interannual variation will be used statistically to demonstrate how different hydrologic regimes affect the prey base. Preliminary numerical analysis indicate that prey fish density and biomass are adversely affected by lower rates of fresh water inputs however, there appears to be some maxima beyond which more fresh water input resulted in no net increases in fish abundance.

Data collections will continue through August, 1997 and at that time the possibility of continuing this type of work will be evaluated. How the amount of fresh water flow into the system impacts the prey base is not well understood and is the subject of associated studies. Surveys of macrophyte vegetation are being conducted in connection with fish collections. The response of the macrophyte community to hydrological conditions may provide insight into the fluctuations in prey base fish populations. Growth enclosure studies are being performed at all sites using 1 m2 enclosures to determine how different hydrologic conditions affect growth of fishes and epiphytes. Stable isotope and gut analyses are being performed on fish specimens collected over the six year period to determine trophic relationships of the prey base community. An evaluation of the impact of the exotic Mayan Cichlid is ongoing. Recent investigations include an analysis of the life history of this species in the southern Everglades. Future experiments will look at the impact of this species on the prey base community and its relative value as a prey item to game fishes and wading birds. Funding for past and ongoing studies has been provided by the South Florida Water Management District, the U.S. Army Corps of Engineers and the John D. and Catherine T. MacArthur Foundation.

Florida Bay as a Nursery Ground: Assessing Larval and Juvenile Fish Communities

Allyn B. Powell and Donald E. Hoss, NOAA, National Marine Fisheries Service, Beaufort, NC; and James Colvocoresses, Florida Marine Research Institute (FDEP), Marathon, FL.

Unprecedented seagrass die-off events accompanied by algae blooms and a reduction in water quality in Florida Bay have caused concern as to the effects these disturbances might have on the function of Florida Bay as a nursery area for fishes. Of particular concern is the spotted seatrout, Cynoscion nebulosus, that spends its entire life history in estuarine waters and whose prime spawning habitat is located in the western portion of Florida Bay, an area where extensive seagrass die-off has occurred. To determine if these disturbances affected the juvenile and larval fish community, we have been comparing the abundance and distribution of juvenile and larval fishes prior to the seagrass die-off (1984-1985) to their present (1994-1995) distribution and abundance.

Since August of 1994 FDEP has been conducting a monthly trawl (6.1 m otter trawl, 3 mm mesh codend liner) and seine (22.3 m center bag seine, 3 mm mesh) survey at 31 fixed sites throughout Florida Bay (Colvocoresses). Qualitative comparisons of the preliminary results of this survey with historical data collected with similar gears indicates that the species composition and distribution patterns of many fish species has remained similar on a bay-wide basis, but that dramatic changes in species composition and relative abundance have occurred at sites in western and central Florida Bay where extensive seagrass die-off has occurred. This observation supports those by staff of NMFS, USGS and DEP in areas of seagrass die-off. Although the populations of small forage fishes and juvenile gamefishes vulnerable to these sampling gears have undoubtedly been impacted by the recent ecological disturbances, especially in the areas that undergone severe structural habitat changes, on a large scale basis these results indicate that the mobility and generally wide environmental tolerances of the estuarine fish community may serve to make it more resilient to these changes than other, more immobile, portions of the Florida Bay fauna. There are indications (see Thayer et al., This Symposium) from sampling with a high speed, small mesh (3 mm mesh cod end) bottom trawls, which more effectively sample the planktivorous fish community that certain planktivores have increased in abundance during recent times, suggesting that a shift in trophic patterns in the fish community has occurred.

To compare the present assemblage of larval fish with the historical assemblage, we sampled ichthyoplankton from September 1994 through August 1995 using the same sampling techniques, stations, and in the same months as in 1984-1985. In addition, we expanded our sampling to include previously non-sampled areas (central and northeastern Florida Bay) to increase our knowledge of spotted seatrout spawning habitat. A decadal comparison will be accomplished when samples, which are being sorted by the Polish Sorting Center, are processed.

The spawning habitat of spotted seatrout has now been defined. The collection of trout larvae at stations previously not sampled, along with data collected in 1984-1985, indicated that spotted seatrout spawn: (1) in the western portion of Florida Bay adjacent to the mainland from Cape Sable to Crocodile Point and adjacent estuarine waters where salinities are greater than 20 ppt; (2) along the edge of the Gulf of Mexico; and (3) in a portion of the central and western interior. Spawning does not appear to occur in waters adjacent to the Keys, in eastern Florida Bay adjacent to the mainland, or in waters approximately <20 ppt. Otoliths of spotted seatrout larvae and juveniles are being examined to (1) analyze hatchdate distributions and growth of larvae and juveniles to determine if differential survival and growth exists among cohorts, (2) identify environmental factors that could influence survival and growth, and (3) compare larval growth and spawning intensity along an east-west transect and compare these with changes in environmental factors (e.g., temperature and salinity).

A protocol for examining daily increments in otoliths was established from 68 juveniles collected from May through September, 1995 with an otter trawl and from 14 larvae collected from June through August, 1995 with ichthyoplankton gear. There was a good relationship between body length and the number of increments in the otoliths, suggesting that the number of increments were proportional to the age of the fish. Variability between size-at-increment number reflected variability in growth. We are presently examining the source of this variability to determine if it is influenced by environmental factors. Juvenile spotted seatrout, which were used to calculate hatchdate distributions, were most abundant in September 1995 collections. August was the dominant birthrate month, suggesting either intensive spawning during that period and/or favorable conditions for cohort survival. Data from ichthyoplankton collections will be used to determine the intensity of spawning by month during the sampling period from September 1994 through August 1995 and from April 1996 through September 1996. A preliminary analysis of the distribution of hatchdates of juvenile trout collected from May 1995 through September 1995 indicated that certain monthly cohorts (April and June 1995) were absent in our collections at the end of the sampling period (September 1995). This indicates a movement out of the sampling area or an indication of weak monthly cohorts. Analysis of otoliths from larvae that have yet to be processed along with the inclusion of more juvenile fishes should allow us to relate the survival of cohorts to environmental factors. When all ichthyoplankton samples are processed we will analyze our spotted seatrout data within the context of temperature and salinity. Available information suggests that strong salinity gradients occur in Florida Bay and salinity may influence spotted seatrout spawning and along with temperature may also influence hatching success. Existing strong salinity gradients may also influence total ichthyoplankton composition.

Gordon W. Thayer and Donald E. Hoss, NOAA, National Marine Fisheries Service, Beaufort, NC; Michael B. Robblee, U. S. Geological Survey, BRD, Miami, FL; and Richard E. Matheson and David K. Camp, FDEP, Florida Marine Research Institute, St. Petersburg, FL.

In the fall of 1987, a widespread, rapid die-off of turtle grass, Thalassia testudinum, began in Florida Bay. Halodule wrightii and Syringodium filiforme habitats appear to have been affected as well. Rapid, but often ephemeral recolonization of many of the die-off areas has occurred often resulting in patchy bottom habitats. It has been hypothesized that these changes coupled with other environmental modifications could have effects on the nursery function and fisheries of the Bay. To address these concerns, staff from the USGS, NMFS and Florida DEP have been conducting research to address the question: What are the changes in the distribution and abundance of living resources that have occurred as a result of habitat changes in Florida Bay? This is being done through decadal comparison of fauna (1994-present vs 1984-1987): Johnson Key Basin bank, basin and near-key habitats (Robblee); banktops representing five different vegetational subenvironments in the Bay (Matheson and Camp); and basins of the Bay southwest of a line from Little Madeira Bay to Tavernier Creek (Thayer and Hoss).

Sampling reported was conducted during January and May 1995 using the same sampling stations and protocol used a decade ago in Johnson Key Basin, one of the earliest areas affected by the die-off. Quantitative 1 m2 throw trap samples were taken and fauna cleared using sweep nets;quantitative measures of seagrass standing crop biomass (above and belowground) and composition also are taken. Analyses are demonstrating that there have been changes in composition and dominance of both plant and faunal communities. Aboveground biomass of Thalassia has decreased 72% relative to 1985; Syringodium has disappeared totally; and there has been no change in biomass of Halodule. Therefore the basin seagrass community has changed from one dominated by Thalassia to one co-dominated by Thalassia and Halodule. Caridean shrimp densities have decreased from about 160 m-2 to ~35 m-2 in 1995. Pink shrimp density in January decreased from a mean of 7 to 3 m-2 over the decade, whereas no difference was observed in May sampling (both ~2 m-2). Fishes also decreased over the decade from averaging about 11 to 4 m-2. Thor floridanus (down 93%) and Periclimenes longicaudatus declined radically among the caridean shrimp, while Lucania parva (down 97%), Floridichthys carpio, and Opsanus beta showed precipitous decreases among the fishes. Other species of carideans and fishes, however, increased in density: e.g., Alpheus heterochaelis, Anchoa mitchilli, Gobiosoma robustum.

We designed the banktop throw-trap sampling program as a comparison with the studies conducted by the National Audubon Society (Powell, Sogard, and Holmquist) during 1984-86. The sampling gear, sites, and methodology of the previous study were used to investigate the effects of environmental changes since 1986 on the resident fishes and crustaceans of these shallow seagrass beds. During 1994-96 we sampled the lee side of banks at six sites in five different subenvironments of the Bay: the Atlantic near the Buchanan Keys, the East Central on Cross Bank, the Northeast near Eagle Key, the Interior near Coon Key and near the Dump Keys, and the Gulf near Oyster Keys. Additionally, in 1995-96 we sampled vegetated and die-off sites on another bank in the Gulf subenvironment: near Rabbit Key. Among the six original bank sites of Powell et al., our data showed that only the Dump Keys site experienced significant seagrass loss. Some faunal changes occurred over the decade at several of the sites, but the most draatic changes occurred at Dump Keys. This bank was productive for resident fauna (relative to other banks) during both studies, but much of the production switched from canopy to benthic species. Thor floridanus and Lucania parva experienced precipitous declines at Dump Keys, whereas Alpheus spp., Gobiosoma robustum, and Microgobius gulosus experiences significant increases. Among the other sites, faunal diversity patterns remained constant (highest diversity at sites with strong marine influence), and abundance of both individual species and ecological groups (canopy vs benthic species) experienced relatively minor fluctuations (especially compared to those at Dump Keys). At Rabbit Key we compared areas with apparently recent seagrass die-off with nearby areas of dense Thalassia. In the die-off areas, benthic and canopy fish declined by factors of approximately 3 and 2, respectively. Canopy shrimp (primarily Thor) also declined by a factor of more than 40, but benthic crustaceans (primarily hermit crabs) increased by a factor of approximately 4. The pink shrimp, Penaeus duorarum, was found at similar densities at both sites, but their average size was greater at the vegetated site. We speculate that the changes observed at our Dump Keys site indicate that this area will be relatively productive with or without seagrass, but that the predator which forages on this bank must be flexible (i.e., not a specialist on canopy dwellers) in order to avoid a serious decline in food resources. At the other five Powell et al. sites, the phytoplankton blooms and episodes of increased turbidity that have plagued the Bay seem to have caused few if any changes in banktop seagrass beds or their fauna. At Rabbit Key sites, recent die-off areas have lost much of their faunal production and have probably declined in value as foraging habitat for all species except those that specialize in benthic crustaceans.

Our basin-wide study is using a stratified random sampling design and sampling protocol established in 1984. The sampling area is subdivided into Atlantic, Central and Gulf strata based on the general distribution of benthic vegetation plus an additional stratum comprised of channels. Fauna are collected at the approximate center of randomly selected cells within each stratum using a 3.4 m otter trawl pulled between 2 boats for 2 minutes at a speed of 3.5-4.5 knots. At this point, we have pooled data from within each of the strata. Pooled bottom habitat data show a decrease in all species within each stratum, with Thalassia having about 65% lower average short shoot densities than during 1984-1985. Decreases were observed for both Halodule, which occurred in each of the strata, and Syringodium which occurs only in the Gulf and Channel strata. Many stations, however, have or are showing evidence of recolonization. Channel habitats as a whole, which generally seem to be ignored as an important habitat, are displaying the most persistent loss of all three of the seagrasses (e.g., 1984-85 mean short shoots m-2 values for Thalassia, Halodule, and Syringodum of about 800, 1100, and 150, respectively compared to respective values of 150, 300, and 30 currently).

Pooled data for fishes indicate a decrease in the mean density for the study area and a shift in composition. During the 1984-85 study we sampled a total of just over 155,000 m2 collecting almost 27,000 fish for an average of 170 per 1000 m-2. A decade later we have sampled in excess of 174,000 m2 of Bay bottom collecting almost 25,500 fish with a mean density of 147 fish per 1000 m-2. These densities are considerably lower than those observed using throw traps, but also cover areas of the Bay, e.g., areas of both the Atlantic and Central strata, that have known low densities of both vegetated bottom and fishes. Only the Channel stratum is exhibiting a decrease in average fish density. Whereas the communities we sampled a decade ago were dominated by Lucania parva, Floridichthys carpio, Lagodon rhomboides, and Eucinostomus spp. (all of which have decreased in abundance since 1984-85), Anchoa mitchilli represents in excess of 60% of the fishes we have collected since 1994. Several species show little change in mean abundance between decadal periods (e.g., Bairdiella chrysoura, Opsanus beta) and several have increased in mean density (e.g., Monacanthus ciliatus, Gobiosoma robustum, Harengula jaquana). Many of the species exhibiting an increased average abundance, particularly the bay anchovy, are planktivores as opposed to benthic feeders, suggesting a shift in trophic patterns of the fishery communities we sampled.

Our faunal-related sampling since late 1994 has been designed as comparisons to work carried out in the same habitats, at the same stations, and with the same gear a decade previous. Many of the sites have displayed no major changes in the seagrass community over the decade, whereas areas of Johnson Key Basin, several banktops, and many of the other basins are displaying both decreases in Thalassia and evidences of seagrass recovery, particularly through colonization by Halodule. The resident and transient fish and invertebrate communities of those areas unimpacted by the seagrass die-off appear not to have changed much over the 10 year period. However, where seagrasses have declined there have been both decreases in total abundance and shifts in faunal composition. In general canopy dwelling fishes and invertebrates and some benthic fishes have declined; benthic crustaceans and at least at least one benthic and a pelagic fish component have become important. Thus, where there are changes in the seagrasses, the resident fishes now appear dominated by a few benthic as opposed to canopy feeders, and the transient fishes (collected by both throw traps and bottom otter trawls) are overwhelmingly dominated by the plankton feeding bay anchovy.

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