While encrusting and excavating sponges of the genus Cliona are important bioeroders on Caribbean reefs, relatively little information exists on large-scale distributional patterns and environmental correlates. A survey encompassing 181 Florida Keys coral reef and hard-bottom sites sampled C. delitrix Pang density and size to explore relationships with habitat-related characteristics such as depth, distance from shore, and coral density and cover. Nine habitats were surveyed across the continental shelf representing a range in depth (1-27 m), cross-shelf position, topographic complexity, and coral abundance. A stratified random sampling design used replicate 8-m x 1-m transects per site to survey 2,896 m2 of coral reef and hard-bottom habitat. Sixty of the 181 sites yielded 189 individuals of C. delitrix. The distribution of C. delitrix was not proportional to the sampling effort, with four of the nine habitats accounting for ~83% of all sightings. Mean sponge density (no. per m2) ranged from 0.01 to 0.24 and differed significantly among habitats. Shallower (< 6 m) and more wave-exposed habitats on the platform margin had statistically lower C. delitrix densities, while densities on the deeper fore reef slope were up to 18 times greater. Cliona delitrix area per m2 and mean size were similarly patterned to density, with significantly greater mean values on patch reef and deeper (> 10 m) fore reef slope habitats. Regression analysis indicated that greater C. delitrix densities were found at greater depths, while larger C. delitrix sponges were found closer to shore in areas of higher coral abundance. Patterns of coral colonization relative to coral availability suggest that C. delitrix occupied ~23% of the coral taxa preferentially relative to availability, especially coral species confined to relatively few habitats or those most abundant on patch reefs or the deeper fore reef slope.

The 1983-1984 caribbean-wide mass mortality of the once ubiquitous long-spined sea urchin Diadema antillarum Philippi, 1845, is one of several factors considered responsible for coral reef change throughout the region. Unfortunately, there is a paucity of pre-mortality event density data for D. antillarum in the Florida Keys, making it difficult to determine pre-1983 population density levels. results from surveys conducted during 1970-1973 in the lower Florida Keys, in shallow (<12 m) fore reef habitats, yielded relatively abundant and widespread D. antillarum densities in qualitative transects at five reefs prior to the 1983-1984 die-off. In quantitative surveys at one reef, Middle sambo reef in 1972, up to 7.9 individuals m-2 were recorded using quadrats in high-relief spur and groove habitat. A second mortality event in the Florida Keys, beginning in April 1991, again depressed urchin densities that had begun to recover from the 1983-1984 mass mortality. by 1992, D. antillarum densities (<0.01 m-2) were two orders of magnitude lower than pre-die-off estimates (range of 0.07-0.57 m-2 from several spur and groove reefs in the lower Florida Keys) and remained so through 2009. The pre-mortality echinoid density estimates detailed in the Florida Keys provide a baseline to compare with their current population status and should help inform managers about realistic recovery or restoration targets for D. antillarum.

Principles of probability survey design were applied to guide large-scale sampling of populations of stony corals and associated benthic taxa in the Florida Keys coral reef ecosystem. The survey employed a two-stage stratified random sampling design that partitioned the 251-km2 domain by reef habitat types, geographic regions, and management zones. Estimates of the coefficient of variation (ratio of standard error to the mean) for stony coral population density and abundance ranged from 7% to 12% for four of six principal species. These levels of survey precision are among the highest reported for comparable surveys of marine species. Relatively precise estimates were also obtained for octocoral density, sponge frequency of occurrence, and benthic cover of algae and invertebrates. Probabilistic survey design techniques provided a robust framework for estimating population-level metrics and optimizing sampling efficiency.

BACKGROUND:The rising temperature of the world's oceans has become a major threat to coral reefs globally as the severity and frequency of mass coral bleaching and mortality events increase. In 2005, high ocean temperatures in the tropical Atlantic and Caribbean resulted in the most severe bleaching event ever recorded in the basin.METHODOLOGY/PRINCIPAL FINDINGS:Satellite-based tools provided warnings for coral reef managers and scientists, guiding both the timing and location of researchers' field observations as anomalously warm conditions developed and spread across the greater Caribbean region from June to October 2005. Field surveys of bleaching and mortality exceeded prior efforts in detail and extent, and provided a new standard for documenting the effects of bleaching and for testing nowcast and forecast products. Collaborators from 22 countries undertook the most comprehensive documentation of basin-scale bleaching to date and found that over 80% of corals bleached and over 40% died at many sites. The most severe bleaching coincided with waters nearest a western Atlantic warm pool that was centered off the northern end of the Lesser Antilles.CONCLUSIONS/SIGNIFICANCE:Thermal stress during the 2005 event exceeded any observed from the Caribbean in the prior 20 years, and regionally-averaged temperatures were the warmest in over 150 years. Comparison of satellite data against field surveys demonstrated a significant predictive relationship between accumulated heat stress (measured using NOAA Coral Reef Watch's Degree Heating Weeks) and bleaching intensity. This severe, widespread bleaching and mortality will undoubtedly have long-term consequences for reef ecosystems and suggests a troubled future for tropical marine ecosystems under a warming climate.

The 1983-84 Caribbean-wide mortality of the urchin Diadema antillarum Philippi was followed by a second mortality event in the Florida Keys in 1991. The demise of this once ubiquitous herbivore is one factor contributing to wider Caribbean reef change during the past 25 years. Over an 8-year period from 1999-2007, we examined densities and test sizes of D. antillarum at 786 sites from the northern extent of the Florida Reef Tract to the Dry Tortugas, including two National Parks and the Florida Keys National Marine Sanctuary. Visual surveys along belt transects were used to enumerate individuals and test sizes in a two-stage stratified random sampling design that incorporated cross-shelf benthic habitats, geographic regions, and no-fishing management zones. While pre-1983 densities were as high as 5 individuals/m 2 , surveys since 1999 from < 1 m to 27 m depth show that current densities are still well below 1 individual/m 2 . During seven different annual sampling periods, the maximum site-level density was only 0.33 individuals/m 2 , with the highest densities of larger (> 5 cm test diameter) individuals reported from only a few locations. The relative contributions of larval survivorship, predation, suitable recruitment sites, and reduced fertilization success to Diadema recovery are still largely unknown.

A one-year experiment was conducted to determine the efficacy of urchin translocations and resulting benthic community effects on Florida Keys patch reefs. Small (1-1.5 cm test diameter) long-spined sea urchins (Diadema antillarum) were collected from back reef rubble zones and transported to two experimental patch reefs during September-December 2001. Changes to community structure were assessed on two experimental and two control patch reefs prior to and one year after the urchin translocation, including percent cover, sponge and cnidarian species richness, and juvenile coral density. Urchin densities on the experimental patch reefs one year after the translocation averaged nearly 1 individual/m 2 , similar to urchin density estimates in the Florida Keys prior to the 1983-84 mass mortality event. The coverage of stony corals and crustose coralline algae increased, while the coverage of brown foliose algae declined on experimental patch reefs. In contrast, stony coral and crustose coralline algal cover declined on control patch reefs, but increased for brown foliose algae. Juvenile coral densities increased at all sites, but density increases were markedly greater on both experimental sites, reflecting greater densities of smaller juveniles (< 1.5 cm diameter), especially Porites astreoides and Siderastrea siderea. Greater juvenile densities on experimental reefs may have resulted from more available space for settlement, lower post-settlement mortality from algal overgrowth, or enhanced settlement sites due to increased coverage of crustose coralline algae compared to control reefs. These results are similar to previous investigations of the effects of artificially enhanced or naturally recovering urchin densities on coral reef benthos, especially as they pertain to changes in algal composition and juvenile coral densities. However, other factors, such as storms that frequented the area during the study, are also possible contributors to the temporal patterns documented. Future surveys will monitor the survivorship of the resident adult urchins on experimental reefs and additional changes to benthic community structure that may occur.

Despite a long history of intensive fishing, information on the spatial extent and biological impacts of fishing gear is lacking in the Florida Keys. We studied spatial distribution, density, and length of lost fishing gear and other non-fishing-related debris at 63 shallow-water (< 8 m) sites. The sites comprised high-relief spur and groove and low-relief hard-bottom habitats; three geographic regions; and three types of management areas: open to fishing, restricted to catch and release fishing by trolling only, and no fishing. Three-hundred pieces of lost fishing gear and other debris were removed from 25,200 m(2) of benthic habitat. Lost hook-and-line gear was the most prominent debris (87%). No significant differences in mean debris densities were detected between habitats studied or among geographic regions. Mean densities of lost hook-and-line gear, lobster trap gear, and total debris were similar among the three management area types in high-relief spur and groove, while lost hook-and-line gear and total debris were significantly greater in no-fishing zones compared to fished areas in low-relief hard-bottom. In designated no-fishing zones, lost fishing was spatially pervasive and comprised the majority of marine debris in the habitats surveyed. Some of the lost fishing gear was probably present before the designation of no-fishing zones in 1997; the preponderance of lost gear in these areas may indicate that they attract anglers. Monitoring of lost fishing gear can help to assess compliance and biological impacts in the Florida Keys and the patterns documented highlight the challenge to patrolling a large marine protected area.

The Florida Keys coral reef ecosystem supports multimillion-dollar commercial and recreational fisheries. The ecological effects caused by fishing gear that is lost when cut or broken after snagging on the bottom is a growing concern to managers and scientists. Few data exist, however, to assess the impacts of lost fishing gear to benthic organisms and habitat structure. In this study, 63 offshore coral reef and hard-bottom sites were surveyed during 2001 to quantify the impacts of lost fishing gear to coral reef sessile invertebrates. Lost hook-and-line fishing gear accounted for 87% of all debris (N=298 incidences) encountered and was responsible for 84% of the 321 documented impacts to sponges and benthic cnidarians, predominantly consisting of tissue abrasion causing partial individual or colony mortality. Branching gorgonians (Octocorallia) were the most frequently affected (56%), followed by milleporid hydrocorals (19%) and sponges (13%). Factors affecting the impacts of lost fishing gear include sessile invertebrate density, the density of lost fishing gear, and gear length. While lost hook-and-line fishing gear is ubiquitous in the Florida Keys, less than 0.2% of the available milleporid hydrocorals, stony corals, and gorgonians in the habitats studied are adversely affected in terms of colony abrasions and partial mortality.

A large-scale assessment of the density and gorgonian host-occupation patterns of the ovulid gastropod Cyphoma gibbosum L. was conducted at 63 shallow (< 7 m), low-relief hard-bottom and spur and groove sites in the Florida Keys. Mean densities of C. gibbosum were not significantly different between habitat types, among geographic regions, or between fished and protected areas. Based on sample allocation analyses, spur and groove sites open to fishing in the lower Keys region yielded more C. gibbosum than expected. Most C. gibbosum occurred individually (58 %) or in pairs (31 %) on gorgonians, indicating more gregariousness than expected based on random distribution. Snails were counted on 127 gorgonians repre- senting 3 families and 12 species. Significant differences in gorgonian host-occupation were detected, with more Plexauridae and less Gorgoniidae occupied than expected. Eunicea tourneforti, Plexaura flexuosa, P. homomalla, and Pseudoplexaura porosa were occupied more than expected, while Gorgonia ventalina was occupied in proportion to its abundance. Pseudopterogorgia americana and other gorgonians were under- occupied, even when relatively abundant. Except for P. homomalla, C. gibbosum density was only weakly correlated with total gorgonian density and individual species densities. Spatial variations in C. gibbosum density are not readily explained by preferred gorgonian availability alone. While variable larval recruit- ment, post-settlement survival, and patchy gorgonian distribution may affect these patterns, differential predation pressure by fishes such as hogfish (Lachnolaimus maximus) could also partially explain the dis- tribution patterns of C. gibbosum in the Florida Keys.

Concern about declining trends in coral reef habitats and reef fish stocks in the Florida Keys contributed to the implementation of a network of no-take marine protected areas in 1997. In support of the efforts of the Dry Tortugas National Park and Florida Keys National Marine Sanctuary to implement additional no-take areas in the Tortugas region in 2001, we expanded the scale of our fisheries independent monitoring program for coral reef fishes in the region. To provide a foundation for the habitat-based, stratified random sampling design of the program, we created a digital benthic habitat map of coral reef and hard-bottom habitats in a geographic information system by synthesizing data from bathymetric surveys, side-scan sonar imagery, aerial photogrammetry, existing habitat maps, and in situ visual surveys. Existing habitat maps prior to 1999 were limited to shallow-water (< 20 m depth) soft-sediment, coral reef, and hard-bottom habitats within Dry Tortugas National Park and did not include deeper areas such as the Tortugas Bank, now partially contained within no-take marine protected area boundaries. From diver observations made during the 1999 survey, we developed a classification scheme based on habitat relief and patchiness to describe nine hard-bottom and coral reef habitats encountered from 1-33 m depth. We provide estimates of area by habitat type for no-take marine protected areas in the Tortugas region. Updated information on the spatial distribution and characteristics of benthic habitats will be used to guide future monitoring, assessment, and management activities in the region. Significant data gaps still exist for the western area of the Florida Keys National Marine Sanctuary and are a priority for future research.

Despite representing the northern extent of Acropora spp. range in the Caribbean, most of the Florida reef line from Palm Beach through the Keys was built by these species. Climatic factors appear to have been important agents of Acropora loss within historic (century) time frames. In the recent past (1980-present), available quantitative evidence suggests dramatic declines occurred in A. cervicornis first (late 70's to 84) with collapse of A. palmata occurring later (1981-86). However, recent monitoring studies (1996-2001) show continued decline of remnant populations of A. palmata. Current trends in A. cervicornis in the Florida Keys are hard to assess given its exceedingly low abundance, except in Broward County, FL where recently discovered A. cervicornis thickets are thriving. While the State of Florida recognizes A. palmata and A. cervicornis as endangered species (Deyrup and Franz 1994), this designation carries no management implications. The current management plan of the FKNMS provides many strategies for coral conservation, among them minimizing the threat of vessel groundings and anchor damage, and prohibitions on collection, touching, and damage from fishery and recreational users. Although Acropora spp. are not explicitly given any special consideration, they are implicitly by Sanctuary management. Restoration approaches undertaken in the Florida Keys include rescue of fragments damaged by groundings and experimental work to culture broadcast-spawned larvae to re-seed natural substrates. Neither of these efforts have yet realized full success. Geological history Prior to the most recent moderate sea-level phase, Florida reef development proceeded under high sea- level conditions in the absence of the sensitive Acropora spp. These species were absent in Florida due to the inimical effects of Gulf waters flowing unimpeded from the shelf (now Florida Bay) over the reef tract. Slow-growing head corals built the Pleistocene Florida reefs. However, the spur-and-groove reef structures which we observe in the Florida Keys today as well as for the three-reef system from Palm Beach to northern Miami-Dade County are all constructions of Acropora palmata (Lighty 1977, Shinn 1988). The rapid growth of this species has allowed this impressive accretion on a short geological time frame, the last 6-7K years when sea level has been low. Early Holocene conditions were perhaps best conducive to rapid Acroporid colonization, but rising sea levels between 5 and 3 thousand years ago led to the demise of the reef system north of Fowey Rocks, and the die-off of Acroporid reef flats due to flooding and formation of Florida Bay. Thus, although geological evidence suggests that coral reef formation has occurred in the absence of Acropora spp. in the distant past under high sea-level conditions, it is clear that future functional absence of these species will severely compromise the ability of Florida reefs to survive anticipated sea level rise (i.e., their ability to "keep up") in the not-too-distant future. Long-term trends (100 yr)

Fishing constitutes one of the most significant threats to marine biodiversity and ecosystem function, documented by a growing body of information on the numerous impacts to populations, community structure, and habitats (Dayton et al., 1995; Roberts, 1995; Jennings and Polunin, 1996). Besides the more obvious effects on species population structure, fishing activities may also reduce the structural complexity of habitats or cause corresponding changes in ecological processes such as competition and predation (Russ, 1991; Jones and Syms, 1998; Auster and Langton, 1999). These patterns are most obvious in areas where explosives, poisons, or other destructive fishing methods are used (Hatcher et al., 1989). However, ecological effects can be expected in any area where traps, mobile fishing gear such as trawls, and, potentially, even large numbers of recreational fishers operate (Russ, 1991; Jennings and Lock, 1996). The Florida Keys (Monroe County, Florida) have a long history of commercial and recreational fisheries that target a great diversity of fish and invertebrate species using a multitude of gears (Tilmant, 1989; Bohnsack et al., 1994). In terms of volume of seafood landed, the Florida Keys is the most important area in the state in landings, dockside value, and numbers of commercial fishing vessels, especially for highly valued invertebrate fisheries (Adams, 1992). There are also significant, but largely undocumented effects of tens of thousands of recreational fishers (Davis, 1977), who target hundreds of species using mostly hook-and-line and spear guns (Bohnsack et al., 1994). Baseline data on fishing gear and other marine debris were collected as part of a larger assessment of benthic community structure in the Florida Keys National Marine Sanctuary, a large (9500 km) marine protected area bordering three national parks in southern Florida (Fig. 1). These data are particularly timely because this coastal ecosystem continues to experience a growing number of recreational fishers, and both commercial and recreational fishers exploit hundreds of invertebrates and fish species (Bohnsack et al., 1994; Ault et al., 1998). This study addressed several issues on marine debris occurrence in shallow-water coral reef and hard-bottom habitats. First, what is the spatial extent and frequency of remnant fishing gear at multiple spatial scales in the Florida Keys? Secondly, what factors, such as habitat type (depth) or management regime (closed or open to fishing) affect the spatial variability of marine debris occurrence? Thirdly, what are the biological impacts of marine debris, especially from remnant commercial and recreational fishing gear, on reef biota such as hard corals and sponges? Forty-five sites were surveyed southwest of Key West to Big Pine Shoal in the lower Keys region of the Sanctuary, spanning 60 km from southwest to northeast and 12 km from nearshore to offshore (Fig. 1). Sites were visited between July and August 2000 and were selected using a two-stage, stratified random sampling design (Cochran, 1977; Ault et al., 1999). Five of the 23 no-fishing zones (designated as sanctuary preservation areas (SPAs), research only areas (RO), and ecological reserves (ER)) in the Sanctuary were surveyed (indicated in Fig. 1). Based on the spatial distribution of coral reef habitat types (FDEP, 1998) and the depth limits of the zones, the following habitat strata were sampled: nearshore hard-bottom, mid-channel patch reef, offshore isolated patch reef, offshore aggregate patch reef, back reef and rubble/hard-bottom matrix, shallow fore reef (4–7 m depth), and deeper fore reef (8–12 m) (Table 1). Two random sites were sampled in each no-fishing zone, within a particular habitat stratum that consisted of predesignated 200 200 m areas randomly selected from a grid constructed using a geographic information system. Corresponding author.