Reference: Biol. Bull. 211: 149-156. (October 2006) ? 2006 Marine Biological Laboratory Caribbean Placozoan Phylogeography ANA Y. SIGNOROVITCH1*, STEPHEN L. DELLAPORTA2, AND LEO W. BUSS1'3 Yale University, l Department of Ecology and Evolutionary Biology, 2Department of Molecular, Cellular, and Developmental Biology, and ^Department of Geology and Geophysics, 165 Prospect Street, OML, New Haven, Connecticut 06511 Abstract. We here address placozoan distribution and phylogeography in five locations in the Caribbean Sea. We performed a coarse-resolution presence/absence survey of placozoans in Belize, Bermuda, Grenada, Jamaica, and Pan- ama and a fine-resolution study of the distribution of pla- cozoans in Twin Cays, Belize. Placozoans were recovered in every country sampled. Animals were sequenced at the mitochondrial 16S rDNA locus, and our analysis identified four of the five previously identified clades present in the Caribbean. In addition, we discovered two new haplotypes within one of these clades, and we found sympatric clades in Belize, Bermuda, Jamaica, and Panama. These studies provide further molecular evidence for species diversity within the Phylum Placozoa. Introduction Since its discovery, Trichoplax adhaerens has been clas- sified as the sole species within the phylum Placozoa (Grell, 1971a), purported to reproduce sexually (Grell, 1971b, 1972), and discovered across the world. Today we know that the Placozoa actually consists of several clades based on 16S rDNA, LSU, SSU, and ITS genotyping (Voigt et al, 2004) and that sexual reproduction is indeed part of the life cycle in at least one of these clades (Signorovitch et al, 2005). While significant progress continues to be made in the field of placozoan biology?as exemplified by the pla- cozoan genome sequencing project presently underway? the phylogeography of this group has remained poorly un- derstood. The first placozoan record dates from 1883 when zoolo- gist F. E. Schulze discovered the microscopic amoeba-like Received 8 March 2006; accepted 11 July 2006. * To whom correspondence should be addressed at Harvard University, Department of Molecular and Cellular Biology, Cambridge, MA 02138. E-mail: asignorovitch@mcb.harvard.edu animals in an aquarium tank containing seawater from the Gulf of Trieste in the Mediterranean Sea. Schulze (1883) aptly named this animal Trichoplax adhaerens for its flat- tened and ciliated body that held fast to the aquarium wall. Because T. adhaerens lacked any definite outline or organs and possessed a uniquely simple yet highly integrated body plan, Schulze declared it the most primitive of all metazo- ans. Almost a century of stasis elapsed before further in- vestigations on this simple animal resumed. The German protozoologist K. G. Grell (1971a) rediscovered T. adhae- rens while examining an algal sample from the Red Sea near Eilat, Israel. By stably maintaining T. adhaerens on a diet of Pyrenomonas salina (Cryptophyceae), Grell and Benwitz (1971) observed in detail its histological organiza- tion. T. adhaerens lacks a basement membrane and has only four somatic cell types. Because of its unique body plan and histological organization, Grell established the phylum Pla- cozoa (Grell, 1971a) to house this singular species. Since the resurgence of interest in placozoan biology in the 1970s, these animals have been discovered in numerous marine aquaria (Miller, 1971; Ivanov etal, 1980; Martinelli and Spring, 2003; Voigt et al, 2004) and seawater systems (Pearse, 1989; Maruyama, 2004), and in the wild where they have been sampled from tropical and subtropical latitudes (Maruyama, 2004; Voigt et al, 2004). In 1980, Ivanov and colleagues discovered placozoans in a marine aquarium in the Moscow Zoo, Russia (Ivanov et al, 1980). They noted slight morphological differences in the so-called "brown bodies" between the Moscow and Red Sea isolates (refer- enced in Aleoshin et al, 2004), and for this reason, the Moscow isolate was named Trichoplax sp. instead of Trichoplax adhaerens. In 1987, Grell and Lopez-Ochoterena (p. 255) discovered placozoans in what they referred to as "a small underground seawater deposit" near a dock in Quin- tana Roo, Mexico. Although the authors were unable to culture these placozoans with the usual Pyrenomonas food 149 150 A. Y. SIGNOROVITCH ET Ah. source, they were successful with the green alga Chlorella. This dietary preference led them to hypothesize that the animals sampled in Mexico belonged to a different group of placozoans from the established Red Sea strain. In 1989, V. B. Pearse reported on the first record of placozoans in Honolulu, Hawaii. These animals were dis- covered on glass microscope slides that had been submersed in running seawater tanks for 2 to 3 weeks. A thin film of bacteria, algae, or both, and an assemblage of microorgan- isms such as foraminiferans, sponges, hydroids, ascidians, copepods, and nematodes characterized these slides. Pearse suggested that these placozoans most likely belonged to the only described species, T. adhaerens, because the morpho- logical features and the growth rate of the placozoans found in Hawaii were consistent with previous observations by Grell (1971a) and measurements reported by Ruthmann (1977). Placozoans have also been documented at widely distrib- uted sites in the Caribbean and throughout the tropical Pacific. The most thorough field study of placozoans to date is a survey on the north coast of Papua New Guinea (V. B. Pearse, UC Santa Cruz, pers. comm., 2005). The survey aimed specifically at understanding the distribution and abundance of placozoans in a variety of environments rang- ing from a laboratory seawater system, to rubble, man- groves, and reefs. Glass microscope slides used as settling plates were either attached to solid substrata or suspended in the water column for 2 to 3 weeks before retrieval and screening. Placozoans were found to be widely distributed among these sampling locations, with the exception of sandy or rubble substrates, which were devoid of placozo- ans. As in the Hawaiian study, placozoans were found associated with the now familiar thin film of bacteria, algal crusts and filaments, ciliates, polychaete worms, calcareous sponges, and hydroids. Like Grell and Lopez-Ochoterena (1987), Pearse also found survival and growth variability among the New Guinea isolates. The longest field study of placozoans was based in Shira- hama, Japan, from 1989 through 2000 (Maruyama, 2004). Placozoans were sampled monthly, using glass microscope slides, from sites about 1 m below the mean tidal level. Natural substrata were also sampled, such as stones, shells, and hard corals. A running seawater tank system containing hard corals was also set up in the laboratory to sample for placozoans. The 11 years of continuous sampling show that placozoans are present year-round on both the slides and substrate material at the field site and on the laboratory slides in the running seawater system. Variation in the abundance of placozoans sampled within each year was reflected in a sharp increase during the late summer and early winter. In a seminal paper by Voigt et al. (2004), DNA sequences from placozoans collected around the world were compared at four loci (SSU, LSU, ITS, and 16S rDNA) to construct a phylogeny of this group. From these studies, the investiga- tors concluded that the Placozoa is composed of at least five highly divergent clades. Thus, although the past records on the occurrence and distribution of placozoans do contribute significantly to our knowledge of the ecology of placozoans, the genetic information necessary to understand the patterns in biogeography of these clades has only just begun to surface from molecular diversity analyses. The work described here aimed at understanding placo- zoan phylogeography with a focus on the Caribbean Sea. Specifically, we report on a molecular survey of partial 16S ribosomal DNA sequences from 64 samples encompassing five Caribbean nations and on a detailed analysis of placo- zoans found in the small mangrove island of Twin Cays, Belize, where we provide the first high-resolution phylo- geographic study of the Placozoa. Materials and Methods We undertook both a coarse-resolution study, designed to sample for the presence and diversity of placozoans in Belize, Bermuda, Grenada, Jamaica, and Panama, and a fine-resolution study, designed to sample for the presence and diversity as well as the distribution and abundance of placozoans in Twin Cays, Belize. All animals were col- lected using glass slides encased in modified plastic micro- scope slide boxes, as designed by Pearse (UC Santa Cruz, pers. comm., 2005). We cut the top and bottom panels from slide boxes (9.5 cm tall by 8.2 cm wide), leaving a narrow 1-cm border. Five glass slides (7.5 cm by 2.5 cm) were evenly spaced inside, and the box tops and bottoms were secured together by cable ties. Either cable ties or rope were used to attach the slide boxes to mangrove roots or dock piles at a depth of 40 cm or greater, and the boxes were exposed to the marine environment for 2 to 7 weeks. Each slide box was recovered, placed separately in its own dis- posable plastic bin while still submerged, and returned to the laboratory for immediate processing. Each slide was then placed into a large petri dish (14 cm diameter by 2 cm high) fitted with two parallel runners to prevent slides from touching the bottom of the dish. Both sides of each slide were screened for placozoans by using a Zeiss Stemi SR or Wild dissecting microscope. After all slides from a single box were screened, the petri dish was thoroughly rinsed with distilled water before the next box of slides was screened. In our Caribbean-wide coarse sampling, we deployed groups of slide boxes largely in mangrove habitats and boat docks. The deployment sites were chosen for their proxim- ity to marine field stations, as easy access to laboratory facilities was necessary for conducting the animal screens. Animals were collected near or at the Smithsonian Institu- tion Marine Station at Carrie Bow Cay in Belize, the Ber- muda Biological Station for Research (BBSR), St. George's University Marine Laboratory in Grenada, Discovery Bay CARIBBEAN PLACOZOAN PHYLOGEOGRAPHY 151 Marine Laboratory in Jamaica, and the Smithsonian Tropi- cal Research Institute (STRI) in Bocas del Toro in Panama. For details of the deployment sites, habitats, and number of boxes deployed, see Table 1. Slides positive for placozoans were individually placed in 50-ml Falcon tubes completely filled with seawater and transported within 2 days to the laboratory at Yale Univer- sity. Animals collected from Twin Cays, Belize, during the summer of 2003 were first cultured at the marine station in Carrie Bow Cay for 3 weeks, then transported as above to Yale. In the laboratory, single isolates from each slide were cultured in glass petri dishes filled with about 250 ml of filter-sterilized artificial seawater (Reef Crystals, Marine- land Labs, Moorpark, CA), salinity 37 psu, supplemented with 250 jui of Micro Algae Grow (Florida Aqua Farms, Dade City, FL) and 2 ml of a stationary phase culture of Pyrenomonas salina. Genomic DNA was extracted from successfully cultured isolates following published protocols (Signorovitch et al, 2005). Since our second goal was to obtain a fine-resolution phylogeographic map of Twin Cays, Belize, we systemati- cally sampled the margins of this island over the course of two summer field seasons (August 2003 and June 2004). We chose Twin Cays because of its small size; proximity to Carrie Bow Cay, the Smithsonian Marine Station in Belize; and prior knowledge that placozoans occur in this region. We sampled at intervals of at least 10 m for a total of 150 slide boxes. Although no particular habitat in Twin Cays was purposely selected for sampling, mangrove trees line most of the island, and therefore, most of our sampling sites reflected the fauna and flora associated with a mangrove ecosystem. One exception was a sandy beach habitat in the northwest part of the island. For each single animal isolate, a region of the mitochon- drial 16S ribosomal DNA (16S rDNA) was amplified by polymerase chain reaction, using the forward 5'-CGAGAA- GACCCCATTGAGCTTTACTA-3' and reverse 5-TACG- CTGTTATCCCCATGGTAACTTT-3' primer pair under the following PCR conditions: 95 ?C denaturation for 2 min; 5 cycles: 95 ?C for 30 s, 63 ?C for 30 s, and 72 ?C for 1 min; 5 cycles: 95 ?C for 30 s, 62 ?C for 30 s, and 72 ?C for 1 min; 20 cycles: 95 ?C for 30 s, 61 ?C for 30 s, and 72 ?C for 1 min; 72 ?C final extension for 10 min. The amplification products were purified using QIAquick (Qiagen) and se- quenced in both directions with PCR primers by either dGTP BigDye or TaqFS dye terminator cycle sequencing reactions using ABI PRISM 3730 DNA sequencers (Ap- plied Biosystems, Inc.) at the W. M. Keck Biotechnology Center at Yale University. ABI chromatogram files and DNA sequences were analyzed by the LaserGene software package, Ver. 6 (DNASTAR, Inc.) and aligned by CLUST- ALW (Chenna et al, 2003) with adjustments performed manually. DNA sequences were deposited in the GenBank database, searchable by strain name or accession numbers: DQ389756-DQ389767 and DQ389769-DQ389885. Placo- zoan samples were also deposited in the Yale University Peabody Museum and are identified by strain name under accession number 10653. Maximum parsimony (MP) phy- logenetic reconstruction was performed only on unique haplotypes in addition to those available from GenBank (accession numbers: AY652522-AY652529 and AY603696) using the heuristic search option in PAUP* 4.0M0 (Swofford, 1998) with default values. Bootstrap values for the MP tree were obtained by running 10,000 bootstrap replicates under the full heuristic search method. Bayesian likelihood inference was carried out using the K80+G model of evolution obtained from Modeltest, ver. 3.7 (Posada and Crandall, 1998). Bayesian posterior prob- abilities were obtained from the software MrBayes, ver. 3.1.2 (Huelsenbeck and Ronquist, 2001; Ronquist and Huelsenbeck, 2003) using Nchains = 4, temp = 0.5, and running 1,000,000 Markov Chain Monte Carlo generations, sampling at every 100 generations with a burn in of 25% (potential scale reduction factor = 1.0 for all parameters). Heretofore, the nomenclature used to describe known placozoan haplotypes was that introduced by Voigt et al. (2004). We here build upon this naming convention by introducing two additional terms. First, we classified the haplotypes identified by Voigt et al. (2004) into five mono- phyletic clades: Clade I contains haplotypes HI and H2; Clade II contains haplotype H3; Clade III is composed of haplotypes H6, H7, and H8; Clade IV contains haplotype H5; and finally, Clade V is represented by haplotype H4. For any new haplotype discovered in this study, we identi- fied its closest known haplotype by sequence homology and named it in a fashion consistent with the already established haplotype nomenclature. For example, a new haplotype highly similar to H4 would be named H4-2 and classified as a member of Clade V. The second term we introduce here is a nomenclature for the individual strains used in this study. All individuals were given a name that begins with their two-letter country code of origin, followed by a labo- ratory strain identifier. For instance, individuals BZ10101, JM511, and GD711e originate from Belize, Jamaica, and Grenada, respectively. Results The results of our sampling efforts are summarized in Table 1. Placozoans were found in all five countries and in the majority of habitat types sampled, but not at every deployment site. Mangrove habitats consistently yielded placozoans regardless of whether the roots possessed a rich epifauna. For example, placozoans were collected from mangrove roots in Mangrove Bay and Walsingham Bay, Bermuda, where the roots were covered with epiphytes but had few epifauna. Placozoans were even found in the least pristine habitats sampled, including the muddy Mangrove 152 A. Y. SIGNOROVITCH ET AL. Table 1 Summaiy of sampling locations and haplotypes discovered Country Deployment site Habitat type Deployment Exposure date time (month, year) (weeks) ?a tt Z M 12 8 0 1 0 0 2 2 0 9 4 7 Clade: haplotype (individual strain names) Belize Carrie Bow Cay Carrie Bow Cay Culew Cay Twin Cays Twin Cays Twin Cays Twin Cays boat dock seawater system mangrove roots mangrove roots sandy beach mangrove roots 08/2002 08/2002 08/2002 08/2002 06/2004 06/2004 5 74 0 35 0 20 I: H2 (BZ514, BZ516, BZ534, BZ611, BZ613) IE: H8 (BZ10101) V: H4 (BZ931) II: H3 (BZ12) III: H6 (BZ227) III: H7 (BZ45, BZ312, BZ651, BZ672) IE: H8 (BZ264, BZ384, BZ46, BZ413) V: H4 (BZ42, BZ49) II: H3 (BZ2423, BZC12, BZD10, BZD12) IE: H6 (BZD5, BZDli) IE: H7 (BZB8, BZE12, BZF2, BZF4) IE: H8 (BZC2, BZC6, BZE8, BZF1) V: H4 (BZB11, BZD3, BZD8, BZE9, BZF3, BZF5) Bermuda Coot Pond Private dock in St. David's Mangrove Bay BBSR Concrete Beach Mangrove Lake Walsingham Pond Walsingham Bay shallow open pond lined with Zostera boat dock mangrove roots concrete boat ramp murky enclosed tidal pond enclosed tidal mangrove roots V: H4-2 (BMCP24, BMCP41, BMCP51) V: H4-3 (BMCP34) 1 1 V: H4-3 (BMJPH) 4 3 V: H4(BMMB11) V: H4-2 (BMMB42) V: H4-3 (BMMB33) 1 0 1 0 V: H4-2 (BMWP53) V: H4-3 (BMWP12, BMWP22, BMWP45) I: H2 (BMWBI1) V: H4-2 (BMWB33) V: H4-3 (BMWB22) boat dock Grenada Grand Anse Grand Mai Westerhall Point Ft. Jeudy Hog Island Jamaica Discovery Bay Mar. Lab. 03/2003 Panama STRI, Bocas del Toro buoys in channel of sandy beach fish and boat dock mangrove roots mangrove roots mangrove roots mangrove roots buoy over sandy bottom fish dock back reef Thallasia bed mangrove roots Thallassia bed 02/2003 06/2002 4 4 1 3 1 0 3 0 0 3 3 1 2 2 0 2 1 0 III: H6 (GD711g) III: H6 (GD1721e) I: HI (JM5U) III: H8 (JM532, JM545) 0 0 1 0 1 5 I: H2 (PNal, PNb4, PNb5, PNb6) IE: H8 (PNa2) 0 0 CARIBBEAN PLACOZOAN PHYLOGEOGRAPHY 153 Lake, also in Bermuda, which emitted foul odors reminis- cent of sewage, and a "fish dock" at Grand Mai, Grenada, where the water was polluted with a layer of oil. Placozoans were not found near the sandy beach habitat in Twin Cays, but one sample was obtained from a buoy near the extensive sandy beach of Grand Anse, Grenada, and another from a buoy over sandy bottom in Jamaica. Slides at the time of sampling typically displayed an early successional marine community comprising a biofilm of bacteria and algae, as well as a rich assembly of epifauna, including juvenile sponges, foraminiferans, tube worms, hy- drozoans, copepods, snails, bryozoans, and ascidians. Pla- cozoans were often observed crawling over the biofilm and on hydrozoan stolons. Their colors varied from almost col- orless to bright pink, depending on the substrate composi- tion and color. Placozoans varied in size from less than 0.5 mm to as large as 3 mm in diameter, and most were found in an actively feeding state, as indicated by their flattened, stationary configuration. Slides on which placozoans were found typically harbored several individuals. The results of our fine-scale sampling in Twin Cays, Belize, are mapped in Figure 1: colors indicate whether placozoans were absent or present at each site. Placozoans were distributed throughout this mangrove island, with the exception of the sandy beach habitat in the northwest re- gion. Of the 159 slide boxes deployed around the perimeter of Twin Cays during the years 2002 through 2004, 68 were positive for placozoans. Most mangrove roots to which slide boxes were attached contained a variety of sponges, colonial ascidians, anemones, and filamentous algae. Turbidity in the water was low, and light intensity varied from high to low throughout the day. Slides displayed an assemblage of or- ganisms similar to that seen in the coarse-resolution study, with the addition of the benthic ctenophore Vallicula mul- tiformis, which was abundant on many of the recovered slides. Using the combined coarse- and fine-resolution 16S rDNA data from this study and the 16S rDNA sequences, available from GenBank, of Voigt et al. (2004) (AY652522-AY652529, corresponding to haplo- types H1-H8, respectively) and Aleoshin et al. (2004) (AY603696), we constructed a partial 16S rDNA phy- logeny of the Placozoa (Fig. 2). We were unable to sequence one region within this locus in part of our samples because of a guanine/cytoseine-rich hairpin sec- ondary structure; therefore, the data have been parti- tioned into separate 5' and 3' regions. We used 73 DNA samples with an average concatenated length of 358 nt and a range from 292 to 406 nt. Due to large insertions or deletions (indels), the aligned data set spanned 480 nt. Removing indel sites reduced the data to 241 nt. Both parsimony (MP) and likelihood (Bayesian inference) methods returned the same tree topology for the well- supported nodes. On the basis of sequence divergence, the 100 m absent present Clade I Clade II Clade III Clade V Figure 1. Map of Twin Cays, Belize. Dots on the map represent the sampling sites. Grey dots indicate sites from which placozoans were not detected. Pink dots represent sites from which placozoans were detected but not genotyped. The remaining colored dots represent sites positive for placozoans from which individual strains were genotyped. The habitat type at the five sampling sites in the northwestern-most part of the map is sandy beach; all other sampling sites are mangrove roots. (Map modified from a figure prepared by Molly K. Ryan for the Caribbean Coral Reef Ecosys- tems Program; used with permission.) bootstrap percentages and Bayesian posterior probabilities, five clades emerged from the combined 73 samples, here- after referred to as Clade I for haplotypes HI and H2; Clade II for haplotype H3; Clade III for haplotypes H6, H7, and H8; Clade IV for haplotype H5; and Clade V for haplotypes H4, H4-2, and H4-3 (Fig. 2). Our coarse-and fine-resolution studies revealed seven of the eight haplotypes already de- scribed by Voigt et al. (2004), and in addition, two new haplotypes, H4-2 and H4-3, were discovered in Bermuda. Haplotype H4-2 differed from the H4 haplotype discovered by Voigt et al. (2004) by only two nucleotide substitutions, while haplotype H4-3 differed from H4 by four nucleotide substitutions. The last column of Table 1 lists the haplotypes detected during this study. In Bermuda (n = 15), 14 samples were found to belong to Clade V, while only one belonged to Clade I. Jamaican samples (n = 3) separated into Clades I and III. Grenadian samples (n = 2) belonged to Clade III, 154 A. Y. SIGNOROVITCH ET Ah. 100/1.00 100/1.00 100/1.00 81/ 0.96 f, AY603696 Sea of Japan H1 Panama (C), Red Sea JM511 L H2 Italy, Panama (C) BMWB11, BZ514, BZ516, BZ534, BZ611, BZ613, PNa1, PNb4, PNb5, PNb6 Clade I H3 Panama (C) BZ12, BZ2423, BZC12, BZD10, BZD12 99/0.98 rf H6 Panama (P) BZ227, BZD5, BZD11, GD711g, GD1721e 88/0.95 (. H5 Italy) Clade IV Clade II r H7 Indo-Pacific, Panama (P) BZ45, BZ312, BZ651, BZ672, 1 BZB8, BZE12, BZF2, BZF4 ? H8 Guam (P), Panama (P) BZ264, BZ384, BZ46, BZ413, BZ10101, BZC2, BZC6, BZE8, BZF1, JM532, JM545, PNa2 Clade H4 Panama (C&P), Venezuela (C) BZ42, BZ49, BZ931, BZB11, BZD3, BZD8, BZE9, BZF3, BZF5 H4-2 BMCP24, BMCP41, BMCP51, BMMB42, BMWB33, BMWP53 LH4-3 BMCP34, BMJP11, BMMB33, BMWB22, BMWP12, BMWP22, BMWP45 Clade V 0.1 Figure 2. 16S rDNA phytogeny of combined placozoan samples from this study and sequences available in GenBank (represented by HI, H2, H3, H4, H5, H6, H7, H8, and AY603896). Haplotypes are represented in boldface type. Haplotypes H4-2 and H4-3 were newly discovered in this study. Individual strains sequenced in this study are italicized and labeled by their two-letter country code of origin (BM = Bermuda, BZ = Belize, GD = Grenada, JM = Jamaica, PN = Panama), followed by laboratory strain identifiers. Localities (countries and seas) sampled by Voigt et at (2004) and Aleoshin et at (2004) are explicitly listed next to haplotypes in bold. (C) = Caribbean Ocean, (P) = Pacific Ocean. Numbers at each internal node represent % bootstrap under maximum parsimony and Bayesian posterior probabilities, respectively (only values above 70% and 0.70 are shown). CARIBBEAN PLACOZOAN PHYLOGEOGRAPHY 155 and Panamanian samples (n = 5) fell into Clades I and III. In Twin Cays (n = 39) we uncovered six haplotypes that fall into Clades I?III and V. Figure 1 shows this haplotype distribution among the sampled sites in Twin Cays. The haplotype groupings (H2), (H3), (H6, H7, H8), (H4) corre- spond to Clades I?II and V, respectively. The three most abundant haplotypes were H7, H8, and H4, and the most diversity was observed in the main channel region that separates the two land masses composing the island. Because only a subset of sampled animals were success- fully established in laboratory culture during the 2003 Twin Cays field season (41%, see Table 1) and because sequence data were generated only for those cultured isolates, it is possible that the molecular data might be biased by differ- ential viability of different clades. To address this possible survival bias in our 2003 data, in 2004 we used a Clone- Saver FTA card (Whatman) to sample 19 placozoan indi- viduals immediately upon their retrieval from the field, thus avoiding an intervening step of laboratory culture (for pro- tocol, see Signorovitch et al, 2005). Sequencing of these 19 DNA samples yielded only representatives of the known clades. Discussion The finding by Voigt et al. (2004) that the Placozoa is not a monotypic phylum raises a number of immediate ques- tions that our study was designed to address. For instance, will extensive sampling result in the discovery of additional placozoan clades? Do the known clades display obvious geographic limits or habitat preferences? And finally, are these clades sympatric on an ecologically relevant spatial scale? Our coarse-scale Caribbean-wide sampling revealed four of the five clades of placozoans previously identified (Voigt et al, 2004). Clades I, III, and V all were widely distributed within the Caribbean basin. Voigt et al. had found Clade I to contain animals from the Red Sea, Italy (Mediterranean), and Panama (Caribbean side), and Aleoshin et al. (2004) identified a sample from the Sea of Japan (Fig. 2). Our study confirmed the presence of Panamanian animals in this clade and further added Belizean, Bermudian, and Jamaican rep- resentatives to it. We have also added Belizean animals to Clade II, which previously contained only a single Carib- bean sample from Panama. Members of Clade III were previously identified only from the Pacific in Panama, the Indo-Pacific, and Guam (Voigt et al, 2004), but we found that this clade displayed the most geographic diversity among the four clades. Indeed, Clade III contained animals from all but one of our sampling locations in the Caribbean and was widely distributed throughout Twin Cays, Belize. Finally, Clade V was represented by Panamanian (both Caribbean and Pacific sides) and Venezuelan samples (Voigt et al, 2004), and our study has now added Belizean and Bermudian placozoans to this group. Because of diffi- culties in establishing animals in laboratory culture, our sampling efforts in different locales were highly uneven, preventing us from assessing geographic population struc- ture across islands. Nevertheless, the fact that our Caribbean- focused study found identical haplotypes (H2, H3, H4, H6, H7, and H8) to those discovered in the worldwide sampling of Voigt et al. (2004) adds further support to the idea that placozoans, like other microbial eukaryotes (Finlay, 2002), are geographically unrestricted because of their small body size, pelagic stage in their life cycle, and large population size. In the course of these studies, samplings were concen- trated around docks, mangrove roots, and to a lesser extent, patch reefs. These habitat types were similar to those pre- viously sampled (Voigt et al, 2004). Inspection of Table 1 reveals no clear restriction of placozoans to any one of these habitats, with the exception of the sandy beach. The absence of placozoans in the sandy beach habitat of Twin Cays agrees with Pearse's observations in Papua New Guinea. The glass slides recovered from the sandy beach habitat in Twin Cays were very clean?almost devoid of any epifauna or flora. Because placozoans are often found on glass slides covered with a noticeable film of bacteria and algae as well a rich assemblage of other microorganisms, it is possible that such a diversity of organisms is necessary for their recruitment. The fact that we found several known haplo- types throughout the Caribbean and discovered only two new haplotypes that fall within an already reported clade suggests that most of the placozoan diversity in these hab- itats is adequately characterized in this region. It bears emphasis, however, that the epifauna of pilings and man- groves is only a tiny fraction of the possible habitats for placozoans, and further discoveries of placozoan diversity might take place in other habitats or in other parts of the world. The lack of any obvious clade-specific habitat preference in our coarse-scale analysis is consistent with the results of our fine-scale study in the mangrove community of Twin Cays (Fig. 1). All four clades were isolated from this 92- hectare island. Members of Clades III and V were wide- spread throughout the island, whereas members of Clades I and II were found only within the Main Channel region dividing the two landmasses composing the island (Fig. 1). Clades III and V were the most abundant and widely dis- tributed clades, while Clade I was especially rare. These results clearly indicate that the four clades occur in sympa- thy in Twin Cays on an ecologically relevant spatial scale. Furthermore, sympatry is likely to be a widespread phenom- enon, as indicated by observations of multiple clades in Panama and Italy (Voigt et al, 2004) and the fact that even a single slide from our Panamanian collection yielded rep- resentatives of two different clades (Table 1). The sympatry 156 A. Y. SIGNOROVITCH ET Ah. of clades raises questions about the ecological dynamics among these grossly similar placozoans. Acknowledgments We are grateful to Klaus Rutzler, Mike Carpenter, Rafael Rosengarten, Matthew Nicotra, Casey Dunn, Erika Ed- wards, Clare Morrall, Oliver Baimer, John Gilbert, Craig Layne, Christina Glastris, Matt Daren, and Joanna Pitt for their help during deployment or collection of placozoan boxes and animal screening. Zack Snable, Aaron Thier, and Chad Kritzberger helped with animal care. We thank Maria Moreno, Vicki Pearse, and James Signorovitch for helpful discussions and critical reading of this manuscript. We also thank two anonymous reviewers whose suggestion greatly improved this manuscript. This work was supported in part by a Caribbean Coral Reef Ecosystem Award (contribution no. 748) and National Institutes of Health Genetics Training Grant 5 T32 GM07499-28 to A.Y.S. and a National Science Foundation Grant EF-0319076 and National Institutes of Health Grant R21-AI066242-01A1 to L.W.B. Literature Cited Aleoshin, V. V., A. V. Konstantinova, M. A. Nikitin, and I. L. Okshtein, 2004. On the genetic uniformity of the genus Trichoplax (Placozoa). Russ. J. Genet. 40: 1423-1425. Chenna, R., H. Sugawara, T. Koike, R. Lopez, T. J. Gibson, D. G. Higgins, and J. D. Thompson. 2003. Multiple sequence alignment with the CLUSTAL series of programs. Nucleic Acids Res. 31: 3497- 3500. Finlay, B. J. 2002. Global dispersal of free-living microbial eukaryote species. Science 296: 1061-1063. Grell, K. G. 1971a. Trichoplax adhaerens F. E. Schulze und die Entstehung der Metazoen. Natmwiss. Rundsch. 24: 160-161. Grell, K. G. 1971b. Embryonic development of Trichoplax adhaerens F. E. Schulze. Natuiwissenschaften 58: 570. Grell, K. G. 1972. Formation of eggs and cleavage in Trichoplax adhaerens, F. E. Schulze, Z. Morphol. Tiere 73: 297-314. Grell, K. G., and G. Benwitz. 1971. Die Ultrastruktur von Trichoplax adhaerens F. E. Schulze. Cytobiologie 4: 216-240. Grell, K. G., and E. Lopez-Ochoterena. 1987. A new record of Trichoplax adhaerens F. E. Schulze (phylum Placozoa) in the Mexican Caribbean Sea. An. Inst. Cienc. Mar Limnol. 14: 255-256. Huelsenbeck, J. P., and F. Ronquist. 2001. MRBAYES: Bayesian inference of phytogeny. Bioinformatics 17: 754-755. Ivanov, D. L., V. V. Malakhov, and A. B. Tzetlin, 1980. Discovery of a primitive multicellular organism Trichoplax sp. Zool. Zhurnal 59: 1735-1738. Martinelli, C., and J. Spring. 2003. Distinct expression patterns of the two T-box homologues Brachyury and Tbx2/3 in the placozoan Trichoplax adhaerens. Dev. Genes Evol. 213: 492?499. Maruyama, Y. K. 2004. Occurrence in the field of a long-term, year- round, stable population of placozoans. Biol. Bull. 206: 55-60. Miller, R. L. 1971. Trichoplax adhaerens Schulze, 1883: return of an enigma Biol. Bull. 141: 374. (Abstract). Pearse, V. B. 1989. Growth and behavior of Trichoplax adhaerens: first record of the phylum Placozoa in Hawaii. Pac. Sci. 43: 117-121. Posada, D., and K. A. Crandall. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818. Ronquist, F., and J. P. Huelsenbeck. 2003. MRBAYES 3. Bioinfor- matics 19: 475-481. Ruthmann, A. 1977. Cell-differentiation, DNA content and chromo- somes of Trichoplax adhaerens F. E. Schulze. Cytobiologie 15: 58-64. Schulze, F. E. 1883. Trichoplax adhaerens, nov. gen., nov. spec. Zool. Anz. 6: 92-97. Signorovitch, A. Y., S. L. Dellaporta, and L. W. Buss. 2005. Molec- ular signatures for sex in the Placozoa. Proc. Natl. Acad. Sci. USA 102: 15518-15522. Swofford, D. L. 1998. PAUP*, Phylogenetic Analysis Using Parsimony (* and other methods). Sinauer, Sunderland, MA. Voigt, O., A. G. Collins, V. B. Pearse, J. S. Pearse, A. Ender, H. Hadrys, and B. Schierwater. 2004. Placozoa?no longer a phylum of one. Curr. Biol. 14: R944-R945.