To address these issues, this report overviews and synthesizes the pertinent literature and data to determine the extent and current status of key resources in the Dry Tortugas region relevant to the condition of the broader fish communities and fisheries of the Florida Keys. The construct of the report was directed to address and answer questions that may arise in the design (location and size) of a marine "no take" reserve in the Dry Tortugas region.

An exhaustive literature survey and data assimilation was conducted to characterize the temporal and spatial distribution and productivity of essential habitats and fisheries in the Dry Tortugas region. The study involved a major collation, synthesis and analysis of a suite of databases from the Dry Tortugas region on biological, physical, chemical and geological environments. We employed a broad array of information technology databases, SAS statistical software, and scientific data visualization technologies to organize and visualize the information relevant to the assessment of the multispecies reef fish community. To facilitate deliberations amongst the National Park Service, the Florida Keys National Marine Sanctuary, and the Marine Reserves Advisory Panel on the design of a marine protected area in the Dry Tortugas region, we produced a digital CD-ROM of all these databases suitable for display on the ARC/VIEW GIS software system.

We characterized habitats critical to sustainable fisheries such as coral reef resources for adults and key inshore environments for juveniles. We found that living coral reef habitats appear to extend down to about 30 m on the Dry Tortugas and eastward to the Marquesas. A majority of the habitats in this geographic area have not been identified either through diver interpretation or by assessment of aerial photographs. In this area it is probably reasonable to assume that areas with depths less than 30 m contain a matrix of high and low relief coral reefs, hard-bottoms and sand proximal to the reefs.

Fisheries-dependent data for commercial, recreational and headboat fishery sectors for the Dry Tortugas region were quantified from survey and assessment data provided by the National Marine Fisheries Service, the National Park Service and the State of Florida. In general, these data seem to indicate significant declines in catches in the region over the last decade, the period for which data were available.

We also analyzed an extensive quantitative fishery-independent database derived from reef fish visual surveys conducted by SCUBA divers from 1979 to 1998, and used these data to develop estimates of population abundance, assemblage composition, and stock structures in relation to key physical and habitat factors. We produced maps of bathymetry overlain with habitats was the base to view animal density for juveniles and average size in the exploitable phase of the stock for adults which reflects stock recruitment success and yield potential, respectively. Spatial estimates of average size in the exploitable phase, a measure of the stock’s production relative to maximum sustainable yield, plus spatial estimates of pre-recruit density which measures the stability and resiliency of a stock. In general, the stocks are distributed in a spatially-heterogeneous way and the abundance, average size, and biomass seem to correlate strongly with key features of the environment.

From these data we conducted baseline multispecies stock assessments for more than 35 economically and ecologically important coral reef fish stocks and spiny lobsters in the Florida Keys using a systems approach that integrates sampling, statistics, and mathematical modeling. We compared the situation in the Dry Tortugas relative to the status of the stocks throughout the broader Florida Keys ecosystem. Exploitation effects were assessed using a new length-based algorithm that calculates total mortality rates from estimates of "average length of fish in the exploitable phase of the stock". These estimates were highly correlated for two statistically independent data sources on reef fish: fishery-independent diver observations and fishery-dependent headboat catches. We used the Reef-fish Equilibrium Exploitation Fishery Simulation (REEFS) model (Ault et al. 1998) and estimates of fishing mortality to assess yield-per-recruit relative to fishing intensity and gear selectivity, and spawning potential ratio (SPR) relative to U.S. federal "overfishing" standards.

Our analyses show that 13 of 16 groupers (Epinephilinae), 7 of 13 snappers (Lutjanidae), one wrasse (Labridae), and 2 of 5 grunts (Haemulidae) are below the 30% SPR overfishing minimum. Some stocks appear to have been chronically overfished since the late 1970's. The Florida Keys reef fishery exhibits classic "serial overfishing" in which the largest, most desirable and vulnerable species are depleted by fishing. Rapid growth of the barracuda population (Sphyraenidae) during the same period suggests that fishing has contributed to substantial changes in community structure and dynamics.

We used black grouper as an example of the effects of overfishing on the fisheries resources of the Dry Tortugas region and the Florida Keys. The net conclusion of these analyses relevant to fishermen is that the average size of black grouper caught in 1999 is 40% of what its historical level (i.e., average of 22.5 lbs circa 1930 versus 9 lbs today). In terms of the stability and resiliency of the black grouper population, the spawning stock biomass is estimated to now be at 5% of what it once was. The current rate of fishing mortality on the black grouper stock is now greater than 4 times the level that would be expected to produce maximum sustainable yield. This situation is similar for a broad segment of the economically and ecologically important reef fish stocks in the Florida Keys.

As a result, at least 23 fish stocks in the Dry Tortugas region may qualify for 'species at risk’ designation due to the serious nature of overfishing in the Florida Keys and the extent to which critical habitats are being degraded or destroyed throughout the range of the unit stocks. More broadly, the region is home to at least 62 fishes that are currently protected under Federal/State regulations, a number of which that are currently under evaluation for increased protection by IUCN and AFS conservation organizations due to critically small population sizes. Characteristically, we noted that the Dry Tortugas region seems to have more and larger fish of the key species (i.e., groupers, snappers, hogfish, grunts, lobsters, etc.) than the rest of the surrounding Florida Keys waters. To this end, we believe that the Dry Tortugas region has served a de facto marine reserve for at least the past 20 years.

Prudent development of a design for marine protected areas in the Dry Tortugas region should carefully consider co-management of fishing mortality and critical habitats. That is, some attention must be given to methods to reduce the areal effect of fishing mortality. In addition, population-dynamic analyses of growth and mortality for these stocks must be undertaken to mitigate the effects of growth overfishing. Finally, the connections between adult stocks, fishing mortality, physical forcing, and critical habitats must also be considered to minimize the potential for recruitment overfishing and to protect those habitats critical to the ontogenetic migrations of reef fish that provide a 'source’ of recruits from adult spawning and the 'sink’ of recruits after their larval life stage. This strategy is required to therefore maximize the survivorship of new recruits to the stocks.

In our analyses of fisheries-dependent, fisheries-independent and habitat databases we found that substantial areas of the Dry Tortugas region are either under-sampled or with no sampling at all. Thus, we suggest a strategic approach focused on development of a systematic, multi-scale methodology to assess the state of the coral reef fish community and the extent of biological-physical coupling as it relates to "habitat" for reef fishes.

To better assess these issues in biophysical coupling and recruitment and stock variability, what is needed are hydrodynamic circulation models for the region to facilitate the study of larval drift from the many spawning aggregation sites identified in the report for the Dry Tortugas region. The coupling of fisheries models to the biophysical environment would allow coherent linkages between adult spawning stock to juvenile recruitment to better understand the bay to reef interconnections of the reef fish life cycle and to optimize the determination of the location and size of proposed marine protected areas.

Finally, we believe what needs to be established is an integrated system of monitoring, assessment and management to meet the goals of building sustainable fisheries and conserving marine biodiversity in the Dry Tortugas region for many generations to come.

Fisheries and Essential Habitats

Thomas W. Schmidt, Jerald S. Ault and James A. Bohnsack

Executive Summary2

Table of Contents5

Figure and Table Legends6

2.0The Biophysical Environment of the Dry Tortugas Region...16

3.0Fisheries-Dependent Information36

4.0Fisheries-Independent Information...53

5.0Acknowledgments102

6.0References.103

Figure Legends

Figure 1.- Three dimensional bathymetry of the northern Caribbean Sea looking from Honduras. Note the precipitous decline of the depths in relatively short distance from land and the intricate channeling which creates some of the fastest ocean current speeds in the world.

Figure 2.- The 3-dimensional seascape of south Florida extending from Miami to the Dry Tortugas. Green indicates land while red indicates the coral reef tract.

Figure 3.- Two dimensional representation of bathymetry for the Dry Tortugas region showing Ft. Jefferson, Loggerhead Key, the Tortugas Bank and Riley’s Hump. Outer white line is the FKNMS boundary and the inner white line is the boundary for the Dry Tortugas National Park.

Figure 4.- Composite habitat map showing habitats in DTNP (via FDEP/DERM aerial photo interpretation), and habitats outside DTNP (via side-scan sonar from NOAA R/V Ferrel cruises). Areas of red and brown (high relief) indicate coral reefs, while green areas are seagrass, and yellow are sand bottoms. Numbered black circles are fish visual sampling sites and red triangles are lobster sampling sites.

Figure 5.- Overlay of habitats on 3-dimensional bathymetry for the DTNP and adjacent waters.

Figure 6.- Composite diagram showing habitats inside the DTNP and areas of raw side-scan sonar data shown to the northwest (large gray banded area) of DTNP.

Figure 7.- Habitat type assignments to the digital side-scan sonar data to the northwest of DTNP shown in Figure 6 classified for hardbottom, high relief reef, low relief reef, pinnacle reef, and sand bottom habitats.

Figure 8.- Side-scan sonar data from the large area to the northwest of the Park relative to FKNMS boundaries (blue lines) classified for hardbottom, high relief reef, low relief reef, pinnacle reef, and sand bottom habitat types. White rectangle in upper left of image is the area of detail shown in Figure 9.

Figure 9.- Zoom-in of the side-scan sonar data shown in the white rectangle in Figure 8 showing details of the raw images.

Figure 10.- Assignment of habitat classifications from the raw data of Figure 7.

Figure 11.- Two-dimensional representation of bathymetry for the Florida Keys and west Florida shelf areas proximal to the Dry Tortugas region.

Figure 12.- Location map showing NMFS statistical sampling grids for the State of Florida used for the collection of commercial fishery catch and effort information.

Figure 13.- Mean carapace length (mm) for spiny lobster sampled from Dry Tortugas commercial fishery catches by National Marine Fisheries Service personnel for the 1985-86 and 1994-95 fishing seasons. The lobster were captured from tow areas. The 95% confidence intervals for each season are indicated (from Harper 1995, Figure 10).

Figure 14.- Length-frequency histogram for spiny lobster (males, females and both sexes combined) from the area northwest of the Dry Tortugas during the 1990-91 and 1991-92 commercial fishing seasons. Lobsters were sampled from six Trip Interview Program (TIP) commercial fishing trips for each season (from Harper 1992, Figure 12).

Figure 15.- Dry Tortugas region 1986-1996 commercial landings (lbs) for: (a) pink shrimp and spiny lobster; (b) yellowtail snapper, red grouper and black grouper; and, (c) mutton snapper, gray snapper and hogfish.

Figure 16.- Combined snapper and grouper species catch-per-unit-effort (number of fish per hour fishing) for the recreational fishing fleet operating in DTNP for the period 1981-1984 (data from National Park Service).

Figure 17.- Combined snapper, grouper and all reef fish species catch-per-unit-effort (number of fish per angler-day) for the headboat fishing fleet operating in the Dry Tortugas region for the period 1981-1997 (data from NMFS Headboat Survey).

Figure 18.- Conceptual diagram of the Florida Keys "coastal bay to coral reefs" interconnections. Adult biomass of the multispecies reef fish community is concentrated on the thin coral reef tract line. Mature adults spawn at the deep-reef edge. Larvae are advected shoreward into coastal bays like Florida Bay and Biscayne Bay. As juveniles mature, they migrate seaward over a mosaic of essential habitats to the coral reefs (Ault and Luo 1998)

Figure 19.- Map of the Florida Keys coastal marine ecosystem running from Key Biscayne southwest to the Dry Tortugas over bathymetry. The numbered circles show the 102 sites where more than 6500 visual survey samples of the reef fish community were taken from 1979 to 1998 using the Bohnsack point census method.

Figure 20.- Map of the Dry Tortugas region marine ecosystem with DTNP and FKNMS boundaries showing the 24 sites where visual census of the fish community were collected using the Bohnsack and Kimmel methods (Numbered orange circles). The red triangles indicate the 66 locations of lobster sampling by Bertelson (1998). Between 1994 the Bohnsack method was used at 24 sites and collected 644 samples (Ault et al. 1998, Bohnsack and McCllelan 1998). Between 1990 and 1995 the Kimmel method was used at 6 sites and 437 samples were taken (Rydene and Kimmel 1998).

Figure 21.- Locations of the 24 reef fish visual sampling sites (denoted by yellow and red flags) in the DTRO overlain on a 3-dimensional bathymetry map. The red flag on the northwest slope of the Tortugas Bank indicates the location of the Sherwood Forest site.

Figure 22.- Conceptual diagram and simulations of the equilibrium relationship of average length of fish in the exploitable phase of the stock dependent on fishing mortality rate using the REEFS model (Ault et al. 1998) for two reef fish taxa, grouper and snapper.

Figure 23.- Estimates of percent spawning potential ratio (SPR%) for 35 species of Florida Keys reef fish comprised of groupers, snappers, grunts, hogfish, and great barracuda (from Ault et al. 1998, Figure 7). Darkened bars indicate stock "overfishing", and open bars indicate the stock is above the 30% SPR U.S. Federal standard.

Figure 24.- The spatial distribution and relative abundance of the Florida Keys black grouper (Mycteroperca bonaci) population: (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; (b) average size of fish in the exploited phase for the entire Keys; and (c) average size of fish in the exploited phase for the Dry Tortugas region.

Figure 25.- Black grouper (Mycteroperca bonaci) adult (exploited) and juvenile (pre-exploited): (a) relative abundance; (b) average size in the life phase; (c) presence-absence; and, (d) CPUE by four north to south geographic regions (Biscayne National Park, north, middle, and south [Dry Tortugas region])of the Florida Keys.

Figure 26.- Overfishing analyses for the Black Grouper following the mandate of the 1998 Reauthorization of the Magnusson-Stevens Fishery Management Conservation Act. The present level of estimated fishing mortality on the Florida Keys black grouper stock is greater than 4 times that level associated with Maximum Sustainable Yield and well-below the Federal minimum SPR level established for overfishing. The fishery is grossly overfished!

Figure 27.- The spatial distribution and relative abundance of the Florida Keys red grouper (Epinephelus morio) population: (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; and, (b) average size of fish in the exploited phase for the entire Keys.

Figure 28.- Red grouper (Epinephelus morio) adult (exploited) and juvenile (pre-exploited): (a) relative abundance; (b) average size in the life phase; (c) presence-absence; and, (d) CPUE by four north to south geographic regions (Biscayne National Park, north, middle, and south [Dry Tortugas region])of the Florida Keys.

Figure 29.- The spatial distribution and relative abundance of all "large" Florida Keys grouper species (i.e., black, gag, nassau, red, yellowedge and yellowfin) populations combined: (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; and, (b) density of fish in the exploited phase.

Figure 30.- The spatial distribution and relative abundance of the Florida Keys yellowtail snapper (Lutjanus chrysurus) population: (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; and, (b) average size of fish in the exploited phase for the entire Keys.

Figure 31.- Yellowtail snapper (Lutjanus chrysurus) adult (exploited) and juvenile (pre-exploited): (a) relative abundance; (b) average size in the life phase; (c) presence-absence; and, (d) CPUE by four north to south geographic regions (Biscayne National Park, north, middle, and south [Dry Tortugas region])of the Florida Keys.

Figure 32.- The spatial distribution and relative abundance of the Florida Keys gray snapper (Lutjanus griseus) population: (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; and, (b) average size of fish in the exploited phase for the entire Keys.

Figure 33.- Gray snapper (Lutjanus griseus) adult (exploited) and juvenile (pre-exploited): (a) relative abundance; (b) average size in the life phase; (c) presence-absence; and, (d) CPUE by four north to south geographic regions (Biscayne National Park, north, middle, and south [Dry Tortugas region])of the Florida Keys.

Figure 34.- The spatial distribution and relative abundance of the Florida Keys gray snapper (Lutjanus analis) population: (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; and, (b) average size of fish in the exploited phase for the entire Keys.

Figure 35.- White grunt (Haemulon plumeri) adult (exploited) and juvenile (pre-exploited): (a) relative abundance; (b) average size in the life phase; (c) presence-absence; and, (d) CPUE by four north to south geographic regions (Biscayne National Park, north, middle, and south [Dry Tortugas region])of the Florida Keys.

Figure 36.- Blue-striped grunt (Haemulon sciurus) adult (exploited) and juvenile (pre-exploited): (a) relative abundance; (b) average size in the life phase; (c) presence-absence; and, (d) CPUE by four north to south geographic regions (Biscayne National Park, north, middle, and south [Dry Tortugas region])of the Florida Keys.

Figure 37.- The spatial distribution and relative abundance of the Florida Keys spiny lobster population in the Dry Tortugas region based on sampling by Bertelsen (FMRI Marathon): (a) density (numbers seen per Bohnsack-method visual sample) for small pre-exploited phase fish; and, (b) average size of fish in the exploited phase for the entire Keys.

Figure 38.- Summary of snapper species (Lutjanidae) spawning aggregation sites near the Dry Tortugas region extending to Key West in the southern Florida Keys coral reef tract identified by commercial fisherman (Lindeman et al. 1999; D. DeMaria, personal communication). Tortugas Bank is 12, Riley’s Hump is 16/7, NW DTNP is 11, Rebecca Shoals is 13, Cosgrove Reef is 9/2, Vestal, Western Dry Rocks and Eyeglass Bar is 5.

Figure 39.- Hydrodynamic simulations of currents in the northern half of the Florida Keys reef tract extending from north of Miami to south of Marathon showing the seasonal reversal of flow along Hawk channel (courtesy of J. Wang, Univ. Miami RSMAS).

Table Legends

Table 1.- Fisheries and fishery ecology studies in the Dry Tortugas region.

Table 2.- Reported commercial fishery annual landings for the period 1979-1996 from the Dry Tortugas region (NMFS Statistical Grid 2). summarizes the available commercial landings and CPUE information from the Tortugas region. Commercial fisheries are prohibited within the boundaries of DRTO. Commercial landings data from the Tortugas region are reported within statistical sampling grids/areas for the Gulf of Mexico (Fig. 15). Because the grid cell area covers 1 degree on a side, assessment of catches can only be gleaned for a relatively large spatial area. In most cases, the actual catch location for the landings data has not been reported.

Table 3.- Number of fishing trips, average fishing party size, average hours fishing per trip and total fish captured for the DTRO recreational fishery for 1981-1984 (National Park Service data).

Table 4.- Number of recreational fishing trip interviews by fishing area in the DTRO for 1981-1984 (National Park Service data).

Table 5.- Total headboat angler-days for the Florida Keys (statistical grids 744, 748, and 1) and Dry Tortugas (statistical grid 2) for the period 1981-1997 estimated from the NMFS Headboat Survey (Dixon and Huntsman 1982; Bohnsack et al. 1994; Ault et al. 1998).

Table 6.- Estimated 1981-1997 annual landings (numbers of fish) for the headboat fishery operating in the Dry Tortugas region. Headboats are chartered vessels that carry 15 or more passengers per trip.

Table 7.- Region, reef number, reef name and samples taken per site during the period 1994-1999 for the Bohnsack-Ault visual Survey (Ault et al. 1998) for the Florida Keys.

Table 8.- Summary of reef fish fisheries-independent surveys of 1081 samples taken in the Dry Tortugas National Park and surrounding waters using visual census methods: (a) Bohnsack-method (n=644), and (b) Kimmel-method (n=437)for the period 1990-1998.

Table 9.- Draft list of tropical marine fish stocks at risk in the Dry Tortugas region.

INTRODUCTION

The Florida Keys support a rich tropical marine ecosystem, a productive multispecies coral reef fishery, and a multi-billion dollar tourist economy. However, the Florida Keys are considered an "ecosystem-at-risk" as one of the nation's most significant yet most stressed marine resources under management of NOAA and the National Park Service. The multispecies reef fisheries of the Florida Keys are under siege from fleet expansions and increased vessel fishing power that threaten to over-exploit, destroy habitat, change marine environments, and reduce biodiversity (Bohnsack and Ault, 1996). Reef fisheries can target a number of economically and ecologically important species (e.g., groupers, snappers, lobsters, conch, sponges and corals). Over the past several decades, public use and conflicts over fishery resources have increased sharply, while some fishery catches from the historically productive snapper and grouper stocks have declined (Bohnsack et al., 1994; Ault et al. 1997, 1998).

The Dry Tortugas (DRTO) region is a critical part of the larger Florida Keys marine ecosystem, and the associated fish and fisheries of the Tortugas Region play an important ecological and fishery support role in both the Dry Tortugas National Park (DTNP) and the broader Florida Keys National Marine Sanctuary (FKNMS). The rich marine fish community found in the DRTO region are represented mostly by sub-tropical and tropical species which support a wide array of biologically, ecologically and socio-economically important commercial and recreational fisheries. Chiappone and Sluka (1996) provide an overview of the Florida Keys region biological and fishery characteristics and issues regarding spiny lobster and pink shrimp and give some perspective on the life history, distribution and occurrence of these species by habitat in the Tortugas region. The high diversity and sheer number of different fish and fisheries in this region is extraordinary. Until recently, the condition or quantitative status of most reef fish stocks was largely unknown because of the great number of species in the fishery and comprising the marine fish community, a lack of fishery-dependent effort and landings data, and the quantity of population dynamics data needed to do traditional stock assessments. In an innovative use of two decades of fisheries-independent and fisheries-dependent data, Ault et al. (1997, 1998) conducted a quantitative retrospective assessment of the multispecies reef fish community in the Florida Keys and showed that fishing mortality levels are very intense, that many stocks are "overfished", and further, exploitation has altered the structure and dynamics of the reef fish community. The importance of the DRTO region is seen in its critical upstream capacity for providing inputs to the rest of the Keys and the documented migrations of many groupers, snappers, pink shrimp and spiny lobsters to the region during spawning season each year. These analyses suggest that the Dry Tortugas region, due to its relatively great distance from ports and attendant low levels of fishing effort, has de facto supported the broader Florida Keys fishery with larval supply and export of adult biomass for over two decades. The picture of exploitation potential is the DTRO is rapidly changing with technological innovations like GPS, better and faster vessel design, and growing size of the fleets and fishermen.

Another concern, perhaps as serious as those from fishing, is the restoration of the Everglades north of the Florida Keys where hydrological projects of historic proportions are expected to substantially change the timing, volume, and location of freshwater outflows into the coastal marine environment (Harwell et al., 1996; ACOE Restudy 1999). Many reef fishes use South Florida coastal bays as juvenile nursery areas for the early part of their life history, before moving out to the reef habitats as adults. These changes could affect the survivorship of juvenile reef fishes in critical shallow nursery areas of Florida Bay and Biscayne Bay, ultimately affecting the productivity of the coral reef ecosystem (Serafy et al., 1997), including the Dry Tortugas region. Furthermore, pressures for increased fishery production escalate as human populations swell in the region. Many species observations made over the past 100 years suggest that many well-known individual species of the Tortugas region are now potentially rare, threatened, or endangered. Some benthic community studies conducted within what are now boundaries of DRTO provided useful information on marine habitats dating back to the 1880's, yet only limited information is available to present on the spatial/temporal distribution and abundance, size distribution, and trophic structure of fishes among habitats in the Tortugas region and adjacent marine environments. Concomitantly, the status and governing biological dynamics of the reef fishery resources are not well understood, and important stock assessment data are not well understood, or in fact, even available.

In response to these types of problems, the National Research Council (1994) recommended development of ecosystem management programs for building sustainable fisheries. Concern about habitat degradation and escalating resource uses from rapid human population growth in southern Florida resulted in the establishment of the Florida Keys National Marine Sanctuary (FKNMS) in 1990. Coral reefs are a particular concern as one of the most complex ecosystems on earth. The diverse reef fish community is influenced by complicated biological and physical interactions. A spatial network of "no take" marine reserves in the Dry Tortugas portion of the FKNMS have been proposed. These, combined with traditional management measures, have the potential to reverse these trends for many species and to allow the long-term goals of building sustainable fisheries and protecting biodiversity to be achieved. An initial 9 mi² reserve area has been established in Sambos area off Key West. A second large "no take" marine ecological reserve in the FKNMS is proposed for implementation in the Dry Tortugas by the year 2000 and working groups will begin drafting boundary alternatives for the Tortugas Ecological Reserve in early 1999. But to fully realize the benefits of spatial management strategies like marine protected areas, it is important to expand the fishery management concept beyond a stock assessment focus to other issues of interest such as the creation of "no-take" marine reserves in the Florida Keys National Marine Sanctuary, which have the potential to build sustainable fisheries and protect marine biodiversity. However, unambiguous assessment of the efficacy of marine reserves will require an improved quantitative framework for integrating data, analyses and models in support of scientific decision making. There is a general paucity of relevant scientific information and little historical data to compare with the removal of extractive activities data in an effort to optimize spatial management strategies for design of marine protected areas.

Thus, the goal of this paper is to organize, analyze and present relevant biological and fisheries data in quantitative spatial information in visual formats to characterize the status of knowledge on fishery and habitat resources to contribute to the prudent development of an 'ecological reserve’ in the Dry Tortugas region. These data are then used to summarize the extent of existing information on life history, ecology, species distribution and abundance, essential habitats, and status of the stocks for the multispecies reef fish fishery. In this report we focus primarily on important reef fishes and macroinvertebrates in the region like groupers, snappers, spiny lobster and pink shrimp that have substantial economic and ecological value to the dynamics, productivity and functioning of the marine environment in the Dry Tortugas region. The life history section will primarily focus on the species composition, abundance, and distribution of these species among various study sites within coral reef and hard-bottom habitats. These deep and shallow reef formations not only support basically different fish faunas but they also support different life history stages (larval, juvenile and adults) which display different distribution patterns among habitats. These fish communities will be discussed regarding variations found within the boundaries of DRTO and contrasted with fish communities located just outside of park boundaries in FKNMS.

To accomplish our goal, this research has objectives that span acquisition, synthesis, and analyses of several classes of biological information relevant to fish and fisheries. Data assimilated, analyzed and presented in this volume are taken from both published literature and gray-technical reports, and analyses of fisheries-dependent and fisheries-independent spatially-explicit databases from programs conducted in the region (cf., Appendix 2). We used firm linkages between fishery management systems and scientific data visualization are fundamental to achieving this objective. The most familiar role of visualization is presentation of results that let us "see" the bigger picture of biology in a physical setting. In this paper we take a systems approach to fishery management and illustrate the role scientific data visualization can play in helping us to define linkages between fisheries research, assessment, and management of the Florida Keys multispecies reef fish fishery. In this paper we cover the critical biophysical habitats, the fisheries-dependent and fisheries-independent monitoring programs, and provide a preliminary assessment of the status of the multispecies reef fishery resources in the DTRO. In doing so we endeavored to produce a study that will be critical to fill in holes in the distribution of resources and characterize habitats for nursery value and will provide additional scientific guidance necessary to make logical and concise decisions on the nature and extent of these boundaries.

THE BIOPHYSICAL ENVIRONMENT OF THE DRY TORTUGAS

?We blended a large number of physical, biological, chemical, geological and fisheries databases from the Dry Tortugas region made available to us from a variety of local, State, University and Federal sources. The data coverages included information on bathymetry, habitat types, water quality, fish communities; benthic macroinvertebrates (cf., Appendices 1 and 2). We created a unique data set from side-scan sonar data recently acquired by NOAA personnel for the Tortugas Bank area in the FKNMS to the northwest DTNP. We then blended all these data into a wide array of "digital layer" coverages suitable for scientific visualization and geographic information systems (GIS). We used the synthesized database with which to drive a visualization show of the various data layers using ARC VIEW geographic information system and distributed this information on a CD-ROM for use by FKNMS and the National Park Service.

The Environmental Setting

Florida and the Florida Keys are situated downstream of the Caribbean Sea where the tropical waters are transported northward to ultimately enter the Straits of Florida after either directly after passing through the Yucatan Channel and moving around the Loop Current in the Gulf of Mexico (Figure 1; Lee et al. 1999). The Florida Keys coral reef ecosystem is a unique tropical coastal marine environment stretching about 370 km from Key Biscayne southwest to the Dry Tortugas (Figure 2). This volume primarily focuses on the area of the Florida Keys National Marine Sanctuary that runs from west of the Marquesas Keys to western FKNMS boundaries on the shelf edge, an area of relatively open water area considerably influenced by the eastern Gulf of Mexico (Lee et al. 1999).

The Dry Tortugas National Park, is a 190 km² area near the western extent of the FKNMS located approximately 120 km west of Key West. The area contains seven small islands composed of coral reefs and sand in a elliptical, atoll-like formation that is approximately 27 km long and 12 km wide (Figures 3, 4 and 5). DTNP is not included in the FKNMS. Water depths within DTNP boundaries range from sea level to about 25 m.

Areas outside DTNP contain the Tortugas Bank which extends upwards to about 15 m to the west, and a series of shallow unexplored and undocumented (by scientists) Banks eastward to the Marquesas. To the west of DTNP is the Tortugas Ecological Reserve Study Area (TERSA). Within the FKNMS, the State of Florida and NOAA are responsible for administering and managing the sea bottom and associated resources in the FKNMS. Water depths range here from 5 to 40 m over deeper benthic habitats.

Compared to inshore waters of the Florida Keys, the Tortugas Region experiences very little variability in water temperature and salinity other than seasonal effects.. Longer water residence times are found in the Tortugas region relative to most other areas in the Florida Keys.. This has considerable implications for larval transport as discussed in Volume 7 in the FKNMS Site Characterization (Lott 1996; Lee et al. 1999). Differences exist in habitat distribution from within the Tortugas region and from east to northeast along the Florida Keys. Within DRTO a mixture of shallow-water seagrassses, deep- water bare sand hard bottoms with moderate shallow water patch and bank reef development prevail. Adjacent to TERSA which includes the Sherwood Forest area, one of the most luxurious coral reefs in the Caribbean and an excellent habitat for reef fishes of commercial and recreational fishing interests. Overall, the north and west areas of Tortugas Bank forms a mosaic of extensive, robust, deep-water reefs adjacent to low relief, hard-bottom and sand-covered areas (Bohnsack & McClellan 1998 and others).

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

New Essential Habitat Maps for the Dry Tortugas

Essential to the understanding of fish distribution and abundance in the Tortugas region are the prevalent oceanographic characteristics and key benthic habitat types of the Florida Keys. These issues have received some discussion in Tables 2 and 3 in Volume 6 of the Florida Keys Site Characterization by Chiappone & Sluka (1996), and in Lee et al.’s (1999) physical oceanography site characterization for the Tortugas region. In general, physical oceanographic processes in the Dry Tortugas region are largely dominated by oceanic currents (e.g. Florida Current), tides, and wind-driven currents (Schomer & Drew, 1982; Lee et al. 1999).

We expanded the basic habitat spatial coverage’s by combining the side-scan sonar data collected by NOAA personnel on the R/V Ferrel during 1998. About 350 km2 of area were surveyed for bottom habitats and depth by collecting data along either side of 100 wide bottom swaths. The data were recorded as two-dimensional horizontal space coordinates, with shadows representing the vertical dimension (Figure 6). The individual side-scan images were then classified into five habitat types: hard-bottom, sand-bottom, Low relief reef, high relief reef, and pinnacle reefs (Figures 7-10).

Essential Habitats and the Reef Fish Life Cycle

The Florida Keys coastal ecosystem is situated parallel to the Florida current and Florida Bay and encompasses many varied habitats including freshwater marshes, estuaries, lagoons, mangrove stands, coral islands, sea grass beds, and coral reefs. The marine and estuarine environments of south Florida and the Florida Keys consists of a crescent-shaped archipelago of some 1,200 islands and a complicated network of relatively small to large embayments or subtropical lagoons (e.g. Card Sound and Barnes Sound), and the eastern Gulf of Mexico, each of which is hydrodynamically linked and subject to influence the larger south Florida shelf and coral reef tract. The chain of islands is comprised of two carbonate formations the Miami oolite formation to the north, and the Key Largo formation in the lower Florida Keys, separates the shallow waters of Florida Bay from the Atlantic Ocean in the east, and partially separates the eastern Gulf of Mexico and the Atlantic Ocean farther west. Seaward of the Florida Keys is the inner (Hawk Channel) and outer shelf margins of the Florida Reef Tract which extend 360 km from Biscayne National Park to the Dry. There is a general pattern of decreasing environmental variability and water residence time from Florida Bay to the Straits of Florida. Differences in circulation, water residence time and physical-chemical characteristics of water in the Tortugas Region generally reflect variability in the Gulf of Mexico Loop Current and the Tortugas Gyre in the Tortugas area (Lee et al. 1999). The Florida Keys living coral reef tract lies approximately 7 to 9 km south and parallel to the archipelago is defined by relatively open circulation and a dominance of patch and bank reefs along with coarse, bare sand areas (after Vaughan 1914) (Figure 2). Generally, it looks like most (or all) of the living coral reef tract lies on the edge of the deep-water escarpment that runs the length of the Keys from the Dry Tortugas north of Miami. The depths to which these reefs extends appears to be limited to about a maximum of 30 m. This is well seen if one compares Figure 4 with Figure 3. If depth is used as a proxy for areas that may contain coral reefs, then a great deal of "potential coral reef tract area" would appear to lie east of the Tortugas to Key West (Figure 11). Much of this area has been lightly or not sampled for coral reefs or reef fishes, while at the same time this area is well-known by fishermen as quite productive for reef fish and lobsters, and contains a number of reef fish spawning sites.

Figure 18

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Florida Bay and adjacent coastal estuaries serve as nursery areas for spiny lobster and many juvenile fishes that occupy reefs as adults including barracuda, hogfish, pink shrimp, many grunts, and most snappers and groupers. Because of this unique life style, we use reef fishes as sensitive indicators of environmental stress and system changes, because during their ontogenetic migrations from coastal bays as juveniles to the coral reefs as adults, fishes integrate all levels of natural and anthropogenic stress found across the coastal ocean ecosystem (see Figure 18). This suggests that the large-scale restoration effort across South Florida which currently focuses on the Everglades and Florida Bay, could affect Biscayne Bay’s productivity, diversity, and nursery habitat role and thereby influence the sustainability of reef fish communities on the coral reefs in unforeseen ways. The linkages between individual population or entire communities of fishes and macroinvertebrates are established by assessing both coastal bay and coral reef "habitats".

History of Fisheries and Ecological Research in DTRO

Over two decades of fisheries-dependent commercial, recreational and headboat data collection programs in the Tortugas region are briefly reviewed regarding: (1) rules and regulations; (2) principal species or taxa targeted by fishermen; and, (3) gear-types employed (hook and line, spearfishing, wire fish trapping, commercial trawling and longlining.

Compared to other areas of the Florida Keys, previous studies of the coral reef fish assemblages of the Tortugas region have been more diverse in their scope and objectives. Table 1 presents a list of previous studies of fishes in the Tortugas Region based on habitat type and study objectives. Most studies have been concerned with inventories on the number of species and number of families of fishes in shallow-water patch reefs and offshore bank reefs within the Tortugas atoll. Seagrass beds, hard bottom areas, and bank reefs outside of park boundaries have been virtually ignored. However, considerable information has been collected on the life history aspects (habits, feeding, foods, behavior, mimicry, etc.) and the taxonomy, and morphological and cytological development of reef fish mostly during the nearly 40 year operation of Carnegie Institution of Washington's (herein referred to as CIW) Tortugas Marine Laboratory on Loggerhead Key . The following sections will describe and compare the results of previous studies and one on-going study with emphasis on species diversity and abundance patterns. Tortugas regional fish assemblages will also be compared to assemblages in other parts of the Florida Keys. It is important to note that data collected over at least five years are needed to detect changes in fish abundance or diversity (Bohnsack et al. 1992).

Figure 11

The following section address the ecological characteristics of those fish assemblages, found principally in the Tortugas Region, the reef fishes Other fish assemblages e. g. nearshore fishes and coastal pelagic species of the Florida Keys and adjacent waters are discussed in detail in and will only be addressed in this report as they pertain to the fisheries-dependent (recreational and commercial) information of the Tortugas area.

Studies reporting on the diversity of the coral reef fishes of the Tortugas Region were first made during the late 1890’s and early 1900’s. In one study of unknown duration, David Starr Jordan and Joseph Thompson (1905) provided a list of 218 species in 62 families collected by Thompson in the Tortugas Region while on duty as a U.S. Naval surgeon on Garden Key during the 1890s’. One long-term (>20 years) study has focused on reef fish assemblages of the region and was initiated during the onset of the CIW Tortugas Marine Laboratory. Longley and Hildebrand (1941) studied the fish fauna for 25 years and described 32 new species (Longley 1940) for the region. Long-term observations yielded 442 species of which 300 were considered associated with the reef structure, but not primary reef species as erroneously reported in Vol. 6.

Elsewhere in the Florida Keys, Starck (1968) conducted a 10-year investigation of fishes in the Alligator Reef area , offshore of Upper Matecumbe Key, and recorded approximately 517 species using destructive and non-destructive techniques. Of the 517 species, he determined that 389 were actually associated with the reef structure while the remaining were not. At Alligator reef, he divided the fish species associated with the reef structure into primary and secondary reef species and recorded 253 primary reef species and .136 secondary reef species. Primary reef species were defined as those fish that spend nearly their entire life history associated with the reef [e.g. wrasses (Labridae), parrotfishes (Scaridae), surgeonfishes (Acanthuridae), butterflyfishes (Chaetodontidae) and angelfishes (Pomacanthidae). Secondary species are normally associated with reef but spent part or majority of their time in other benthic communities.(e.g., seagrass beds, soft/bare bottom shoal areas) Secondary species include gobies (Gobiidae), blennies (Blennidae), cardinalfishes (Apogonidae), grunts (Haemulidae), groupers (Serranidae), snappers (Lutjanidae), and squirrelfishes (Holocentridae). .

Recently, during the summer of 1975, Jones and Thompson (1978) evaluated the species composition, species diversity, and relative abundance of fish at six natural reef sites and two shipwreck, artificial "reef" sites, divided equally between the Upper Florida Keys (Key Largo) and DRTO. They found significant differences between the number of species recorded in the fish communities at Key Largo (John Pennekamp State Park) and the Tortugas (see also Table 6, Chiappone and Sluka 1996). They observed a total of 146 species in the Upper Florida Keys and 134 species in Park waters. During the following summer, Schmidt observed, as part of the Thompson and Schmidt (1977) survey, a total of 134 species from 3 natural and one shipwreck site. Species richness was highest at Long Key Reef, followed by White Shoal and Loggerhead Key.

During the 1990-1994 period, Rydene and Kimmel (unpublished, undated report) observed at six Park sites, a total of 166 species from a total of 375 samples. They found significant differences in the number of species found among six sites in Park waters. Although there was a significant overall difference in species diversity (H= 38.8, 8df, p<0.0001) and evenness indices (H=55.8, 8df) among habitat types, no significant differences were found over the five year study period.. McKenna (1997) sampled four Park reef areas and observed a total of 74 species during September 1992. Species richness within a site ranged from 26 to 46 while Diversity (H’) at the most disturbed grounding site averaged 1.23 and at Pulaski Shoal averaged 2.60. Relatively undisturbed sites near the grounding site were characterized by high diversity (2.27) and intermediate species assemblages. Bohnsack and McCllelan (1998) reported a total of 162 species from 375 visual census samples collected during 1994-1997.

Table1.- Fisheries and fishery ecology studies of fishes in the Dry Tortugas Region.

Fish Biogeography and Species Diversity

The geographic description of fishes varies over time. Each fish species is partly a product of regional oceanography, coastal geomorphology, habitat availability, and natural disturbance.

The demersal fishes of the Tortugas region can be classified into four basic types based on habitat descriptions and species distribution as discussed by Longhurst and Pauly (1987). The four categories are: (1) sciaenid assemblages (drums, croakers, groupers), (2) lutjanid assemblages (snappers), (3) active, large-eyed species adapted to clear water/high illumination (grunts, mojarra), and (4) highly evolved genera specific to reefs (e.g. triggerfishes, boxfishes, pufferfishes). The sciaenid assemblages occur in the warm temperate turbid waters to tropical areas in the western Atlantic Although the tropical Sciaenid assemblages have not been reported in Florida, the subtropical sciaenid assemblages does occur in the Florida/Tortugas area and is represented by families/species from the northern Gulf of Mexico to Cape Hatteras (Longhurst and Pauly, 1987) including Sciaenidae (drums/croakers), Serranidae (groupers), Clupeidae (herrings), Mullidae (goatfishes), and Gerreidae (mojarra). The lutjanid assemblage inhabits rock, coral, and coral sand habitats from Florida to Brazil and includes species from the families Lutjanidae (snapper), Serranidae (grouper), Balistidae (triggerfishes), and Haemulidae (grunts) These species are found primarily offshore from the Tortugas region northward to west central Florida. In addition to the species specific to reefs (e.g. triggerfishes, trunkfishes, ) the Florida Keys/Tortugas Region is considered a faunal transitional zone based on the presence of one or more demersal assemblages (Schomer & Drew, 1982). Starck, (1968) described assemblages of fish as either insular (reef-associated species from abiotically stable environments) or continental as represented by species found over muddy bottoms or turbid waters. The merging of temperate and tropical species is also apparent in other taxa e.g. invertebrates and benthic algae as reported in Chiappone and Sluka (1996). This unique convergence of abiotic and biotic factors provides for a diverse and variable fish communities relative to the more tropical (Caribbean) and more temperate (e.g. northern Gulf of Mexico) environments in the western Atlantic.

In this volume two different bodies of information are used to evaluate fish communities or populations of fishery targeted species: (1) fishery dependent information, primarily collected from fishermen and (2) fisheries independent data - information collected during ecological/oceanographic studies conducted by scientists.

Scales of information and definitions of fish assemblages found in the Tortugas Region vary considerably. Each scale relates to the degree to which a species is linked to the bottom or a benthic habitat. (micro-scale, refers to site specific study e.g. patch reef (<1 m2) or small reefs where fishes are confined to a particular habitat, meso-scale refers to shallow-water species that are less associated with the bottom but may range along a coastline for tens of kilometers, and thirdly coastal pelagic species associated with the open-ocean (e.g. across Caribbean Basin movements), movements of hundreds of kilometers.

FISHERIES-DEPENDENT INFORMATION

Fisheries management for the Dry Tortugas region is split between State of Florida Marine Fisheries Commission and two Federal Fishery Management Councils (FMC); the South Atlantic FMC and the Gulf of Mexico FMC. The State manages territorial waters 9 miles from in the Gulf of Mexico and is regulated by the Florida Marine Fisheries Commission (now under the Fish and Wildlife Conservation Commission). Waters under Federal regulation are those from the seaward edge of the State territorial waters out to the US Exclusive Economic zone. This jurisdiction excludes areas under management by the National Park Service which is responsible for managing the fisheries of the DTNP. The NPS manages fisheries in DRTO. The purpose of this section is to review the existing recreational fisheries data from the Tortugas Region and identify general retrospective fisheries trends.

Concern exists that excessive fishing could be deleterious to individual species, disrupt marine ecosystems, and damage the overall economy of the Florida Keys (Bohnsack et al. 1994). The fisheries of the Dry Tortugas region are comprised of three sectors: commercial fisheries; recreational fisheries ; and, the headboat fishery. Data sources for these analyses were from NMFS commercial landings, NMFS shrimp landings, FMRI TIP commercial landings, and the NMFS headboat surveys collected and assigned to regional statistical grids (Figure 12a), and the Dry Tortugas region is NMFS statistical grid 2 (Figure 12b). For most of the commercial fisheries resources we noted persistent declines in catches over the last decade (Table 2).

Commercial Fisheries

?Commercial fisheries of southern Florida and the Dry Tortugas region have been described previously by Bannerot (1990), Bohnsack et al. (1994), and Chiappone and Sulka (1996). Analyses of commercial and recreational sector fisheries operations within the FKNMS, including the Tortugas area, are described by Bohnsack et al. (1994). Two of the most important macroinvertebrate fisheries in the Dry Tortugas region are for pink shrimp and spiny lobster.

The Dry Tortugas region has been the principal fishing grounds for pink shrimp, and represents one the most valuable commercial fisheries in Florida waters. Pink shrimp appear to favor sediments comprised of calcareous- and sand-bottoms in waters between 9 and 44 m deep. The main commercial gears are double-winged trawls. The fishery was developed in the early 1950's, and the pink shrimp fishery grew to average annual landings that varied around 10 million lbs until the mid to late 1980's (Klima et al.1986), when they declined to a low of around 4 million pounds and remained there until about 1992. Since 1992 landings have been creeping back to the 10 million pound average (Figure 15a). Areal closures have been the primary measures used for managing the pink shrimp population off south Florida and the Tortugas grounds. Pink shrimp spawn year round, and juveniles settle inshore in the low salinity environments of coastal bays, tending to get larger (and mature) as they move further from shore (Ault et al. in press). Pink shrimp have high growth and mortality rates, and relatively high individual fecundity. Larvae are transported to inshore nursery grounds where they spend 2 to 6 mos before the estuarine-dependent juveniles migrate to offshore waters. Florida Bay in ENP represents a major nursery area for the Tortugas pink shrimp.

Figure 12a

Figure 12b

Table 2(1)

Table 2(2)

Table 2(3)

The Tortugas has an active spiny lobster fishery which was described by Harper (1992, 1995). The fishery apparently expanded in the early 1990 to an area northwest of the Tortugas where exceptionally large individual lobster were being landed (Harper, 1992). Average carapace size for landings from the Tortugas is the largest reported for Florida. Harper (1992, Figure 13) showed a comparison of average carapace lengths from statistical Grid 2 (Tortugas) with statistical Grid 1 (Florida Keys) and compared sizes from the Tortugas region in 1990 and 1991 (Harper, 1992, Figure 14). Average carapace length increased annually from the1987-88 season from 86 mm to 94 mm in the 1990-91, but dropped sharply to 83 mm the next year before increasing again to 87 mm in 1994-95 (Harper, 1995, Fig. 13, Area 2).

?A commercially trawl-fished concentration of deeper-dwelling royal red shrimp, Peloticus robustus, exists off the edge of the continental shelf in waters depths of 280 to 600 m south-southeast of the Dry Tortugas in the Straits of Florida (GMFMC 1981).

?Fish trapping occurred in the region since the late 1970s. Fish traps were prohibited in Florida waters in 1980 and in SAFMC waters in 1990. Amendment 16A under review by the GMFMC proposes regulations to prohibit the use of fish traps south of 25.05 degrees north latitude after 7 February 2001. An obvious point of concern is that catches of red grouper, black grouper, yellowtail snapper, mutton snapper, and gray snapper have all declined in the Dry Tortugas region over the past decade. Catches of hogfish have remained relatively low over that entire period (Figure 15).

Recreational Fisheries

Recreational fishing is an important and widely practiced activity in the Dry Tortugas region. Reef-fish fishing is a traditional attraction not only in the nearshore waters of the Florida Keys but on the offshore Reef Tract as it extends to the Tortugas Region. Bohnsack et al. (1994) showed that although snapper and grouper stocks appear to have declined in the Florida Keys; snapper, grouper headboat landings from the Dry Tortugas did not show distinct trends, as measured by numbers of fish, and were highly variable between 1981 and 1992. Yet, between 1986 and 1991, more grouper were landed from the Dry Tortugas than from the rest of the Florida Keys. Assessing trends in fish stocks from recreational fisheries-dependent information is tenuous at best. This is particularly true for the Dry Tortugas region, where reports are either sparse, not available, or not easily obtained. Recreational anglers in the Region operate from private boats and vessels for hire, which include charter boats, small guide boats and party boats (headboats). Of the few charterboats operating in FKNMS/TERSA, none have reported their catches to the NMFS Charterboat Survey (B. Palko, NMFS Panama City 1997, pers. comm.).

Figure 13

Figure 14

Figure 15

We used DRTO recreational hook-and-line catch data to identify fisheries-dependent trends in Park waters. A computerized data search, carried out at Everglades National Park's, South Florida Natural Resources Center Oracle Database for Fisheries, yielded a total of 593 angler interviews (1981-84) from nine fishing areas in DRTO. Funding constraints restricted angler interviews to the 1981-1984 period. More recent attempts in 1991 to reinstitute angler surveys at DRTO met with a similar fate. Variables collected in the DTNP recreational fishery survey included: numbers of fish landed, number of fish released, nominal fishing effort (hours fished), number of people fishing, species preference, area fished, trip hours, and place of angler residence. These data were summarized to provide a brief description of the DRTO fishery (Table 3), and to calculate catch-per-unit-effort (CPUE) information for combined snappers and groupers species by year. Catch rates of individual trips were calculated after Malvestuto (1983). Only those anglers successful in catching a species were used to calculate mean annual catch rates to avoid bias in the possible change in the proportion of effort applicable to a species each year.

Two species groups, combined snapper and combined grouper, dominated the catches from the DRTO recreational fishery, and had catch numbers marginally sufficient to conduct a CPUE analysis by year. The snappers were comprised primarily of yellowtail and gray snapper. Grouper species made up the overwhelmingly majority of the combined grouper category. Combined snapper and grouper CPUE increased by year (Figure 16) possibly reflecting the small number of annual interviews/trips (range: 85 to 280). A similar pattern in effort (angler-days) was shown for Dry Tortugas headboat survey data for the years, 1981-84. Overall, recreational fishing activity based on angler interviews reported from areas 1-9 indicates that area 5 (Garden Key) is the single most heavily used area (63% of all trips) in DRTO (Table 4). The next most popular area is area 2 near Loggerhead Key (16%) followed by area 8 (7%). Over the entire period of record, more than 76% of the catches of combined snapper species came from areas 2, 5, and 8.

Table 3.- Description of DRTO recreational fishery from 1981-84.

Table 4.- Number of recreational fishing interview-trips by fishing area in DRTO from 1981-1984.

Figure 16

Headboat Fishery

We used data from the 1981-97 NMFS headboat survey (Bohnsack et al. 1994; Dixon and Huntsman, 1992) to provide information on FKNMS fisheries trends. Headboats are large vessels where anglers pay "by the head" to fish (Bohnsack et al. (1994). Fisheries biologists compile operator logbook records and make length measurements of the catch. The recreational headboat data for Monroe County are divided into geographical sub-areas, one for the Florida Keys and the other for the Dry Tortugas (FKNMS). Headboat data provide total numbers of individuals in the catch (landings) as well as total weight in the catch by species by year. The landings data were further grouped into reef fishes and non-reef fishes. Reef fishes included both primary and secondary reef species as after Starck (1968). Mean annual catch rates were calculated by dividing total annual estimated catch (landings) for each combined species group by total annual estimated effort (angler days) (Table 5).

Bohnsack et al.(1994), reported that annual estimated headboat angler-days for the Dry Tortugas (TERSA) were relatively constant through 1992, however since 1993 estimated angler-days have decreased, substantially from the peak years of 1987 (21,667) and 1988 (23,600), to 10, 209 in 1997, following the elimination of headboats in DRTO in 1995 (Table 6). Headboat CPUE analysis for the Dry Tortugas (TERSA) indicated that combined grouper catch rates remained relatively constant through 1997. Between 1986 and 1991 more grouper were landed in the Dry Tortugas than from the rest of the Florida Keys (Bohnsack et al. 1994). Combined snapper catch rates/landings and total reef catch rates/landings have declined substantially since the late 1980’s to less than half of what was reported during the earlier years (Figure 17).

Reef fish dominated the 1981-92 headboat catch (Bohnsack et al. (1994) From 1981-92 reef fish landings averaged 98% of the total headboat landings, however from 1992 to 1997 reef fish dominance was reduced to an average of 92% of the total catch. Between 1982 and 1992, king mackerel (a non-reef species), catches totaled 6,362 fish compared to 14,795 landed between 1993 and 1997 (Table 6). From 1993-1997, dominant reef fish species in decreasing order of abundance (landings) were yellowtail snapper, white grunt, gray snapper, and mutton snapper. Total reef species (snappers) declined from an estimated 304,257 fish in 1988 to 36,570 fish in 1997, but this decline was probably due to a major reduction in effort (23,600 angler days to 10, 209 angler days.

Except for NPS hook & line data, recreational fisheries non-headboat data from the cooperative state-federal marine recreational fisheries statistical survey (MRFSS) were not available for the Dry Tortugas area because the smallest geographical reporting units for Florida were for the east and west coasts.

Federal and state fishery management actions from 1976 to 1992, affecting Monroe County and FKNMS are reviewed in Bohnsack et al. (1994). Those regulations enacted since 1992 which affect the dominant recreational species in TERSA are as follows: for 1994, the minimum size (TL) for mutton snapper and hogfish was increased to 16" and 12", respectively while a 10"commercial minimum size for vermillion snapper was established, then changed to 12" for SAMFC, while, for the State of Florida, 8 " is the minimum size for the Gulf side and 10" is the minimum size for the Atlantic side. King mackerel regulations also have varied considerably. Between 1992-1997, bag limits have changed from 5 to 10 to 2 fish per person per day. There are also different areal closures and recreational/commercial allocations of catch permitted for every two years (Bennett and Weinberger in prep.) creating, at times, mass confusion for fishermen and scientist alike. Additional modifications in minimum size limits for many species of the snapper-grouper complex for State, GMFMC, and SAFMC waters are expected to be imposed by mid-year 1999.

Table 5.- Total estimated headboat angler-days for the Florida Keys and Dry Tortugas, 1981-97: NMFS Headboat Survey (Dixon and Huntsman 1982; Bohnsack et al. 1994).

Figure 17

Table 6

FISHERIES-INDEPENDENT INFORMATION

Tropical ecosystems are characterized overall by relatively large number of species with very complex inter-relationships. These are usually not well represented through exclsuive use of fisheries-dependent data. Thus, strategic use of well-designed fisheries-independent survey sampling is warranted. The Florida Keys coral reef tract has served as the focal point for reef fish assessment surveys in the Florida Keys. The clear, warm-water and high diversity of reef fish in the Florida Keys coral reef tract provides a unique environment to assess multispecies fisheries using fishery-independent survey sampling from the coastal bay to coral reef environments. With accurate and unbiased estimates of reef fish abundance available, we then used the correlation structure of fish stocks and sizes of fish relative to environmental 'habitat’ features in a design-based approach to obtain high precision data for fish stock assessment at relatively low cost (Cochran, 1977; Ault et al., 1998c; Ault and Smith, 1998) and to facilitate effective decision making and to improve fishery management performance (Ault and Fox, 1989; Rothschild et al., 1996).

Visual Assessments of the Reef Fish Community Assemblages

Fisheries-independent coral reef fish research in the Dry Tortugas region has primarily focused on juvenile and adult assemblages and typically relies on visual census survey techniques using SCUBA that can employ photography or videography (Alevizon and Brooks, 1975 and others) as well as fixed base or roving methodologies to quantify area and time (Jones and Thompson, 1978; Thompson and Schmidt, 1977; Clark, 1986; Bohnsack and Bannerot, 1986; Ault et al. 1997, 1998). Several fisheries-independent databases were used to present a view of the status of the resources in the Florida Keys and Dry Tortugas region. These data included the NMFS stationary visual censuses of reef fishes and habitat at multiple sites in the Tortugas area using the Bohnsack/Bannerot technique (1994-1998). The FMRI reef fishes visual census at multiple sites in the Tortugas area using the Kimmel technique (1990-1995). The FDEP/FMRI spiny lobster census at multiple sites in the Tortugas area (Appendices 1 and 2). These studies are mostly repeated surveys from a single reef site or a few studies which include many different reef sites contrasting the reef fish species of the Tortugas region with other areas in the Florida Keys and tropical western Atlantic locations. Results vary by study objectives and have not always been comparable among sites.

Fishery-independent visual estimates of the abundance and size distributions of multispecies reef fish populations were taken along the Florida Keys reef track continuously from 1979 to 1998 (Figure 19 and Table 7) by 12 highly-trained and experienced divers using the stationary visual survey method of Bohnsack and Bannerot (1986). This non-destructive method provides reliable quantitative estimates of species abundance, frequency-of-occurrence, and size structure for the reef fish community. Divers recorded the abundance and the minimum, mean and maximum lengths of each species seen in 5 min within randomly selected 7.5 m radius circular quadrats. Underwater visual estimates of reef fish size and abundance are frequently made (Bellwood and Alcala, 1988; Harvey and Shortis, 1996); however, obtaining accurate and precise visual estimates of fish length requires well-trained and experienced observers because objects in water appear magnified and closer than their actual range (Bell et al., 1985; Harvey and Shortis, 1996). To improve accuracy, divers continuously calibrated their length estimates using a 30 cm ruler attached perpendicular to the far end of a meter-stick. Divers without calibration sticks have been shown to obtain a mean accuracy of 86% for length estimates (St. John et al., 1990). Fish assemblages were quantitatively described using the following parameters: total species present at a given time-site (or given habitat); abundance or biomass of fishes at a given time in a given habitat; size-frequency distribution of fishes; seasonal changes in total number of species, abundance, or biomass in a given habitat; and, presence of spawning aggregations or number, size, and fecundity of spawning fishes.

Figure 19

Table 7(1)

Table 7(2)

The spatial and temporal distribution of reef fish and spiny lobster visual sampling is shown (Figure 20 and Table 8). A total of 1081 visual reef fish samples have been taken in the Dry Tortugas from 1990-1998. About 600 samples from 60 sites were also taken for spiny lobster. A 3-dimensional characterization of the reef fish sampling sites over Dry Tortugas bathymetry is given in Figure 21, showing the location of Sherwood Forest (red flag) on the northwest slope of the Tortugas Bank. Available fishery-independent information were summarized/synthesized to describe the spatial-temporal distribution, abundance, and species composition and relationships between fishes of the Tortugas region and habitats.

Rationale for Multispecies Stock Assessments

The goal of this section is to present a technically sound quantitative method for multispecies management assessments in the Dry Tortugas and Florida Keys. For this purpose, we present an integrated baseline assessment to reference the status of the multispecies fishery in order to eventually evaluate effects of management changes in FKNMS (U.S. Department of Commerce 1996). Using fishery-independent data, we conduct a 18-year retrospective analytical yield assessment of economically important Florida Keys reef fish stocks to elucidate the effects of fishing and to help define an effective fishery management strategy.

?A key to our ability to assess reef fish stocks was the use of "average size’ (in length) of fish in the exploitable phase of the population" or 'average size’ as an indicator of stock status using either reef fish visual survey data or headboat landings data. Headboats are party boats that take more than 15 anglers fishing per trip (Dixon and Huntsman, 1992). The use of 'average size’ in stock assessment has deep roots in traditional fisheries management (Beverton and Holt 1956, 1957; Ricker 1975). The statistic provides a population level metric that integrates individual metabolic variables such as inter-dependent growth, mortality and reproductive processes. The 'average size’ statistic also is an important index of fishing effects, because persistent heavy fishing reduces the average size of the population over time, making the stock younger through a process known as "juvenescence" which successively eliminates older, more fecund size classes (Ricker, 1963) (Figure 22). This is extremely important in the context of stock and recruitment, since the fecundity potential of individuals increases exponentially with size. In general, the average length of fish in the exploitable phase (i.e., between the size at first capture L’ and the maximum size L_max ) is highly correlated with average population size, and so reflects the rate of fishing mortality operating in the fishery.

Theoretically, the average size of fish landed for any given species should be equal to the average size in the exploited phase of the remaining population just after fishing. In other words, we hypothesize that the fishery-independent diver visual survey estimates of average length should equal fishery-dependent estimates derived from headboat angler catches. The greater the correlation between the two independent estimates of 'average size’, the more robust 'average length’ should be as an indicator of stock status subject to exploitation.

Figure 20

Table 8

Figure 21

Figure 22

To calibrate the fisheries-independent visual assessments of reef fish, we also used the 1981-1995 NMFS headboat catch and effort data (Bohnsack et al., 1994; Dixon and Huntsman, 1992) to provide fishery-dependent population estimates comparable to those from the visual survey. Headboat data provides total numbers of individuals in the catch as well as total weight in the catch by species by year. Spatial and temporal patterns in abundance and correlative linkages to habitat types were qualitatively analyzed using 3-D visualization software (IDL -- Interactive Data Language, Research Systems, Inc., Boulder, CO) by reef site throughout the Florida Keys for various survey years. Multivariate statistical analysis (Johnson and Wichern, 1992; Venables and Ripley, 1994) was used to assess variance-covariance and correlation structures between reef fish density and selected environmental and fish community auxiliary covariates.

Abundance and Recruitment Patterns -- Status of the Stocks

?A stock assessment indicator variable is a quantitative measure that reflects the status of a population subjected to fishing or other environmental changes. Because reef fishes integrate aspects of the coastal ocean environment over their lifetime, a robust measure of population "health" or status can provide a sensitive indicator of direct and indirect stress on the stock, and perhaps the regional marine ecosystem (Fausch et al., 1990). Population health for reef fish communities can best be described using the metabolic-based pool variable 'average length in the exploitable phase of the stock’. Exploitation effects were assessed using a new length-based algorithm that calculates total mortality rates from estimates of "average length of fish in the exploitable phase of the stock". These estimates were highly correlated for two statistically independent data sources on reef fish: fishery-independent diver observations and fishery-dependent head boat catches.

?The REEFS model was configured to assess two fishery management decision making endpoints, yield-per-recruit (YPR), and spawning potential ratio (SPR). Fishery management endpoints are relatively robust measures of potential yields and recruitment. As such, they help to focus on biological (size) and fishing (intensity) controls for managing current and future fishery production. Yield-per-recruit (YPR) is then calculated by scaling yield to average recruitment from the right hand side of the above equation. Spawning stock biomass SSB is a measure of the stock’s reproductive potential or capacity to produce newborn, ultimately realized at the population level as successful cohorts or year classes.

Spawning potential ratio (SPR) is a contemporaneous management endpoint that measures the stock’s potential capacity to produce optimum yields on a sustainable basis. SPR is a fraction expressed as the ratio of exploited spawning stock biomass relative to the equilibrium unexploited SSB. Resultant estimated SPRs are then compared to the U.S. Federal standards which define 30% SPR as the "overfishing" threshold (Rosenberg et al., 1996). The Reef-fish Equilibrium Exploitation Fishery Simulation (REEFS) model was used estimates of fishing mortality to assess yield-per-recruit relative to fishing intensity and gear selectivity, and spawning potential ratio (SPR) relative to U.S. federal "overfishing" standards.

The analyses of Ault et al. (1998) show that 13 of 16 groupers (Epinephilinae), 7 of 13 snappers (Lutjanidae), one wrasse (Labridae), and 2 of 5 grunts (Haemulidae) are below the 30% SPR overfishing minimum (Figure 23). Some stocks appear to have been chronically overfished since the late 1970's. The Florida Keys reef fishery exhibits classic "serial overfishing" in which the largest, most desirable and vulnerable species are depleted by fishing. Rapid growth of the barracuda population (Sphyraenidae) during the same period suggests that fishing has contributed to substantial changes in community structure and dynamics.

To further examine the affects of exploitation Keys-wide for a few representative groupers (black, red, and all large grouper combined), snappers (yellowtail, gray and mutton), grunts (white and bluestriped), and lobsters, we estimated the adult 'average size’ (a measure of yield potential) and the juvenile pre-exploited phase density (a measure of population stability and resiliency). We compared these spatially-explicit estimates for the Keys with those from sampling in the Dry Tortugas region.

Black Grouper.- Black grouper density (measured as number of fish seen per visual sample) of the pre-recruit phase (i.e., juveniles) ranged from 1 fish per thousand samples to 1 fish every 2 samples (Figure 24a). There were many sites where no juveniles were seen (black triangles). The greatest concentration of small fish was in the Dry Tortugas. The average size of adult fishes reveals that many sites have no adult fishes present at all (Figure 24b). Only one site in Key Largo had an adult-phase fish, and it was very near to the minumum size of first capture. Adult black grouper tend to be rarely observed in the spatilly-synoptic sampling regime. The greatest number of size of legal fish is found in the Dry Tortugas. Interestingly, the largest average size of specimens observed was found within the boundaries of DTNP and on the northwest side of the Tortugas Bank (Figure 24c). Both adult and juvenile abundance is highest in the southern end of the Keys (Figure 25). This seems to indicate the National Park in the Tortugas has served a de facto reserve function for the black grouper population. The current level of exploitation estimated from the 'average size’ of black grouper was compared against the Federal definitions of overfishing stated in the reauthorization of the Magnusson-Steven’s Fishery Management Conservation Act of 1998 (Figure 26). The act requires that the stock’s theoretical "maximum sustainable yield " (or MSY) be established directly or by proxy. Proxy estimates of MSY are detailed on an analytical yield diagram as either: F_=M, the fishing mortality rate that equals the natural mortality rate; F_0.1, the point of the YPR curve where the slope is equal to 10% of the slope of the YPR curve at the origin; or, F_max, the maximum point of the YPR curve. The current estimate of fishing mortality F₉₇ for the black grouper is 0.63. That means that the fishing mortality rate, measured as the nominal fishing mortality (number of anglers and commercial fisherman) times the catchability coefficient (the fraction of the stock removed per unit of nominal fishing effort) is about 4.2 times the MSY level. This indicates the black grouper stock is currently seriously overfished. What this means to the fishermen using black grouper as an example is that the average size of black grouper caught in 1997 is 40% of what it once was (22.5 versus 9 lbs). What it means to the stability and resiliency of the black grouper population is that the spawning stock biomass is now 5% of what it once was (Figure 26)! In other words, stock is currently at 5% SPR.

Figure 23

Figure 24a

Figure 24b

Figure 24c

Figure 25

Figure 26

Red Grouper.- Red grouper density of the pre-recruit phase (i.e., juveniles) also ranged from 1 fish per thousand samples to 1 fish every 2 samples (Figure 27a). There were many sites where no juveniles were seen (black triangles). The greatest concentration of small fish was in the Dry Tortugas. The average size of adult fishes reveals that many sites have no adult fishes present at all (Figure 27b). Only one site in Key Largo had an adult-phase fish, and it was very near to the minimum size of first capture. Adult red grouper tend to be rarely observed in the spatially-synoptic sampling regime. The greatest number of size of legal fish is found in the Dry Tortugas. Interestingly, the largest average size of specimens observed was found within the boundaries of DTNP and on the northwest side of the Tortugas Bank. Both adult and juvenile abundance is highest in the southern end of the Keys (Figure 28). This seems to indicate the National Park in the Tortugas has served a de facto reserve function for the red grouper population. The current level of exploitation or fishing mortality estimated from the 'average size’ of red grouper is 0.74, or about 5 times the level required to obtain MSY. Stock is currently at 22% SPR.

All Large Gropuer Combined.- Total large groupers combined density of the pre-recruit phase (i.e., juveniles) also ranged from 1 fish per thousand samples to 1 fish every 2 samples (Figure 29a). There were many sites where no juvenile groupers were seen (black triangles). The greatest concentration of small fish was in the Dry Tortugas. The average size of adult fishes again reveals that many sites have no adult fishes present at all (Figure 29b). This is a very serious problem that suggests on average the sighting of a legal grouper tends to be a relatively rare event. Most frequent sightings occur in the Dry Tortugas.

Yellowtail snapper.- Yellowtail snapper juvenile (pre-recruit) densities are shown (Figure 30b). Juveniles were more readily seen throughout the Keys. However, the distribution of adult yellowtail snappers as measured by average size shows a somewhat bleak picture of far fewer sites occupied and the average size of fish at those sites observed is near the minimum size of first capture, indicating a very high fishing mortality rate (Figure 30a). Adults were more abundant in the north Keys (Figure 31). Stock is currently at 45% SPR.

Figure 27a

Figure 27b

Figure 28

Figure 29a

Figure 29b

Figure 30a

Figure 30b

Figure 31

Gray Snapper.- Gray snapper juvenile (pre-recruit) densities are shown (Figure 32b). Juveniles were more readily seen throughout the Keys. However, the distribution of adult gray snappers as measured by average size shows a somewhat bleak picture of far fewer sites occupied and the average size of fish at those sites observed is near the minimum size of first capture, indicating a very high fishing mortality rate (Figure 32a). Adults and juveniles were present throughout the Keys, but experiencing a very high rate of fishing mortality (Figure 33). Stock is currently at 24% SPR.

Mutton Snapper.- Average size of adults (Figure 34a). Density of juveniles (Figure 34b). Stock is currently at 52% SPR.

White Grunt.- Adult white grunt appear to be more abundant in the north and middle Keys, while juveniles appear to be more abundant in the middle and southern Keys (Figure 35). Stock is currently at 15% SPR.

Bluestriped Grunt.- Both adult and juvenile bluestriped grunt appear to be more abundant in the north and middle Keys (Figure 36). Stock is currently at 60% SPR.

Quantitative data on abundance of reef fishes in the Tortugas Region are available from 4 previous fishery-independent studies and one on-going study. Thompson and Schmidt (1977), using the roving visual census technique, reported that nine reef fish families dominated the fauna in Park waters: Pomacentridae, Scaridae, Haemulidae, Labridae, Chaetodontidae, Lutjanidae, Gobiidae, and Acanthuridae. (See also Table 6, Chiappone and Sluka 1996). Jones and Thompson (1978)) using the same census method, observed similar families of fishes between the Upper Florida Keys and Park waters. They will be discussed in a following section on comparisons of fish fauna between the Florida Keys and the Tortugas Region.

Rydene and Kimmel, using the modified Bohnsack-Bannerot visual census method (Bohnsack and Bannerot 1986) found substantial differences in the numbers of individuals found per census among Park habitat sites. By family, the gobies (Gobiidae) dominated by Coryphopterus personatus were numerically dominate at fore reef and patch reef sites. Excluding gobies, the dominant families were Pomacentridae, Scaridae, Labridae, and Haemulidae. No significant trends were found for fish abundance over the five-year period when C. personatus was included in the analysis. However, when C. personatus was excluded from the abundance analysis, a significant decline in fish

abundance was found for 1994. Based on a analysis of selected larger predatory species (red grouper/black grouper/yellowtail snapper) and ecologically important species (cocoa damselfish/Striped parrotfish), they found that the number of red and black grouper did not vary significantly during the overall period while only black grouper abundance varied significantly by habitat, being higher at Pulaski Shoal than at Loggerhead Key. Yellowtail snapper, striped parrotfish and cocoa damselfish abundance varied significantly among years, but when compared together with the other two species they did not exhibit a consistent trend in abundance over the five years.

Figure 32a

Figure 32b

Figure 33

Figure 34a

Figure 34b

Figure 35

Figure 36

During September 1992, McKenna (1997) examined fish assemblages from four Park sites (Texas Rock, Pulaski Shoal, Bird/Long Key, and the Mavro Vetranic grounding site) while using the modified Bohnsack-Bannerot visual count survey technique. He reported that schools of small species associated with the bottom including wrasses (Halicchores bivattatus) and juvenile parrotfish (Scarus croicensis) dominated a disturbed site (Mavro Vertranic grounding site) whereas a variety of schooling species including C. personatus and juvenile grunts (Haemulidae) along with those species of a more pelagic disposition (Caranx ruber) dominated undisturbed sites.

Bohnsack and McClellan (1998) and Ault et al. (1998) have assessed reef fish assemblages both inside and outside of DRTO boundaries using stationary point visual survey method of Bohnsack and Bannerot (1986) and a 15-minute roving predator search. This methodology provides reliable quantitative estimates of species abundance, frequency of occurrence, and size structure of the reef fish community (Ault et al. 1998). Between 1994 and 1997, a total of 518 stationary samples were collected from 9 reefs inside the Park and seven reefs outside the Park (six reefs in FKNMS and one reef (Sherwood Forest) outside FKNMS) (Bohnsack and McClellan 1998). A total of 162 species (76,408 individual fish) were found in this effort (Bohnsack and McClellan 1998). Results indicated that the masked goby C personatus, Striped parrotfish Scarus croicensis, and the bluehead wrasse Thalassoma bifasciatum were, overall, the most abundant species. Recently, frequency of occurrence has been suggested as a better indicator of abundance especially when making comparisons among other studies with different visual census techniques. Thus, excluding the masked goby (which is often found less frequently 41% as in this study, due to schooling in high densities in deeper habitats), by frequency of occurrence the striped parrotfish (91%), bluehead wrasse (81%), and the cocoa damselfish, Pomacentrus variabilis (74%) were the species of greatest abundance. Comparisons of reef fish assemblages inside and outside Park boundaries are presented in a following section.

Spiny Lobster.- The density of adult lobsters sampled in the Dry Tortugas region is given (Figure 37a). The highest density of lobsters is clearly seen within DTNP boundaries. The average size of adult lobsters sampled in the Dry Tortugas region is given (Figure 37b). The highest concentrations of large adult lobsters is clearly seen within DTNP boundaries.

Biological and fishery characteristics of the spiny lobster (Panulirus argus) in Florida waters is relatively well studied as reviewed in Volume 6. Studies conducted in the Tortugas Region are particularly important because they have clearly shown the effects of recreational harvesting on protected populations of spiny lobster. Davis (1974), during a 29-month study with a seasonal closure (2 lobsters per person per day) noted a 35-60 percent reduction in lobster den density, indicating that a recreational harvest can significantly impact the stock (Davis 1974, 1977). After the initial 8-month closure to recreational harvest, an estimated 26,500 lobsters were removed, a 57 percent decline in CPUE and a 78 percent decline in pre-harvest levels were recorded (Davis 1977). Recovery of the stock following closure was moderate, and the study suggested that the Dry Tortugas stock could not sustain harvests of this kind (Davis 1977). Information on the life history of spiny lobster indicates the present minimum legal size (76 mm CL) effectively prevents most lobsters from reaching full sexual maturity in the Florida Keys (Davis, 1977). The fishery has greatly reduced the age and size distribution of the Florida Keys stock (Davis 1975, 1981). Davis (1974) and Bertelsen and Hunt (1996) reported that the size at maturity of a protected lobster population in the Dry Tortugas is greater (79 mm CL) than the fished population in the Florida Keys. Overexploitation of spiny lobster in the Keys has caused a shift in reproductive maturity towards a smaller size (minimum 70-mm CL) whereas most are reproductively active at 105 mm CL in the Dry Tortugas (Bertelsen and Hunt 1996).

Figure 37a

Figure 37b

The impacts of over-exploitation resulting in a declining CPUE and reproductive effects is clearly evident in the Florida spiny lobster fishery, yet a relatively stable harvest rate suggests constant recruitment to the Florida Keys region. Due to the uncertainty regarding the stock origin, management of this fishery is quite complex. Although Florida Keys spiny lobster stock only spawns during the summer months, year-round recruitment suggests a possible Caribbean source to the Florida Keys. The fate of larvae spawned in Florida waters is unknown, however the protected stocks in DRTO (Davis and Doldrill 1980) suggest a possible source of larvae to the Keys as reported by Lee et al, (1994, 99). Without the complete understanding and control of parental stocks, yield-per-recruit losses imposed by growth retardation will continue to be a major concern to the Florida Keys fishery.

To evaluate the stock status of spiny lobster relative to the 30% SPR U. S. Federal standard of overfishing, lobster reproductive activity (eggs per recruit ratios) and abundance/size multiyear studies are underway between an unexploited population in DRTO and areas within the fished populations along the Florida Reef Tract including the Upper Keys, Middle Keys, Lower Keys and west of Key West ( Bertelsen and Hunt 1996, 1997, 1999). Other small unexploited areas (Looe Key reef and FKNMS marine reserves) are being added to the study to support investigations in DRTO. In 1995 during the first year of study, Dry Tortugas sampling did not distinguish between inside and outside Park boundaries and data were combined as "Dry Tortugas". In 1996 they divided the Dry Tortugas into one deepwater site within the Park southern boundary and one site outside the Park’s southern boundary and reported results separately.

Large differences in spiny lobster abundance were found between the Dry Tortugas and the other four regions along the reef tract. Based on preliminary information for 1995, the peak abundance (legal-sized lobsters observed per hour) in the Dry Tortugas (90) was 5 to 16 times the peak abundance’s found in all other regions. In 1996, lobster abundance estimates dropped substantially compared to 1995 data for the Dry Tortugas, possibly due to an artifact in sampling. Poor weather and failure to sample shallow-water patch reefs because reproductive activity was non-existent (although they contained many lobsters in the previous year in the Dry Tortugas), may have caused the change in lobster abundance between the 2 years (Bertelsen and Hunt 1997). In the lower Keys abundance estimates increased dramatically compared to the previous year.

Differences in lobster size and egg production were found among regions. Bertelsen and Hunt (1996) reported that the average size lobster producing eggs in the Dry Tortugas is roughly equal to the largest lobsters producing eggs among all regions studied. The lobster reproductive season in the Dry Tortugas peaks earlier, with a higher percentage of females carrying eggs than in all areas. From data collected in 1995, they reported that the typical female Dry Tortugas/DRTO lobster produces 900,000 eggs per clutch whereas those in all other areas (Upper Keys to Key West) typically produce one-third the number of eggs per clutch. Davis (1975) suggested a pattern of senescence in older, larger females (>126 mm CL), however Bertelsen and Hunt (1996) reported that all larger females (115-145 mm CL) bore eggs.

From data collected during the second year of an on-going multiyear spiny lobster population study in the Florida Keys, Bertelsen and Hunt (1997), report that among all regions, by far, the largest lobsters were found inside and outside Park boundaries. Sizes were also similar between inside and outside Park sampling sites, as the largest lobsters, up to 180, and 179 mm CL, were observed in, and outside Park boundaries, respectively. The smallest lobster was also reported from inside Park waters. They suggested that if the relatively larger number and size of lobsters found outside the Park boundary originate wholly or partly within park waters then perhaps a similar fate awaits the now relatively depauperate offshore area in the lower Keys.

Regional Biodiversity of DTRO, TERSA and the Florida Keys

A list of the common and scientific names of marine fish species/families of the Tortugas region recorded from fishery studies evaluated in this volume are in Appendix 3. Schmidt and Pikula (1997) compiled an annotated bibliography on the scientific studies conducted at DRTO. For a list of fishes recorded by habitat elsewhere in the Florida Keys, adjacent environs and in the Florida Bay area the reader is referred to Chiappone and Sluka (1996) which includes some but not all of the following reports for (1) the Florida Keys: Springer and McErlean, 1962; Roessler 1965, 1967; Starck, 1968; Voss et al. 1969; Jones and Thompson 1978; Bohnsack and Talbot, 1980; Jones and Clark, 1981, and for (2) Florida Bay: Tabb & Manning, 1961; Roessler, 1967; Schmidt, 1979, 1993; Thayer et al, 1987, 1989, Sogard et al., 1987, 1989a,b; Ley, 1992, Matheson, 1995.

The general characteristics of reef fish assemblages including species diversity, trophic structure, and factors affecting reef fishes such as recruitment and post recruitment processes have been described in detail for the Florida Keys region and adjacent environs in Chiappone and Sluka (1996). To briefly summarize, relative to Tortugas reef fish assemblages: reef fish constitute a highly diverse fauna, packing in a considerably high number of species (~ 400 in the Florida Keys to as many as 900 to 1,500 species in the Indo-Pacific) into a relatively small spatial scale (m² to km²) They may reside near reef structure during the day and feed in adjacent seagrass beds (e.g. grunts and snappers) at night. Trophic classifications for reef fish indicating the general type of prey items they consume include herbivores, planktivores, benthic invertivores, benthic carnivores, pelagic carnivores, coralivores, omnivores, and detritus feeders. Generally, they are mostly herbivorous bottom feeders and feed primarily during the night. Larger carnivorous species are usually nocturnal and prey upon invertebrates and fishes. Most reef fish spawn in the water column and have pelagic reproductive strategies. Spawning aggregations may decrease adult predation and typically occur in areas where their increased probability of dispersal via oceanic currents. The larval stages may range from weeks ( gobies, grunts) to months (surgeonfishes). Reef fish are highly fecund and can expend more energy in egg production than in parental care. Post-recruitment success can vary spatially and temporally and is affected by competition and predation. Abundance of reef resources and growth will affect distribution patterns. Slower growing fish will become more available to predation and increase their mortality rate. Bohnsack (1983b) concluded that censuses of reefs over different time periods will yield similar changes in the overall composition of the reef fish assemblage, while differences will be apparent in the turnover rate. Thus results of studies concerning the spatial and temporal distribution and post recruitment processes will vary based on the life histories of the species involved.

Based on data collected through 1997, Bohnsack and co-workers (1998) found that the largest differences in reef fish assemblages were found between reefs inside and outside the Park. Of the 162 species found inside and outside Park waters, 114 species were common to both areas. The reefs inside had 26 species not observed outside, while 20 species were only observed outside.

There were obvious habitat differences between sites inside and outside Park waters. They found that reef sites inside the Park were shallower and were more likely to have sand and seagrass bottom than outside where the substrate was comprised mostly of coral or "rock". Inside, average depths ranged from 3.3 m at Little Africa, adjacent to Loggerhead Key to 13 m ft at Pulaski Shoal whereas average depths varied from13.5 m at Twin Peaks to ~26 m at Black Coral Reef outside park waters..

Trends in abundance were examined. They reported that reefs outside the Park tended to have higher numbers of individuals such as planktiviorus feeding damselfishes and total fish biomass. Selected species/groups of recreational importance such as yellowtail snapper, total grouper, and mutton snapper were found to vary somewhat similarly inside and outside Park waters. Hogfish and black grouper were observed more frequently inside the Park than outside.

Between 1994 and 1997, a total of 127 roving piscivorous predatory surveys were conducted on 14 reef sites inside the Park and 2 reefs outside (FKNMS). Thirty-nine predatory species were reported from all samples combined. However, the vast majority (98% ) of the samples were taken in Park waters. Thirty-eight species were common to both areas. Due to the lack of samples taken from outside Park boundaries no real differences in predatory surveys can be presented between reefs inside and outside of Park waters. In addition no finfish spawning aggregations were found during surveys inside and outside of Park waters.

Based on earlier studies, the number of reef fish species found among different reefs among the Florida Keys can vary significantly as a result of study duration and survey techniques (destructive versus non-destructive) employed. Starck (1968) recorded greater diversity of species found at Alligator Reef (517) relative to the Tortugas Region (442), but this can be partially explained by the greater effort using both destructive and non-destructive sampling techniques. Starck postulated that their were no significant differences in the fish fauna between the two areas.

In short-term, roving visual census surveys, Jones and Thompson (1978) compared four reefs in the Upper Florida Keys with four reefs in the Dry Tortugas (National Park)-Long Key, White Shoal, Loggerhead Reef, and French Wreck. They found 165 species between both areas, with 115 species common to both reefs in the Upper Keys and the Dry Tortugas. The reefs in the upper Florida Keys had 31 species not observed in the Dry Tortugas while 19 species were observed only in the Dry Tortugas.

They reported that the Pennekamp communities were generally higher in relative species richness, and abundance than in the Tortugas communities. For selected grouper species, red grouper (E. morio) and Myctoperca venenosa were recorded in the Tortugas but not in Pennekamp communities. In Park communities, White shoal and Loggerhead reefs showed the lowest overall diversity but the greatest similarity with one another. Both sites were similar with regards to species diversity with Long Key Reef which is an entirely different structure in terms of vertical profile (spur & groove formation) water depth, and coral species composition, and physiography. They implied that the natural reef communities of the Tortugas are conservative with regard to their fish communities.

To summarize, in the number of fish species observed between the Dry Tortugas region and elsewhere in the Florida Keys, similarities do exist in species composition, yet differences may be attributed to sampling methodology and duration of study. For example, approximately 74 to 188 species were identified on reefs from the Florida Keys to Dry Tortugas in short-term (< 2 years), visual census surveys whereas between 300 and 389 species were observed in studies of longer duration using a variety of sampling techniques.

Tortugas Region Fishes at Risk.

Until only very recently have marine fish species been proposed for inclusion on national and international rare/endangered animal lists (Musick 1998; Hudson and Mason 1996). Although lists of rare and endangered animals have been compiled by various federal/state/( as applicable to the Endangered Species Act) and international conservation institutions and published over the past 10-25 years, new criteria were needed, then developed and applied to determine the extinction risk for marine fish species. Estuarine/marine fish have been poorly represented (only six out of 1,321 taxa, as of 1994, on the federally endangered species list)) with not a single wholly oceanic marine finfish species included (Huntsman 1994). Because life histories of marine fish species are not well known (Gilbert 1992), Florida has virtually ignored all marine fish in their lists of rare and endangered biota as prepared by the Florida Committee of Rare and Endangered Plants and Animals (FCREPA),. This situation exists partly because (1): there is societal and scientific doubt that marine fish species can become extinct, and (2) they have traditionally been of lower conservation concern than their terrestrial counterparts.

However, with the advent of the marine reserve movement, particularly through the efforts of the Plan Development Team Report (1992), anonymously authored by Jim Bohnsack, increasing attention has been placed on the realization that marine ecological/fishery reserves can provide the framework for protecting individual species of both ecological and economic value.

In 1996, the World Conservation Union (ICUN) and the World Wildlife Fund (WWF) published rare/threatened marine species criteria and revised animal lists to include marine fish species. Concurrently, the American Fisheries Society (AFS) has evaluated the risk of extinction for marine fish species using new quantitative criteria adopted by IUCN (Hudson and Mace 1996), and published a list of 68 marine fish species at risk in North America (Musick 1998). Of the 68 marine finfish species listed, eight species are found in the Tortugas region. The IUCN list includes 12 species from the Region..

In the Tortugas region, marine fish species receive a certain amount of protection by residing within Park boundaries. Commercial fishing activities and recreational lobster fishing activities are banned in Park waters. Recreational (hook & line) finfish activities in Park waters are regulated by Park fishing regulations which follow recreational fishing regulations for the State of Florida. In addition, there are Park fishery regulations in place which protect juvenile ornamental fish species, many of which are potentially rare reef associated species.

The Dry Tortugas Region, one of the world's most ecologically diverse areas, is home to at least 62 marine fish species that are currently protected by federal/state regulations as well as those species that are under evaluation for protection by international conservation organizations, (IUCN and AFS), along with those species which we have suggested as potentially rare based on existing criteria (critically small population size) used by IUCN and AFS (Table 2). Except for two species, large-scale lizardfish and antenna codlet, potentially rare fish found principally in deeper benthic habitats (>100 ft.) are not included in the Tortugas List.

Of the 62 species, 47 show critically small population sizes associated with life history/habitat limitations while 15 have undergone population declines usually associated with overexploitation. Some listed species may be protected by all of the listed conservation agencies/organizations, while others will need further evaluation by fishery scientists from various federal/state fishery resource agencies and NGO's to determine extinction risk.

Table 9.- Draft list of tropical marine fish stocks at risk in the Dry Tortugas region.

Recruitment, Physics and the Parent Stock

Recruitment is defined as the addition of newborn to a stock each year. In the tropics, recruitment can occur over a protracted portion of the year (Ault 1988; Ault and Fox 1990). Spawning aggregations often bring together specific conditions of biological cycles, physical oceanography and habitat. A number of spawning aggregation sites have been identified in the Dry Tortugas region (Figure 38). These areas concentrate fish during the spawning season and serve as the source points for larvae which then drift advectively and then behaviorally until they become competent juveniles and settle to a take on a benthic existence. A suite of different species occupy the different spawning sites at different times. For example the snapper species, Lutjanus griseus, Lutjanus cyanopterus, Lutjanus analis, Ocyurus chrysurus, Lutjanus jocu, and Lutjanus apodus are all thought to use the Riley’s Hump area as a spawning site. It is critical to protect the integrity of the spawning sites and spawners during the reproductive periods of the year, and to protect the habitats critical to the survivorship of the settling juveniles.

Most tropical marine reef fishes of the Florida Keys and the Dry Tortugas Region have pelagic larvae that are dispersed by hydrodynamic currents driven by winds, tides and bathymetry. Recruitment of juveniles into a particular habitat or environment, e.g. the inshore coastal bays, nearshore barrier islands or the coral reefs tract, of this region are dependent upon the nature of the water flows. Interestingly, new evidence from physical oceanographers suggests gyre formations and diametric current reversals occur seasonally that facilitate the transport and retention of larvae to suitable settling areas (Figure 39). Spawning sites in the Dry Tortugas region were identified by commercial fishermen and others (Lindeman et al. 1999). Thus the probability of successful recruitment is a function of the size of the parent stock, the number of gravid fish spawning at a particular location, and the physical environment prevalent during the period of spawning and transport. In general, the biophysical processes involved in recruitment and survivorship of the larvae and juveniles is often the most poorly understood portion of the life history of reef fishes.

Figure 38

Figure 39

Relatively few studies of reef fishes in the Florida Keys have examined the recruitment and post-settlement of fish larvae near the Tortugas Region. Recent studies by Cha et al (1994) and Limouzy-Paris et al (1994) have examined the distribution and biodiversity of reef fish larvae from the Upper Florida Keys to Cosgrove Reef near the eastern boundary of TERSA (Marquesas Keys). Of the 68 families of reef fishes compiled by Starck (1968) at Alligator Reef in the Middle Keys, larvae of 43 families were collected in plankton tows from May 31 to June 5, 1989 (Limouzy-Paris et al. 1994). The nine most common reef fish families (most frequently occurring among stations, and in the top 10% in terms of abundance) were Paralichthyidae (flounders), Scombridae (mackerel/tunas), Gobiidae (gobys), Bregmacerotidae (codlets) Myctophidae (lanternfishes), Serranidae (seabasses), Carangidae (jacks), Bothidae (lefteye flounders).

Fishery Management and Marine Reserves

?Traditional fishery management approaches have often failed in complex tropical marine ecosystems due to the fact that reef fish communities are extremely vulnerable to fishing because of the large number of species, their complex biological community interactions and population-dynamic characteristics that include long lives, low adult natural mortality rates, sedentary adults, and high vulnerability to size-selective fishing gears. Ault (1996) proposed an integrated fishery management system for managing Gulf of Mexico fishery resources that is supported by recent Federal legislation to characterize 'essential fishery habitats' in all US Fishery Management Plans. Unfortunately, definitions of 'essential' and 'habitat' vary among biologists, ecologists, and managers. In the case of tropical coral reef fishes of economic and ecological value, these identifications are critical to the sustainability of the resources. Spawning and settlement areas of exploited tropical reef fishes (e.g., snappers, groupers) are essential fish habitats that are often spatially discrete. During spawning, many species aggregate on deep reefs to concentrate the sources of recruits and focus their settlement in particular inshore habitats. Identification and protection of spatially discrete habitats that facilitate ontogenetic migrations within much broader species ranges provides clear foci for habitat management and explicit criteria for the design of marine reserves.

We have developed a quantitative framework for addressing the dynamics and productivity of the Florida Keys multispecies reef fishery and have employed that approach in assisting the design and implementation of marine protected areas in the FKNMS. Our research is based on the premise that fish harvesting must be linked with the concept of ecosystem management, and that productive fisheries are a key indicator of a "healthy" coastal ocean environment. To do so we must take a systems view of the resources, emphasizing strategy over tactics in the development of a Florida Keys fishery management system. In developing models of reef fish community spatial dynamics, we are evaluating the space-time regime of natural and anthropogenic stressors that may affect economically and ecologically important fishes to define spatial management and restoration strategies to build sustainable reef fisheries. An important aspect of this work involves extensive use of scientific data visualization tools to let the 'unseeable’ to be seen. Because visualization technologies provide a high degree of functionality in sampling design, exploratory data analysis, model development and model validation, they will play an increasingly strategic role in fishery management science and research.

Acknowledgments

We thank Mario Alvarado, Peter Fischel, David McClellan, Colin Schmitz, Barry Wood and James Zweifel for their contributions to this report. We also want to express our sincere appreciation to those persons who graciously provided data: David Rydene and Joseph Kimmel of FMRI for reef fish visual census data; Jim Fourqueren of FIU for seagrass habitats; Joe Boyer of FIU for water quality; Eric Frankin and Mike Crosby at NOAA for benthic and coral habitats; Andy Beaver of NOAA for side-scan sonar data; Chris Friel and Frank Sargent at the Florida Department of Environmental Protection for habitat data from aerial surveys; Chris Glendhill of NMFS Pascagoula, Mississippi for bathymetry and reef census data; Rod Bertleson of FMRI Marathon for lobster surveys; and, Ken Lindeman and Don DeMaria for information on reef fish spawning locations. Preparation of this document was partially funded by the U.S. Man and the Biosphere Marine and Coastal Ecosystems Directorate, NOAA Coastal Ocean Program, National Park Service and the Florida Keys National Marine Sanctuary.