2073_C012.fm Page 219 Friday, April 7, 2006 5:00 PM 12 Lessons Learned from Experimental Key-Species Restoration Margaret W Miller and Alina M Szmant CONTENTS 12.1 12.2 Introduction .219 Sexual Propagation and Seeding of Broadcast-Spawning Corals .220 12.2.1 Acropora palmata 221 12.2.2 Montastraea spp .223 12.2.3 Larval Culture 223 12.2.4 Seeding Efforts 227 12.3 Restocking of Diadema antillarum in the Florida Keys 227 12.4 Conclusions .230 Acknowledgments 231 References 232 12.1 INTRODUCTION The mortality of reef-building corals, particularly in the Caribbean, has been a persistent phenomenon over the past two decades1 despite increasing implementation of conservation measures, including the establishment of protected reef areas.2 Increasing conservation efforts, therefore, have coincided with practically monotonic reef decline Conservation measures alone thus seem insufficient to prevent or reverse the decline in coral cover We suggest that restoration is a necessary next step to rebuild coral populations and reestablish reef resiliency As large reef-building corals have died, new ones have failed to replace them Where coral recruitment has been noted, it is generally of the smaller brooding coral species that are minor contributors to reef framework construction There is evidence of dramatic decrease in overall coral recruitment success in the Caribbean,3 suggesting a generalized loss of resilience of Caribbean reef systems.4 Several interacting factors likely contribute to this recruitment failure in reef-building corals As live coral cover declines by the loss of adult corals, and/or adult corals are stressed and have reduced fecundity, the areal production of coral gametes decreases,5 which can increase the likelihood of fertilization limitation in broadcast-spawning species (the so-called Allee effect) Fertilization success has also been found to be low when the spawning corals have been recently bleached.6 Thus, a smaller supply of larvae will be available to settle onto reefs, and recruitment may be supply-side limited There is concern, but few data to support the possibility, that exposure to xenobiotics or other aspects of poor water quality may compromise reproductive effort and effectiveness.7 Lastly, Caribbean reefs have manifested dramatic changes in substrate quality following 219 © 2006 by Taylor & Francis Group, LLC 2073_C012.fm Page 220 Friday, April 7, 2006 5:00 PM 220 Coral Reef Restoration Handbook the mass mortality of the important grazing sea-urchin, Diadema antillarum, in 1982 to 1983,8 namely, the increase of macroalgal cover and loss of substrate types associated with intense grazing (e.g., live corals, crustose coralline algae [CCA] and fine turfs) Some corals exhibit a requirement for certain types of CCA as substrate cues to induce settlement;9–11 therefore, loss of CCA cover may further limit coral recruitment rates Recent observations of increased coral recruitment (though mostly small brooders, not broadcast-spawning, reef-building species) in patchy areas where Diadema have for the most part recovered12 indicate that this so-called “phase shift” from coral to macroalgal dominance may be reversible and suggest that Diadema grazing may be an important aspect of reef recovery and resilience In responding to acute coral loss, such as that caused by anthropogenic physical disturbances (e.g., ship groundings), most restoration efforts have focused on transplantation of coral colonies or fragments from adjacent reefs,13 although the net ecological benefit of transplantation has been questioned.14 More recently, response to acute disturbances has included rescue and reattachment of impacted corals.13,15 Active propagation of coral fragments to build source populations for transplant efforts is also under study.16 However, these strategies are particularly compromised in the Caribbean by a dearth of healthy donor populations and a species pool which is depauperate in fast-growing, branching species that lend themselves to asexual propagation and transplantation In fact, both of the main fast-growing branching corals in the Caribbean, Acropora cervicornis and A palmata, are in the process of being listed as threatened under the U.S Endangered Species Act (Federal Register, 50 CFR Part 223) Given the clear need for intervention to help speed up the restoration of coral abundance on Caribbean reefs, the practical limitations of asexual propagation and transplantation in this region, and the clear importance of the grazing urchin D antillarum in maintaining substrate quality suitable for successful coral recruitment, we have chosen to pursue experimental restoration approaches that include restocking Diadema to restore substrate quality, followed by seeding of restored substrate with sexual propagules (planula larvae) of reef-building coral species 12.2 SEXUAL PROPAGATION AND SEEDING OF BROADCAST-SPAWNING CORALS Scleractinian corals display two broad categories of life history strategies (Figure 12.1) One includes opportunistic species with generally small colony size, which have internal fertilization and brood their larvae to a fairly advanced stage and release planula larvae that can settle and metamorphose shortly after release from the mother colony (Cycle II in Figure 12.1) These species, known as brooders, generally recruit quite effectively The second broad category includes the majority of coral species that spawn their gametes into the water column Known as broadcastspawners, these species undergo fertilization and a week or more of larval development in the water column, subject to its dilution, currents, and predators The potential for fertilization limitation (especially in populations with low adult density), together with advection away from reef areas, and expected high mortality for planktonic larvae, could lead to poor recruitment by these species Broadcasting species are the corals that tend to attain large colony size and thus, are responsible for most reef accretion In the Indo-Pacific region, broadcasters achieve high rates of recruitment, even in recent years.17–20 By contrast, most of the major broadcasting reef-builders in the Caribbean are infrequently observed in recruitment studies.21,22 The main justification for our interest in developing a methodology for larval seeding of coralpoor reef substrates is based on the premise that a major bottleneck in coral life history and, therefore, a major aspect of recruitment failure of reef-building species, lies in the presettlement phase, specifically insufficient larval supply (for all the reasons described above) Therefore, if a large number of viable larvae can be introduced to a substrate in need of restoration, coral recruitment of reef-building species can be enhanced In our approach, fertilization success and larval survival are increased by collecting and maintaining high gamete concentrations during the © 2006 by Taylor & Francis Group, LLC 2073_C012.fm Page 221 Friday, April 7, 2006 5:00 PM Lessons Learned from Experimental Key-Species Restoration 221 Pattern I Planulae spend days to weeks in plankton Broadcast spawners 1–2 cycles per year many gametes released Brooders: Many cycles few larvae per cycle Pattern II Hours-days to settlement to years to reproduction Decades to reproduction FIGURE 12.1 Two major reproductive cycles of Scleractinian corals: I Broadcasting of gametes or gamete bundles into the water column where fertilization and larval development take place This pattern is generally associated with a long-lived, large-colony-size life history strategy; II Brooding of gametes, where fertilization is presumably internal, and embryos are retained for weeks to months until they develop to a fully competent planula larva stage Our restoration efforts have focussed on Pattern I species fertilization phase and avoiding the high natural mortality expected in the planktonic phase Based on two decades of work,23–25 we can predict the spawning nights within a narrow window Thus, we deploy teams of divers to collect broadcast-spawned gametes as they are released, fertilize them in buckets or coolers, culture them through the planktonic phase to competency under protected conditions (either in the laboratory or in field enclosures), and settle them onto experimental tiles or directly onto natural or “restored” reef substrates Our efforts have focused on the two most important (and poorly recruiting) reef-building corals in the Caribbean, Montastraea annularis complex and Acropora palmata 12.2.1 ACROPORA PALMATA Acropora palmata, along with its congener A cervicornis, has been responsible for much of the structural accretion of Caribbean reefs in the Holocene.26 Both species have branching morphologies and high growth rates and were formerly highly abundant However, these species have undergone such drastic decline over the past two decades that they are in process of being listed as threatened under the U.S Endangered Species Act Disease, bleaching, and various sources of physical disturbance have been instrumental in this decline.27,28 At the present time, most remaining populations of A palmata are small, consisting largely of isolated colonies Because A palmata often propagates by fragmentation, some of these populations have extremely limited genotypic diversity This species requires out-crossing for successful fertilization;24 populations consisting of a single clone will thus not produce any larvae These factors of rarity, isolation, and limited genotypic diversity suggest that its potential for successful sexual reproduction may be at present limited The same considerations apply to A cervicornis We have focused our efforts in the Florida Keys on A palmata because A cervicornis is so rare that no adult source populations were available In the Florida Keys, A palmata is predicted to spawn in the range of three to five nights after the full moon in August or September However, the first challenge to working with this species is its unreliability in spawning time and synchrony (summarized in Table 12.1) In some years © 2006 by Taylor & Francis Group, LLC 2073_C012.fm Page 222 Friday, April 7, 2006 5:00 PM 222 Coral Reef Restoration Handbook TABLE 12.1 Recent Spawning Observations for Acropora palmata in Key Largo, Florida Year 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Site Key Largo Dry Rocks Ball Buoy Reef, BNP; Key Largo area; others Key Largo Dry Rocks Key Largo Dry Rocks Horseshoe Horseshoe, Florida and Atravesado, Puerto Rico Horseshoe Sand Island and French Sand Island Horseshoe Little Grecian Horseshoe, Little Grecian, Key Largo Dry Rocks (KLDR) Date and Timing None on August 25 to 26 (5 to nights AFM): watched from 2130 to 2300 August 14 to 18 (4 to nights AFM); 2150 to 2230; a few colonies spawned August to (4 to nights AFM); 2145 to 2230 September to (4 to nights AFM); 2150 to 2230 August 20 to 21 (2 to nights AFM) August 11 (3 nights AFM); 2150 to 2245 No observations None on August to (3 to nights AFM) August (5 nights AFM); 2233 to 2300 None on August 24 to 27 (2 to nights AFM) July 21 (9 nights AFM) 3/60 colonies on Sand Island (SI) August 17 (5 nights AFM) 8/60 colonies on SI; 50% on Horseshoe and Little Grecian; 2200 to 2230 Heavy spawn (>80% of colonies) at Horseshoe and Little Grecian (fewer at KLDR) on Aug (4 nights AFM); 2200 to 2230 Note: AFM = After full moon; BNP = Biscayne National Park (e.g., 2002), no spawning was observed throughout the entire predicted window In 2003 we had observations for the predicted window for both July and August One colony at Sand Island reef, Key Largo, released some spawn on nights and with three additional colonies spawning on night after the full moon in July (I Baums, personal communication) In August, the main spawn occurred on night with about 50% of colonies at Horseshoe and Little Grecian reefs spawning but only eight of 60 colonies spawning at Sand Island At least one large colony at Sand Island had ripe gametes based on histological examination earlier in July (I Baums and S Colley, personal communication) but was not observed to spawn, suggesting that additional asynchronous spawning occurred on nights we were not present to observe Besides timing, another challenge in culturing A palmata larvae has been to collect adequate amounts of spawn to work with It is currently quite rare, and colony condition is poor in many areas Up until 1998, we had successfully collected large quantities of viable A palmata spawn at Key Largo Dry Rocks, which had extensive thickets of this species The site also has mooring buoys and is shallow and somewhat protected, making nighttime field operations feasible there under most weather conditions The A palmata stand at Key Largo Dry Rocks was decimated in 1998 by severe bleaching and Hurricane Georges Thus, in 1999 to 2001 we shifted our collecting efforts to nearby Horseshoe Reef, where a robust and fairly dense A palmata stand persisted We were able to make large collections in 2000 and 2001, but fertilization was poor and no larvae resulted New information on A palmata population genetics became available in 2002 The extensive A palmata thicket at Horseshoe Reef is in fact monotypic, comprised of a single genetic individual.29 © 2006 by Taylor & Francis Group, LLC 2073_C012.fm Page 223 Friday, April 7, 2006 5:00 PM Lessons Learned from Experimental Key-Species Restoration 223 This hermaphroditic species cannot self-fertilize (Szmant24 and later observations) so it is no surprise that our collections in 2000 and 2001 failed to produce larvae In subsequent years, we have doubled our logistical burden to collect spawn at two different sites in order to assure outcrossing During the summers of 2003 and 2004, we were able to raise several thousand larvae of A palmata, some of which were used in laboratory experiments,30 with the remainder seeded onto substrates that were deployed at the Wellwood grounding site (see Section 12.2.4 below) 12.2.2 MONTASTRAEA SPP The reef coral M faveolata is another primary reef-building coral in the Florida Keys and throughout the Caribbean, often occurring as enormous colonies several meters high Many of these colonies are hundreds of years old and have large cavities that are important shelter and substrate for reef fishes and invertebrates There are also large numbers of mid-sized colonies m or so in diameter, but smaller (