Reef coral reproduction in the equatorial eastern Pacific: Costa Rica, Panamá, and the Galápagos Islands (Ecuador). VII. Siderastreidae, Psammocora stellata and Psammocora profundacella

Reef coral reproduction in the equatorial eastern Pacific: Costa Rica, Panamá, and the Galápagos Islands (Ecuador). VII. Siderastreidae, Psammocora stellata and Psammocora profundacella P. W. Glynn, S. B. Colley, J. L. Maté, I. B. Baums, J. S. Feingold, J. Cortés, et al. Marine Biology International Journal on Life in Oceans and Coastal Waters ISSN 0025-3162 Mar Biol DOI 10.1007/s00227-012-1979-5

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag. This e-offprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication.

1 23

Author's personal copy Mar Biol DOI 10.1007/s00227-012-1979-5

ORIGINAL PAPER

Reef coral reproduction in the equatorial eastern Pacific: Costa Rica, Panama´, and the Gala´pagos Islands (Ecuador). VII. Siderastreidae, Psammocora stellata and Psammocora profundacella P. W. Glynn • S. B. Colley • J. L. Mate´ • I. B. Baums J. S. Feingold • J. Corte´s • H. M. Guzma´n • J. C. Afflerbach • V. W. Brandtneris • J. S. Ault

•

Received: 5 April 2012 / Accepted: 31 May 2012 Ó Springer-Verlag 2012

Abstract Two zooxanthellate, scleractinian species present in the equatorial eastern Pacific, Psammocora stellata and Psammocora profundacella, were examined in terms of their reproductive biology and ecology at four study sites, non-upwelling (Can˜o Island, Costa Rica, and Uva Island, Panama´), upwelling (Gulf of Panama´, Panama´), and seasonally varying thermal environments (Gala´pagos Islands). Both species were gonochoric broadcast spawners lacking zooxanthellae in mature ova. Mature gametes and spawned gonads are present around full moon; however, no spawning was observed naturally or in outdoor aquaria. Mature gametes occurred in P. stellata at Can˜o Island for nearly 6 months, and year round at Uva Island, both non-upwelling sites. Reproductively active colonies occurred mostly in the warmer months in

the Gulf of Panama´ and Gala´pagos Islands. In the Gala´pagos Islands, where collecting effort was greatest for P. profundacella, mature gametes were also most prevalent during the warm season. Annual fecundity was high in both species, 1.3–1.8 9 104 ova cm-2 year-1 in P. stellata and 1.2–2.0 9 104 ova cm-2 year-1 in P. profundacella. Compared to other eastern Pacific corals, P. stellata was relatively resistant to ENSO-related bleaching and mortality, especially populations inhabiting deep (12–20 m) coral communities. Rapid recovery and persistence of Psammocora spp. can be attributed to several factors: (a) relative resistance to bleaching, (b) deep refuge populations, (c) broadcast spawning, (d) protracted seasonal reproduction, (e) high fecundity, and (f) asexual propagation. Introduction

Communicated by J. P. Grassle.

Electronic supplementary material The online version of this article (doi:10.1007/s00227-012-1979-5) contains supplementary material, which is available to authorized users. P. W. Glynn (&)  J. C. Afflerbach  V. W. Brandtneris  J. S. Ault Division of Marine Biology and Fisheries, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, USA e-mail: [email protected] S. B. Colley Louisiana Applied Coastal Engineering and Science Division, Office of Coastal Protection and Restoration, 450 Laurel St., Ste. 1200, Baton Rouge, LA 70801, USA J. L. Mate´  H. M. Guzma´n Smithsonian Tropical Research Institute, PO Box 0843-03092, Balboa, Ancon, Republic of Panama´

Knowledge of the sexual reproductive biology of eastern Pacific scleractinian corals has increased significantly since

I. B. Baums Department of Biology, The Pennsylvania State University, 208 Mueller Laboratory, University Park, PA 16802, USA J. S. Feingold Nova Southeastern University Oceanographic Center, 8000 North Ocean Drive, Dania Beach, FL 33004, USA J. Corte´s Centro de Investigacio´n en Ciencias del Mar y Limnologı´a (CIMAR), and Escuela de Biologı´a, Ciudad de Investigacio´n, Universidad de Costa Rica, San Pedro, 11051-2060 San Jose´, Costa Rica

123

Author's personal copy Mar Biol

the 1980s. This is especially true for the equatorial eastern Pacific (Colley et al. 2000; Glynn et al. 1991, 1994, 1996, 2000, 2008, 2011) and more recently for the Mexican coral fauna (Vizcaı´no-Ochoa 2003; Mora-Pe´rez 2005; Rodrı´guez-Troncoso 2006; Cha´vez-Romo and Reyes-Bonilla 2007; Lo´pez-Pe´rez et al. 2007; Carpizo-Ituarte et al. 2011; Rodrı´guez-Troncoso et al. 2011). As in the equatorial eastern Pacific, patterns of coral sexual activity are now known for several species inhabiting upwelling and nonupwelling environments from northern- to southern-most Mexican Pacific sites. The asexual regeneration of corals surviving disturbance events has been observed in several eastern Pacific areas and has been well documented for branching and massive species that survive ENSO disturbances (e.g., Guzma´n and ´ ngel et al. 2001; Glynn and Fong Corte´s 2001; Vargas-A 2006; Glynn et al. 2009, 2011). Guzma´n and Lo´pez (1991) noted that pufferfish corallivores likely contribute to the propagation of Psammocora spp. while feeding on these colonies. The pufferfish Arothron meleagris is a known predator of P. stellata in the Gala´pagos Islands (Feingold 1996), and may likewise promote asexual reproduction. In contrast to other coral reef biogeographic regions, the great majority of eastern Pacific corals are broadcast, asynchronous (no multispecies spawning) spawners, and reproductively active year round in non-upwelling environments. Only two brooding corals inhabit equatorial eastern Pacific coral reefs, Tubastraea coccinea, an azooxanthellate cryptic species (Glynn et al. 2008), and Porites panamensis, a zooxanthellate small, encrusting species (Smith 1991; Glynn et al. 1994). This study investigates the reproductive biology of two siderastreid zooxanthellate species that are widespread and often locally abundant on and adjacent to eastern Pacific coral reefs, Psammocora stellata and Psammocora profundacella (Glynn and Ault 2000; Reyes-Bonilla 2002). These species are generally not important in terms of framework construction, but P. stellata often forms unconsolidated mounds in both shallow and deep reef zones that provide shelter for diverse coral reef-associated species (Corte´s 1990; Feingold 1996; Bezy et al. 2006). With the goal of better understanding post-disturbance resilience and recovery, the following aspects of the biology and ecology of the two siderastreid species were investigated: (a) sexual systems and mode of reproduction, (b) reproductive activity under contrasting oceanographic regimes, (c) seasonal and lunar spawning cycles, (d) fecundity, and (e) recovery of ENSO-impacted populations. With this publication, studies of the reproductive biology/ ecology of 13 major scleractinian reef corals present on equatorial eastern Pacific reefs will have been completed (e.g., Colley et al. 2000; Glynn et al. 2011).

123

Materials and methods Species and collections Morphological and molecular evidence demonstrate clear and consistent differences between Psammocora stellata and Psammocora profundacella (Benzoni et al. 2007). Psammocora stellata displays a branching colony habit with relatively large calice diameters (mean & 2.0 mm), wide enclosed petaloid septa (mean & 0.25 mm), and relatively closely spaced calices without encircling ridges (Stefani et al. 2008). Psammocora profundacella typically grows as encrusting, submassive to massive colonies, with relatively small calice diameters (mean & 1.5 mm), narrow enclosed petal septa (mean & 0.15 mm), and relatively well separated calices surrounded by ridges (Benzoni et al. 2010). Psammocora superficialis, originally reported from various eastern Pacific localities, has been synonymized with P. profundacella (see Benzoni et al. 2010) and samples that we collected as P. superficialis are here assigned to P. profundacella. All collected colonies of P. stellata were free living, usually present on mixed sandy and rubble substrates; P. profundacella colonies were generally firmly cemented to stable substrates (see Electronic Supplementary Material, Appendices 1, 2). Both species were collected over relatively large areas (50–100-m search paths along isobath), with colonies usually separated by 5–10 m or more, in order to avoid clonemates. Sampling of Psammocora stellata was performed by cutting or breaking sections of colonies. Samples usually included 3–5 branchlets, and sampled colonies 10–20 branches. Approximately 2-cm2 sections were chiseled or pried from Psammocora profundacella colonies. Sampling was performed on 3–5 different colonies during each collection, but this was sometimes not possible for P. profundacella due to its relative scarcity. Repeated sampling of known colonies was not performed with one exception. A large (&0.25 m2) male colony of P. stellata at Taboga Island, Gulf of Panama´, was re-sampled in every calendar month during 1990–1991. Generally, the same populations were re-sampled on multiple occasions without any discernible effects on species’ abundances. The general locations of the nine principal study areas in Costa Rica, Panama´, and Ecuador (Gala´pagos Islands) span only about 10o latitude, but are subject to very different environmental conditions (Fig. 1). In Costa Rica, samples were collected at two fringing reefs at the northeast end of Can˜o Island, site 1 (Guzma´n and Corte´s 1989). Sampling in Panama´ was performed in two areas: the Gulf of Panama´ (seasonal upwelling) and the Gulf of Chiriquı´ (nonupwelling). Sampling in Chiriquı´ was performed on the Uva Island reef (site 2) or \1 km north on a rocky promontory. Additional samples were collected at Taboga Island

Author's personal copy Mar Biol

Fig. 1 Collection sites in Costa Rica, Panama´ (Gulf of Chiriquı´ and Gulf of Panama´), and the Gala´pagos Islands. Gala´pagos Islands sites include Itabaca Canal and Baltra Island (site 5), Robinson Cove (site 6) and Punta Estrada/Academy Bay (site 7) on Santa Cruz Island, north end of Floreana near Punta Cormorant (site 8), and eastern end

of Espan˜ola Island, Gardner Bay (site 9). All Gala´pagos study sites are thermally similar, located within the SE bioregion (Wellington et al. 2001; Edgar et al. 2002). Solid recurved arrows denote upwelling centers

(site 3) and Saboga Island (site 4) in the Gulf of Panama´. The Gala´pagos Islands sites included Itabaca Canal and Baltra Island (site 5), Robinson Cove (site 6) and Punta Estrada/Academy Bay (site 7) on Santa Cruz Island, the north end of Floreana Island near Punta Cormorant (site 8), and the eastern end of Espan˜ola Island (site 9). These five central and southern island sites are located within Harris’s (1969) relatively mild temperature zones 1 and 2, which demonstrate very similar seasonal patterns (Wellington et al. 2001). The collecting effort for both Psammocora species at all sites and years is noted in Appendix 3.

less than 2 h. Within 1 h after collection, samples were fixed in seawater/Zenker’s solution with 5 % formaldehyde for 18–24 h. Decalcification and histological tissue preparation were performed as previously described in Glynn et al. (1994, 1996, 2000). Tissues were embedded in Paraplast Plus and oriented so that the branches lay horizontally. The paraffin blocks were then sectioned into approximately 7-lm-thick slices. One slide of serial sections was prepared from each of 3 different levels of the mid-polyp region, approximately 70, 200, and 500 lm below the mouth. These tissues were then mounted on slides and stained with a slightly modified Heidenhain’s aniline-blue method (Luna 1968) using Azocarmine G. One section per slide was outlined with a fine marker pen and examined under 100–4009 magnification with a light microscope.

Histology After collection, samples were transported underwater in wide-mouth bottles or netted bags during dives lasting

123

Author's personal copy Mar Biol

Spawning Attempts to observe spawning in the field were performed most consistently for Psammocora stellata at Uva Island in Panama´ from 1990 to 2003 by monitoring colonies in (1) natural reef settings, (2) buckets, aquaria, and water tables aboard ship at anchor, and (3) sealed polyethylene bags tethered to reef substrates. Procedures 1 and 2 were carried out mostly in the dry season (February, March) during lunar days 15–18 with observations concentrated between sunset and midnight. Sample size usually ranged from 10 to 15 colonies, and occasionally to 5–10 colonies. Aquaria and water tables were supplied with running sea water and colonies in 20-l buckets were freshened with approximately hourly seawater changes. Suspected spawned gametes were checked microscopically; ova appearance and colors were noted and diameters measured. Procedure 2 was also performed on several occasions in the Gala´pagos Islands over the course of this study. Systematic monitoring of Psammocora stellata was also performed on colonies collected at Saboga Island, Gulf of Panama´, and then transported to flow-through sea water tables at Naos Island (about 60 km distant). Twenty-one colonies approximately 20 9 50 mm (height by length) were collected 3–4 days before lunar day 15 in July, September, and October 2011 and monitored continuously every 15 min from sunset to sunrise through lunar day 19. On August 4, new colonies were collected and monitored with those remaining from the July collection. Each colony under observation was positioned in a 10-cm-diameter cylinder lined with 64-lm mesh nylon netting topped with an inverted funnel and test tube egg collector. Coral abundances Psammocora stellata abundances are reported for five sites sampled in Costa Rica, Panama´, and the Gala´pagos Islands. Sampling methodologies for the Costa Rican sites are noted in Guzma´n and Corte´s (2001, 2007). Colony abundances were sampled annually or every 2 years in a 288-m2 plot at Uva Island. Colony counts were made in randomly placed 15–25 1-m2 quadrats. Two Gala´pagos sites were sampled repeatedly: Devil’s Crown (Onslow Island), north side of Floreana Island since 1983, and Xarifa Island, east end of Espan˜ola Island, since 2004. At Devil’s Crown, the area covered by P. stellata-dominated patches was determined by field measurements from 1976 to the early 2000s (Glynn and Wellington 1983; Glynn 1994) and with high precision area measurements with CPCe (Kohler and Gill 2006) in 2011. Colony densities were determined using randomly placed 0.5 9 0.5-m quadrats in 2004 (n = 60) and 2011 (n = 75). Since 2004, at Xarifa Island, P. stellata abundance was determined from percent live cover and

123

numbers of colonies from quadrat sampling in a shallow (0.5–3.0 m) 34-m2 plot dominated by this species. This population has been present at Xarifa since first observed in the 1980s (T. DeRoy, personal communication) and has persisted through two very strong ENSO events (1982–1983 and 1997–1998). Statistical analyses Chi-square analyses were performed on the occurrence of spawned gonads in relation to lunar phases in histological samples, and sex ratios at all study sites. The seasonal and lunar timing of gametogenesis was tested employing chisquare analyses and the Fisher exact probability test, the latter when sample sizes were small. Due to the limited sample sizes for calculating fecundity metrics in some collections, bootstrapping methods (Davidson and Hinkley 1997; Efron and Tibshirani 1998; Manly 2007) were used to assign unbiased measures of accuracy (e.g., standard error of the mean) to statistical estimates of coral fecundity. Bootstrap samples y ¼ y1 ; . . .; yn were obtained by randomly sampling n times, with replacement, from the original data points y to produce a bootstrap sample that has the same sample size as the original sample (Appendix 4). Confidence intervals of the mean were computed using the central limit theorem where the estimate followed a normal density function. Standard graphical and statistical diagnostics, such as quantile–quantile and residual plots, and Shapiro–Wilk and Kolmogorov–Smirnov tests, were used to assess normality of the particular computed sample distributions and whether two or more sample distributions differed (Zar 1999; Crawley 2005; Kutner et al. 2005; Alder 2010). Sample distributions that failed to meet the assumption of normality were transformed to normality using the Box-Cox ladder of powers procedure (Kutner et al. 2005; Appendices 5–8). Parametric Student’s t, nonparametric Kruskal– Wallis, and Wilcoxon rank sum tests were used to evaluate the equality of sample means. All statistical procedures and tests were implemented in the R Project for Statistical Computing (http://www.r-project.org/, Crawley 2005; Qian 2010; Alder 2010). Other details of the analysis are outlined in the fecundity ‘Results’ section below.

Results Gonad development and gametocyte classification Detailed descriptions of gamete development in female and male colonies are available in Appendices 9 and 10. No discernible interspecific differences were found in gonad development or gamete maturation in the two Psammocora

Author's personal copy Mar Biol

species. Female gamete diameters ranged from &10–25 lm (Stage I oocytes) to &90–130 lm (Stage IV ova) in histological preparations. Zooxanthellae were not observed in any mature ova. Spermary diameters ranged from &5–25 lm (Stage I) to &60–120 lm (Stage IV). Cellular organization appeared disrupted and empty in spent gonads. Several spent spermaries were observed in colonies of Psammocora stellata collected in Panama´ (Gulfs of Chiriquı´ and Panama´) and the Gala´pagos Islands (Appendix 11). Reproductive condition Relatively high percentages of Psammocora spp. contained gonads, spanning most seasons and lunar phases. Over half of sampled colonies at non-upwelling Uva Island were reproductively active—57.8 % of Psammocora stellata and 64.4 % of Psammocora. profundacella (Table 1). The population of P. stellata at seasonal upwelling Saboga Island demonstrated the lowest proportion of reproductive colonies (20.4 %), but P. profundacella at non-upwelling Can˜o Island also exhibited relatively few reproductive colonies (25.0 %). All colonies but one (P. stellata) of both species were gonochoric. Planula larvae were never observed in histological preparations, only gametes in all stages of development and occasionally spent gonads. Nearly equal numbers of P. stellata male and female colonies were sampled at Can˜o Island (51.5 % male and 48.5 % female), and the single large colony repeatedly sampled at Taboga Island contained only spermaries. Sex ratios of P. stellata at Uva and Saboga Islands were significantly skewed with predominantly female colonies (85.4 and 84.2 %, respectively), and male colonies predominated in Gala´pagos samples (71.4 %). Female colonies significantly dominated P. profundacella samples at Uva Island (73.7 %). A preponderance (59.4 %) of P. profundacella female colonies was sampled in the Gala´pagos Islands, but there was no significant deviation from a 1:1 sex ratio. Too few reproductively active colonies of P. profundacella from Can˜o Island were available for reliable statistical testing. Seasonal and lunar activity Psammocora stellata demonstrated significant seasonal differences in the presence of mature gonads at all study sites (Figs. 2a, b, 3, 4, 5a, b; Table 2). At Can˜o Island, all colonies with mature ova (Stage IV) were concentrated in the wet season, with male colonies bearing mature spermaries mostly during the wet season, but also in the dry season. This pattern was reversed at Uva Island where both male and female colonies contained ripe gonads from January to March of the dry season. In the Gulf of Panama´,

nearly all mature gonads occurred in the wet season following seasonal upwelling. Stage IV spermaries were present in only a single colony in February 1992. In the Gala´pagos Islands, mature gonads were most prevalent in male and female colonies in the wet/warm season. A few male and female colonies contained mature gonads in November 1987, during a mild El Nin˜o year (Podesta´ and Glynn 2001). The presence of mature gametes in Psammocora stellata was marginally significantly (p \ 0.05) related to full moon, within 5–6 days at Can˜o and Uva Islands (Figs. 2c, d, 3c, d; Table 2). Gulf of Panama´ (combined Saboga and Taboga sites) and Gala´pagos locations did not show a significant correspondence with any particular lunar phase (Figs. 4c, d, 5c, d; Table 2). Psammocora profundacella demonstrated a significant increase in reproductive activity at Can˜o Island during the dry season, from mid-December to the end of February (Fig. 6a, b; Table 3). No lunar pattern was detected in Costa Rica (Fig. 6c, d; Table 3). The limited collections (4 mo) from Uva Island indicated reproductive activity in the dry and wet seasons, but without statistical significance (Fig. 7a–d, Table 3). The prevalence of Stage IV gametes during lunar days 14–20 suggests a spawning period near full moon (Fig. 7d; Table 3), but not at a statistically significant level. In the Gala´pagos Islands, P. profundacella contained a significantly high proportion of gametes (Stages I–IV) in the wet/warm season (Fig. 8a, b; Table 3). Mature gametes were detected at nearly all lunar phases in the Gala´pagos Islands (Fig. 8c, d). Limited spawning was observed under artificial conditions at Uva Island, Panama´, and at Devil’s Crown (Onslow Island), Gala´pagos Islands. Two of 10 colonies of Psammocora stellata confined to sealed transparent bags overnight at 4 m depth on the Uva reef spawned numerous large (120–150 lm diameter) ova during lunar days 21 and 22, February 2–3, 1994. These eggs were either pearly white or muddy green with a corona of microvilli in the latter. Two colonies of P. stellata spawned at Devil’s Crown on April 9, 1992 at 16:03 h (6 days after new moon). These were collected at 15 m depth and held in aerated buckets following staining with Alizarin Red S vital stain. Microscopic examination of the spawn was not possible, but it appeared to be sperm. Empty gonads, presumably recently spawned, were observed in histological samples from several colonies of Psammocora stellata over a wide range of lunar phases at all major study sites (Appendix 11). Nearly equal numbers of males and females demonstrated this condition. Statistical testing failed to reveal a clustering of spent gonads at any of four lunar phase periods (Chi-square test, p [ 0.90). Spawned gonads were found in just a single female colony of Psammocora profundacella during new moon at Can˜o Island.

123

Author's personal copy Mar Biol Table 1 Psammocora stellata, Psammocora profundacella Location

n

% with gonads

No. of colonies with ooc/sperm

Sex ratio

Male

Female

#:$

v2 (p)

Psammocora stellata Can˜o

78

42.3

17

16

1:0.94

0.8026

Uva Island

154

57.8

13

76

1:5.85

\0.0001

1:5.33

Saboga

93

20.4

3

16

Taboga Gala´pagos Islands

95

37.9

36

0

1:0

0.0028 \0.0001

93

45.2

30

12

1:0.4

0.0055

Psammocora profundacella Can˜o 28

25.0

5

2

1:0.4

0.257

Uva Island Gala´pagos Islands

59

64.4

10

28

1:2.8

0.0035

88

36.4

13

19

1:1.46

0.289

Percent colonies with gonads, sex ratios, and Chi-square analyses. Number of colonies sampled (n), percentage of colonies with gonads, and number of colonies with oocytes or spermaries (ooc/sperm) from five (P. stellata) and three (P. profundacella) sampling localities. Deviation of sex ratio from 1:1 (#:$) was tested with v2 analyses

a

c

b

d

Fig. 2 Psammocora stellata. Reproductive activity at Can˜o Island, Costa Rica, in relation to season (a, b) and lunar phase (c, d), based on 25 collections and 78 colonies examined (1985–1989). All male (top panels in each set) and female (bottom panels) colonies that contained gonads at any stage of development were included in a and

c; only colonies with mature (Stage IV) gonads were included in b and d (Embedded numbers in parentheses denote number of overlapping data points; 5 and 6 denote multiple collections within a given year); c, d abscissas lunar days, full moon occurred near Lunar Day 15. Gray bar denotes zero values

Fecundity

of ova per cross-section of mesentery; (5) number of ova per longitudinal section of mesentery; and (6) number of polyps per cm2 colony surface area. Histological distinction between polyps was often difficult because mesenteries appeared to be shared among several polyps. Gastrodermal canals extended across corallites whose walls were absent or weakly defined (Wells 1956).

Empirical sample data from four locations (1) Can˜o Island, Costa Rica; (2) Uva Island, Panama´; (3) Gulf of Panama´, Panama´; (4) Gala´pagos Islands, Ecuador, were analyzed for six variables: (1) ovum diameter; (2) ovum volume; (3) number of mesenteries per polyp; (4) number

123

Author's personal copy Mar Biol

a

c

b

d

Fig. 3 Psammocora stellata. Reproductive activity at Uva Island, Gulf of Chiriquı´, Panama´, in relation to season (a, b) and lunar phase (c, d), based on 41 collections and 154 colonies examined (1985–2004). Further details as in legend to Fig. 2

a

c

b

d

Fig. 4 Psammocora stellata. Reproductive activity in Gulf of Panama´ (Saboga and Taboga Islands), Panama´, in relation to season (a, b) and lunar phase (c, d), based on 40 collections and 188 colonies examined (1985–1998). Further details as in legend to Fig. 2

Therefore, in some gravid samples with low n, it was not possible to match ova numbers in mesenteries with their corresponding polyps.

A mean ovum diameter of 98.8 lm in Psammocora stellata at Can˜o Island was significantly greater (Kruskal– Wallis rank sum test, p \ 0.001) than in Panama´

123

Author's personal copy Mar Biol

a

c

b

d

Fig. 5 Psammocora stellata. Reproductive activity in Gala´pagos Islands, Ecuador, in relation to season (a, b) and lunar phase (c, d), based on 34 collections and 93 colonies examined (1985–1990). Further details as in legend to Fig. 2

(Uva Island) or the Gala´pagos Islands (Appendix 12). At collection sites in Panama´ (Uva Island and Gulf of Panama´) and the Gala´pagos Islands, mean ovum diameters ranged from 92.5 to 94.7 lm and there was no statistically significant difference between them. Mature ova diameters in Psammocora profundacella at 3 sites (Can˜o Island, Uva Island, and Gala´pagos Islands) ranged from 90.5 to 94.7 lm and were not statistically different (Kruskal–Wallis test, p = 0.234; Appendix 13). Mean gamete-bearing mesenteries in both species ranged between about 9 and 11. In the Gulf of Panama´, P. stellata polyps contained about one additional fecund mesentery than polyps at Uva Island. There was no statistically significant difference in numbers of fecund mesenteries in P. profundacella among localities. Bootstrapped probability distributions were computed for various sample sizes to produce unbiased estimates for the means and standard errors (Appendices 12, 13). Two additional joint probability estimates of reproductive output were subsequently computed by combining certain individual variable estimates consisting of 10,000 bootstrapped trials each to produce unbiased estimates of overall multiplicative mean and standard error. These were number of stage IV ova per polyp (attribute #7, Appendices 12, 13): ova mes2 ova #3  #4  #5 ¼   cs ls polyp mes2 ova ova ¼ #7 ð1Þ ¼  ¼ mes2 polyp polyp

123

and, ova-per-unit surface area (cm2) of colony (attribute #8, Appendices 12, 13): #6  #7 ¼

polyps ova ova  ¼ ¼ #8 colony polyp colony

ð2Þ

A visual example of the sequence of bootstrapped distributions individually generated to compute the synthesized fecundity estimates is shown for Psammocora stellata at Can˜o Island for attribute #7 in Appendix 5 and attribute #8 in Appendix 6. Parametric statistical tests of significance of mean differences in computed fecundity (e.g., attribute #8) for P. stellata and Psammocora profundacella required transformation of the original bootstrapped estimates (Appendices 7, 8) using the Box-Cox procedure. Boxplots of the distributions of the bootstrapped estimates of fecundity (number of stage IV ova cm-2 of colony surface) are shown for the two species (Appendix 14a, b). All among-site fecundity estimates demonstrated highly significant differences for Psammocora stellata (X2 = 19,717, p \ 2.2 9 10-16), with median values ranging from \1,000 ova cm-2 at Uva Island to [3,000 mature ova cm-2 in the Gulf of Panama´ (Appendix 14a). A posteriori testing indicated significant site differences in fecundity in descending order (Appendix 14a): Gulf of Panama´, Can˜o Island, Gala´pagos Islands, Uva Island (Kruskal–Wallis test, p \ 2.2 9 10-16). Psammocora profundacella also demonstrated highly significant intersite

Author's personal copy Mar Biol Table 2 Psammocora stellata Environmental setting location

Samples (n colonies) Totals

Stable thermal regime Can˜o Is., Costa Rica

78

Test

With gametes

With stage IV

Period

33

29

Wet/Dry Lunar

Uva Is., Gulf of Chiriquı´ Panama´

154

88

83

p Gamete stages

I–IV

\0.005

IV

\0.001

I–IV

\0.05

IV

\0.05

Wet/Dry

I–IV

\0.05

Lunar

IV I–IV

\0.05 NS

IV

\0.05

Seasonal upwelling Saboga Is., Gulf of Panama´ Panama´

Taboga Is., Gulf of Panama´ Panama´

Combined, Gulf of Panama´ Panama´

Variable thermal regime Gala´pagos Is., Ecuador

93

95

188

93

19

36

55

42

9

34

43

37

I–IV

\0.001

IV

\0.001

Lunar

I–IV

\0.001

IV

\0.005

Wet/Dry

I–IV

\0.001

Wet/Dry

IV

\0.001

Lunar

I–IV

\0.05

IV

\0.01

Wet/Dry

I–IV

\0.001

IV

\0.001

Lunar

I–IV

NS

IV

NS

I–IV

\0.001

IV

\0.005

Wet/Dry Lunar

I–IV

NS

IV

NS

Seasonal and lunar patterns of gametogenesis at six equatorial eastern Pacific study sites. Fisher exact test employed in all analyses except for Uva Island, and Gulf of Panama´ lunar periods for both stages I–IV and IV, which were tested employing Chi-square analyses. When p B 0.05, reproductive activity was non-random over period tested [I–IV presence of Stages I–IV inclusive; IV presence of Stage IV gametocytes alone; wet/dry wet season, 15 April to 14 December (8 months) and dry season 15 December to 14 April (4 months) (in Gala´pagos, wet and dry seasons occur at opposite times of year to Costa Rica and Panama´); lunar four equal periods with lunar cycle beginning with new moon]

differences in fecundity (X2 = 19,717, p \ 2.2 9 10-16) with a high median value of slightly \4,000 ova cm-2 at Uva Island and a low median of just [2,000 ova cm-2 in the Gala´pagos Islands (Appendix 14b). Three lines of evidence from histological analyses indicated multiple annual spawning cycles at all study sites: (1) a high incidence of mature gonads in most collections, (2) prevalence of spent gonads, and (3) presence of mature testes throughout most of the year in a frequently sampled large colony. First, Stage IV gonads were present in 192 of 218 (88.1 %) samples of Psammocora stellata and in 69 of 77 (89.6 %) samples of Psammocora profundacella (Tables 2, 3). This suggested the presence of several generations of mature gametes in a colony at any given time, indicative of continuous gamete production.

Secondly, spent gonads occurred in several colonies of P. stellata in 7 of 12 calendar months during sampling in the 1980s and 1990s (Appendix 11). Finally, approximately biweekly sampling of a single male colony of P. stellata at Taboga Island (Gulf of Panama´) revealed the presence of Stage IV spermaries over a nearly 8-month non-upwelling period (Fig. 4b). Based on this evidence, probable lunar and annual spawning activities were denoted for both species of Psammocora at the four study sites (Table 4). There was a tendency for spawning to occur around new and full lunar phases, but this was often shifted by a few days as noted previously. Spawning in P. stellata occurred during most of the year and year round at Uva Island, and in P. profundacella from a few to 6 months in the Gala´pagos Islands.

123

Author's personal copy Mar Biol

a

c

b

d

Fig. 6 Psammocora profundacella. Reproductive activity at Can˜o Island, Costa Rica, in relation to season (a, b) and lunar phase (c, d), based on 11 collections and 28 colonies examined (1985–1991). Further details as in legend to Fig. 2 Table 3 Psammocora profundacella Environmental setting location

Samples (n colonies) Totals

Stable thermal regime Can˜o Is., Costa Rica

28

With gametes

7

Test With stage IV

6

Period

Wet/Dry Lunar

Uva Is., Gulf of Chiriquı´ Panama´

Variable thermal regime Gala´pagos Is., Ecuador

59

88

38

32

38

25

p Gamete stages

I–IV

NS

IV

\0.05

I–IV

NS

IV

NS

Wet/Dry

I–IV IV

NS NS

Lunar

I–IV

NS

IV

NS

Wet/Dry Lunar

I–IV

\0.05

IV

NS

I–IV

NS

IV

NS

Seasonal and lunar patterns of gametogenesis at three equatorial eastern Pacific study sites. Fisher exact test employed in all analyses. See caption in Table 2 for details regarding p and durations of seasonal and lunar periods

Recovery Changing abundances of Psammocora stellata at several equatorial eastern Pacific sites, mainly in response to ENSO warming disturbances, have resulted in persistently low to moderate to high population densities (Appendix

123

15). Two shallow nearshore (5–9 m) sites at Can˜o Island, Costa Rica, have demonstrated no increases in live coral cover or number of colonies over a 15-year period following the 1982–1983 and 1997–1998 bleaching events. Two deep (9–14 m) sites demonstrated recovery of live coral cover and colony abundances after 1984, but then

Author's personal copy Mar Biol

a

c

b

d

Fig. 7 Psammocora profundacella. Reproductive activity at Uva Island, Gulf of Chiriquı´, Panama´, in relation to season (a, b) and lunar phase (c, d), based on 9 collections and 59 colonies examined (1997–1998). Further details as in legend to Fig. 2

a

c

b

d

Fig. 8 Psammocora profundacella. Reproductive activity in Gala´pagos Islands, Ecuador, in relation to season (a, b) and lunar phase (c, d), based on 27 collections and 88 colonies examined (1986–1997). Further details as in legend to Fig. 2

experienced declines following the 1997–1998 ENSO event. At Cocos Island, two shallow sites (3–18 m) experienced declines in live cover while there was an increase

at the deep site (9–24 m); overall, there was a minor change in cover (Guzma´n and Corte´s 2007). At Uva Island, Panama´, a 288-m2 shallow (2–3 m) plot monitored since

123

Author's personal copy Mar Biol

1983 demonstrated a gradual increase from 0 to 10.6 colonies m-2 over a 19-year period. Although present in this plot before the 1982–1983 ENSO mortality event, abundances are unknown pre-1982. In the Gala´pagos Islands, an aggregate area of P. stellata of 2,295 m2 at the Devil’s Crown patch reef in 1976 experienced total mortality in 1983 and has since recovered to about one-third of its former cover after 28 years (Appendix 16). A second site in the Gala´pagos (Xarifa Island) has demonstrated substantial increases in live cover (&48 %) and mean number of colonies (&138 %) over a 7-year period (Brown and Feingold unpubl data, Appendix 16).

Discussion The two eastern Pacific Psammocora species in this study (P. stellata and P. profundacella) are gonochoric as are the majority of species in the Siderastreidae (Richmond and Hunter

1990; Baird et al. 2009a; Harrison 2011; Kerr et al. 2011). Indirect evidence suggested that P. stellata in Hawaii was a brooder. Swimming planulae were observed in an aquarium with isolated colonies of P. stellata, which were assumed to be the maternal source colonies of the larvae (Kolinski and Cox 2003). Our study strongly suggests, however, that P. stellata in the eastern Pacific is a broadcast spawner, based on four lines of evidence: (1) observed spawning of colonies, albeit under stressed conditions, (2) presence of ova in sealed polyethylene bags containing corals, (3) presence of male and female spawned gonads in histological preparations taken throughout the calendar year, and (4) absence of planula larvae in hundreds of examined histological samples. Although only a single colony of Psammocora superficialis (= ? profundacella) was sampled at the Solitary Islands in eastern Australia, Wilson and Harrison (2003) also concluded that this species was a broadcast spawner. Both brooding and broadcast spawning corals are known in the Siderastreidae and even within species in the genus

Table 4 Psammocora stellata, Psammocora profundacella Species

Location

Months

No. months year-1

Ova cm-2 year-1

Psammocora stellata

Can˜o I, Costa Rica Uva I, Panama´

Jul–Jan

C7

C18,418 ± 2,731

Jan–Dec

12

12,835 ± 1,294

May–Dec

C8

C13,233 ± 1,414

Nov–Jul

C9

C18,193 ± 2,568 C12,175 ± 2,214

Gulf of Panama´, Panama´ Gala´pagos Is., Psammocora profundacella

Can˜o I, Costa Rica Uva I, Panama´

Dec–Feb

C3

Jan–Mar, Sep–Oct

5

20,338 ± 2,813

Gala´pagos Is.,

Dec, Feb–May

C6

C16,473 ± 2,936

Annual fecundity estimates for all study localities based on number of months per year gonads contained mature gametes (attribute #8 ? 25 % Appendices 12 and 13 9 respective no. months year-1). Mean ovum densities calculated by increasing number of polyps cm-2 by 25 % to compensate for use of non-processed tissue

Fig. 9 Seasonal, lunar, and diel timing of 11 broadcast spawning zooxanthellate corals predominantly at Uva Island reef, Gulf of Chiriquı´, Panama´. Data are from Stage IV gonadal presence in male and female colonies, and spawning observations (Glynn et al. 1991, 1994, 1996, 2000, 2011; Glynn 1999). Diel spawning times of

123

Pocillopora inflata and Pavona gigantea were observed in Gulf of Panama´ and Gala´pagos Islands, respectively. Spawning in Psammocora stellata may occur sometime from 2300 to 0700, based on mature ova observed in sealed bags when collected soon after sunrise

Author's personal copy Mar Biol

Siderastrea (Szmant 1986; Harrison and Wallace 1990; Baird et al. 2009a). In the Caribbean, the mode of reproduction has generally been attributed to colony morphology and environmental conditions (Szmant 1986). Large adult size and predictable environments were hypothesized to favor broadcast spawners whereas brooding species occurred more commonly in unstable habitats where high adult mortality is frequent. This pattern is consistent with Porites in some eastern Pacific areas where the large broadcast spawner Porites lobata typically occurs in predictable deeper reef zones, and Porites panamensis, a small brooding species, characteristically occupies shallow habitats subject to frequent disturbances such as bleaching and extreme low tidal exposures (Glynn 1976; Glynn et al. 1994). At Can˜o Island, however, P. lobata microatolls are abundant on shallow reef flats, and P. panamensis occurs from 2 to 15 m depth (Guzma´n and Corte´s 1989; Guzma´n, personal communication). Psammocora stellata at Uva Island inhabits multiple reef zones including the deep (20 m) forereef rubble plain, while Psammocora profundacella more often occurs on basalt substrates along waveexposed shores. In the Gala´pagos Islands and the Gulf of Panama´, P. stellata is also found in deeper reef habitats. Both Psammocora species conform to the protracted seasonal reproductive patterns previously reported for zooxanthellate corals in the equatorial eastern Pacific (Fig. 9, e.g., Glynn et al. 2011). In thermally stable environments (Can˜o and Uva Islands), seasonal reproductive activity is mostly protracted throughout the year. In environments subject to seasonal upwelling (Gulf of Panama´), or marked annual variations in temperature (Gala´pagos Islands), reproductive activity was most pronounced during warm periods. This pattern of gamete development over extended reproductive periods, perhaps simply a response to sea temperature variation, may enhance survival in disturbed environments (Glynn et al. 1991, 1994, 1996, 2000, 2011; Colley et al. 2006). After considerable effort over many years, neither species of Psammocora was observed spawning under natural conditions. The documented negative results reported here, however, should serve to narrow the timing of spawning that will hopefully be determined in future studies. The presence of mature male and female gametes at all lunar phases at all study sites (but one) suggests that spawning may be diffuse and not closely linked to any particular lunar phase. Psammocora profundacella at Uva Island was the one exception with mature gametes clustered over a 7-day period near full moon (lunar days 14–20). It is also possible, however, that this pattern is an artifact due to limited sampling. Multispecific synchronous spawning has not been observed in the eastern Pacific, adding to the diversity of biogeographic reproductive patterns reported in coral

assemblages (Guest et al. 2005; Baird et al. 2009a; Harrison 2011). Even though mature gametes in several species, suggestive of impending spawning, were clustered between lunar days 15 and 19 (Fig. 9), no simultaneous group spawning was ever observed. The closest approach to spawning synchrony in the eastern Pacific occurs in Pavona varians and Pavona chiriquiensis, which spawn during full moon, but 12 h out of phase (Glynn et al. 2000). This pattern of temporal reproductive isolation of 11 broadcast spawning species is similar to that reported for the northern-most Red Sea (Shlesinger and Loya 1985; Shlesinger et al. 1998). More recent studies in Kenya (Mangubhai and Harrison 2009), Japan, and the Great Barrier Reef (Baird et al. 2009b) have added to our knowledge of latitudinal differences in reproductive behavior and called into question the prevalence of mass spawning events. Patterns of seasonal spawning duration and synchrony have been shown to vary geographically; the equatorial eastern tropical Pacific exhibits a complete lack of synchrony and a greatly extended spawning season as compared to other regions of the world. Ongoing studies at higher latitudes in the Gulf of Tehuantepec, Me´xico, a seasonally upwelling environment, report a relatively narrow reproductive period of 4 months for two broadcasting species—Pocillopora damicornis, Pavona gigantea—and one brooding species—Porites panamensis (Rodrı´guez-Troncoso et al. 2011). Egg maturation occurred in the two broadcast spawning species in August, about 4 months after upwelling when annual sea temperature peaks. During the very strong 1997–1998 ENSO in the non-upwelling Gulf of Chiriquı´, the numbers of reproductive colonies of P. damicornis were significantly less than in non-ENSO years (Colley et al. 2006). However, Psammocora stellata collections contained female and male reproductive colonies in the Gala´pagos cool season in 1987, during a moderately strong El Nin˜o event. These results underline the importance of site-specific environmental differences (e.g., seasonal temperature variability and tidal amplitude), interannual variability of oceanographic conditions (e.g., ENSO and Pacific Decadal Oscillation), and geographic population difference in determining gametogenesis and spawning. Psammocora spp. appear to possess four adaptive characters that would facilitate recovery following ENSOrelated mortality: (1) relative resistance to bleaching/mortality, especially below 10–12 m depth, (2) persistence of surviving deep source populations that can potentially promote recruitment into decimated shallow reef habitats, (3) prolonged seasonal reproductive periods and high fecundities, (4) asexual reproduction. Compared to most zooxanthellate corals, Psammocora stellata experienced

123

Author's personal copy Mar Biol

less bleaching/mortality in Panama´ (Glynn 1983, 1984) and the Gala´pagos Islands (Robinson 1985) during the severe 1982–1983 ENSO warming event, and in Costa Rica (Jime´nez et al. 2001) and Me´xico (Carriquiry et al. 2001) during the equally strong 1997–1998 ENSO. Bezy et al. (2006) have offered evidence of the relative resistance of Psammocora spp. to stress and remarked on their ability to recolonize disturbed areas in Costa Rica. This also was observed at Gorgona Island, Colombia, following the 1982–1983 ENSO (Guzma´n and Lo´pez 1991). Deep (10–25 m) occurring colonies in the Gala´pagos Islands bleached but suffered minimal mortality and regained normal pigmentation in 5–6 months following the disturbance event (Glynn 1990; Feingold 1996). This depthrelated survivorship pattern also was observed in Baja California (Reyes-Bonilla 2001). Both Psammocora species are highly fecund with annual ova production ranging from 1.2 9 104 to 1.8 9 104 cm-2 at all localities. These values greatly exceed all Indo-Pacific and Caribbean species listed in Harrison and Wallace (1990) and are similar to those reported for other eastern Pacific poritid and agariciid species (Glynn et al. 1994, 1996, 2000, 2011). Finally, fragmentation resulting from pufferfish corallivory could increase asexual reproduction (Guzma´n and Lo´pez 1991; Feingold 1996). The relative contributions of sexual and asexual reproduction to the post-disturbance recovery of Psammocora populations are unknown. The small sizes and mobile habit of eastern Pacific Psammocora, especially P. stellata, make it difficult to distinguish between sexual and asexual (fragmentation) recruitment (Glynn 1974). Some studies, however, are suggestive of an important role of sexual larval recruitment from surviving deep populations. In Costa Rica (Can˜o Island), P. stellata was more abundant in deep compared with shallow habitats, and Guzma´n and Corte´s (2001) noted that sexual recruitment was responsible for increases in abundance at Can˜o Island after the 1992 ENSO disturbance. Jime´nez and Corte´s (2003) also observed high mortality of shallow-occurring P. stellata at Costa Rican mainland sites and suggested that recovery would depend on the recruitment of sexual propagules from deep populations. A slow but steady increase in P. stellata on the shallow Uva reef, primarily on rubble substrates previously lacking this species, occurred after the 1997–1998 ENSO. Less than 100-m downslope, at 12–18 m on the forereef, was a large population of P. stellata that survived both ENSO events and likely served as a source population. The recovery of shallow (2–3 m) P. stellata patches within Devil’s Crown in the Gala´pagos Islands also appears to be due to settlement of larvae originating from nearby (15–25 m) upstream source populations at 15 to 25 m depth (Feingold 1996, 2001). A molecular genetic

123

analysis aimed at establishing the degree of connectivity between deep and shallow corals should help to clarify this question. Additionally, determination of relatedness of colonies at various spatial scales would help pinpoint if population increases are primarily due to asexual or sexual reproduction or some combination of the two processes. The reproductive biology of equatorial eastern Pacific scleractinian coral communities is different from most other regions of the world. Although only a few species have been seen releasing gametes, most are broadcast spawners, but do not form bundles except possibly for the Pocilloporidae. Each species has its own seasonality that can vary with locale. To date, only two species are proven brooders, two species are simultaneous hermaphrodites (Pocilloporidae), four species are sequential cosexual hermaphrodites (Agariciidae), and five species display stable gonochorism including one agariciid (Pavona clavus), two species of Psammocora (this study), one poritid (Porites lobata), and one fungiid (Diaseris distorta) (Glynn et al. 1991, 1994, 1996, 2000, 2008, 2011; Colley et al. 2000). Recent molecular genetic studies indicate that coral sexual traits in several taxa demonstrate strong phylogenetic relationships. For example, the two brooding species, Tubastraea coccinea (azooxanthellate) and Porites panamensis (zooxanthellate, endemic), are sister-group species in clades II and III; Pavona and Porites are members of the Complexa clade; and Psammocora and Diaseris are members of clade XI (Fukami et al. 2008; Kitahara et al. 2010; Huang 2012). Therefore, it follows that reproductive mode and gamete development may be more closely related to phylogeny than to environmental conditions. Temperature stability and its variation may, however, be an important factor in the success or completion of gamete cycling, and due to upwelling and ENSO effects, may show a varied pattern in at least some species on a yearly basis. Acknowledgments We thank Adrienne M. S. Correa, Peggy Fong, Christiane Hueerkamp, Priscilla Martı´nez, and Fernando Rivera for help in the field, and Rebecca Ball-Bailey, Kathy Black, Kathryn Brown, Erin Kapostasy, Susan Laessig, Alison Moulding, Juan Pen˜a, David Smith, Joy H Ting, Bonnie Tucker, and Joyce A Yager for various analytical tasks in the laboratory. Francesca Benzoni kindly assisted in the identification of Psammocora species. The following host countries, agencies, and institutions granted permission and variously assisted during the course of this study: Costa Rica, Centro de Investigaciones en Ciencias del Mar y Limnologı´a (CIMAR), ´ reas de ConUniversidad de Costa Rica, and Sistema Nacional de A servacio´n, Ministerio del Ambiente y Energı´a; Panama´, Departamento de Biologı´a Acua´tica (Universidad de Panama´), Autoridad Nacional del Ambiente (ANAM), and Smithsonian Tropical Research Institute; Ecuador, Charles Darwin Research Station, and the Gala´pagos National Park Service. Research support was provided by the U. S. National Science Foundation, Biological Oceanography Program, grant OCE-0526361 and earlier awards.

Author's personal copy Mar Biol

References Alder J (2010) R in a nutshell: a desktop quick reference. O’Reilly Media, Sebastopol Baird AH, Guest JR, Willis BL (2009a) Systematic and biogeographical patterns in the reproductive biology of scleractinian corals. Annu Rev Ecol Evol Syst 40:551–571 Baird AH, Birrel CL, Hughes TP, McDonald A, Nojima S, Page CA, Prachett MS, Yamasaki H (2009b) Latitudinal variation in reproductive synchrony in Acropora assemblages: Japan vs. Australia. Galaxea 11:101–108 Benzoni F, Stefani F, Stolarski J, Pichon M, Mitta G, Galli P (2007) Debating phylogenetic relationships of the scleractinian Psammocora : molecular and morphological evidences. Contrb Zool 76:35–54 Benzoni F, Stefani F, Pichon M, Galli P (2010) The name game: morpho-molecular species boundaries in the genus Psammocora (Cnidaria, Scleractinia). Zool J Linn Soc 160:421–456 Bezy MB, Jime´nez C, Corte´s J, Segura A, Leo´n A, Alvarado JJ, Gille´n C, Mejı´a E (2006) Contrasting Psammocora-dominated coral communities in Costa Rica, tropical eastern Pacific. In: Proceedings of 10th international coral reef symposium, Okinawa, vol 1, pp 376–381 Carpizo-Ituarte E, Vizcaı´no-Ochoa V, Chi-Barraga´n G, Tapia-Va´zquez O, Cupul-Magan˜a AL, Medina-Rosas P (2011) Evidence of sexual reproduction in the hermatypic corals Pocillopora damicornis, Porites panamensis, and Pavona gigantea in Banderas Bay, Mexican Pacific. Cienc Mar 37:97–112 Carriquiry JD, Cupul-Magan˜a AL, Rodrı´guez-Zaragoza F, MedinaRosas P (2001) Coral bleaching and mortality in the Mexican Pacific during the 1997–98 El Nin˜o and prediction from a remote sensing approach. Bull Mar Sci 69:237–249 Cha´vez-Romo HE, Reyes-Bonilla H (2007) Sexual reproduction of the coral Pocillopora damicornis in the southern Gulf of California, Mexico. Cienc Mar 33:495–501 Colley SB, Feingold JS, Pen˜a J, Glynn PW (2000) Reproductive ecology of Diaseris distorta (Michelin) (Fungiidae) in the Gala´pagos Islands, Ecuador. In: Proceedings of 9th international coral reef symposium, vol 1, pp 373–379 Colley SB, Glynn PW, May AS, Mate´ JL (2006) Species-dependent reproductive responses of eastern Pacific corals to the 1997–1998 ENSO event. In: Proceedings of 10th international coral reef symposium, vol 1, pp 61–70 Corte´s J (1990) The coral reefs of Golfo Dulce, Costa Rica: distribution and community structure. Atoll Res Bull 344:1–37 Crawley M (2005) Statistics: an introduction using R. Wiley, England Davidson A, Hinkley D (1997) Bootstrap methods and their application. Cambridge Univ Press, New York Edgar GJ, Farin˜a JM, Calvopin˜a M, Martı´nez C, Banks S (2002) Comunidades submareales rocosas II: Peces y macroinvertebrados mo´viles. In: Danulat E, Edgar GJ (eds) Reserva marina de Gala´pagos. Lı´nea base de la biodiversidad. Fundacio´n Charles Darwin and Servicio Parque Nacional Gala´pagos, Santa Cruz, pp 68–97 Efron B, Tibshirani RJ (1998) An introduction to bootstrap. Chapman & Hall, Boca Raton Feingold JS (1996) Coral survivors of the 1982–1983 El Nin˜oSouthern Oscillation, Gala´pagos Islands, Ecuador. Coral Reefs 15:108 Feingold JS (2001) Responses of three coral communities to the 1997–1998 El Nin˜o-Southern Oscillation: Gala´pagos Islands, Ecuador. Bull Mar Sci 69:61–77 Fukami H, Chen CA, Budd AF, Collins A, Wallace C, Chuang Y–Y, Chen C, Dai C-F, Iwao K, Sheppard C, Knowlton N (2008) Mitochondrial and nuclear genes suggest that stony corals are

monophyletic but most families of stony corals are not (Order Scleractinia, Class Anthozoa, Phylum Cnidaria). PLoS ONE 3:e3222. doi:10.1371/journal.pone.0003222 Glynn PW (1974) Rolling stones among the Scleractinia: mobile coralliths in the Gulf of Panama´. In: Proceedings of 2nd international coral reef symposium, Brisbane, vol 2, pp 183–198 Glynn PW (1976) Some physical and biological determinants of coral community structure in the eastern Pacific. Ecol Monogr 46:431–456 Glynn PW (1983) Extensive ‘bleaching’ and death of reef corals on the Pacific coast of Panama´. Environ Conserv 10:149–154 Glynn PW (1984) Widespread coral mortality and the 1982–1983 El Nin˜o warming event. Environ Conserv 11:133–146 Glynn PW (1990) Coral mortality and disturbances to coral reefs in the tropical eastern Pacific. In: Glynn PW (ed), Global ecological consequences of the 1982–1983 El Nin˜o-Southern Oscillation. Oceanogr Ser 52. Elsevier Amsterdam, pp 55–126 Glynn PW (1994) State of coral reefs in the Gala´pagos Islands: natural vs anthropogenic impacts. Mar Pollut Bull 29:131–140 Glynn PW (1999) Pocillopora inflata, a new species of scleractinian coral (Cnidaria: Anthozoa) from the tropical eastern Pacific. Pac Sci 53:168–180 Glynn PW, Ault JS (2000) A biogeographic analysis and review of the far eastern Pacific coral reef region. Coral Reefs 19:1–23 Glynn PW, Fong P (2006) Patterns of reef coral recovery by the regrowth of surviving tissues following the 1997–1998 El Nin˜o warming and 2000, 2001 upwelling cool events in Panama´, eastern Pacific. In: Proceedings of 10th international coral reef symposium, Okinawa, vol 2, pp 624–630 Glynn PW, Wellington GM (1983) Corals and coral reefs of the Gala´pagos Islands. University of California Press, Berkeley Glynn PW, Gassman NJ, Eakin CM, Corte´s J, Smith DB, Guzma´n HM (1991) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama´, and the Gala´pagos Islands (Ecuador). I. Pocilloporidae. Mar Biol 109:355–368 Glynn PW, Colley SB, Eakin CM, Smith DB, Corte´s J, Gassman NJ, Guzma´n HM, Del Rosario JB, Feingold JS (1994) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama´, and the Gala´pagos Islands (Ecuador). II. Poritidae. Mar Biol 118:191–208 Glynn PW, Colley SB, Gassman NJ, Black K, Corte´s J, Mate´ JL (1996) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama´, and the Gala´pagos Islands (Ecuador). III. Agariciidae (Pavona gigantea and Gardineroseris planulata). Mar Biol 125:579–601 Glynn PW, Colley SB, Ting JH, Mate´ JL, Guzma´n HM (2000) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama´, and the Gala´pagos Islands (Ecuador). IV. Agariciidae, recruitment and recovery of Pavona varians and Pavona sp. A. Mar Biol 136:785–805 Glynn PW, Colley SB, Mate´ JL, Corte´s J, Guzman HM, Bailey RL, Feingold JS, Enochs IC (2008) Reproductive ecology of the azooxanthellate coral Tubastraea coccinea in the equatorial eastern Pacific: part V. Dendrophylliidae. Mar Biol 153:529–544 Glynn PW, Riegl B, Correa AMS, Baums IB (2009) Rapid recovery of a coral reef at Darwin Island, Gala´pagos Islands. Gala´pagos Res 66:6–13 Glynn PW, Colley SB, Guzman HM, Enochs IC, Corte´s J, Mate´ JL, Feingold JS (2011) Reef coral reproduction in the eastern Pacific: Costa Rica, Panama´, and the Gala´pagos Islands (Ecuador). VI. Agariciidae, Pavona clavus. Mar Biol 158:1601–1617 Guest JR, Baird AH, Goh BPL, Chou LM (2005) Seasonal reproduction in equatorial coral reefs. Inv Repr Dev 48:207–218 Guzma´n HM, Corte´s J (1989) Coral community structure at Can˜o Island, Pacific Costa Rica. Pubb Staz Zool Napoli (I: Mar Ecol) 10:21–43

123

Author's personal copy Mar Biol Guzma´n HM, Corte´s J (2001) Changes in reef community structure after fifteen years of natural disturbances in the eastern Pacific (Costa Rica). Bull Mar Sci 69:133–149 Guzma´n HM, Corte´s J (2007) Reef recovery 20 years after the 1982–1983 El Nin˜o massive mortality. Mar Biol 151:401–411 Guzma´n HM, Lo´pez JD (1991) Diet of the corallivorous pufferfish Arothron meleagris (Tetraodontidae) at Gorgona Island, Colombia. Rev Biol Trop 39:203–206 Harris MP (1969) Breeding seasons of seabirds in the Gala´pagos Islands. J Zool Lond 159:145–165 Harrison PL (2011) Sexual reproduction of scleractinian corals. In: Dubinsky Z, Stambler N (eds) Coral reefs: an ecosystem in transition. Springer, Berlin, pp 59–85 Harrison PL, Wallace CC (1990) Reproduction, dispersal and recruitment of scleractinian corals. In: Dubinsky Z (ed) Coral reefs, ecosystems of the world 25. Elsevier, Amsterdam, pp 133–207 Huang D (2012) Threatened reef corals of the world. PLoS ONE 7:e34459. doi:10.1371/journal.pone.0034459 Jime´nez CE, Corte´s J (2003) Coral cover change associated to El Nin˜o, eastern Pacific, Costa Rica, 1992–2001. PSZN Mar Ecol 24:179–192 Jime´nez C, Corte´s J, Leo´n A, Ruiz E (2001) Coral bleaching and mortality associated with El Nin˜o 1997/1998 event in an upwelling environment in the eastern Pacific (Gulf of Papagayo, Costa Rica). Bull Mar Sci 69:151–169 Kerr AM, Baird AH, Hughes TP (2011) Correlated evolution of sex and reproductive mode in corals (Anthozoa: Scleractinia). Proc R Soc B 278:75–81. doi:10.1098/rspb.2010.1196 Kitahara MV, Cairns SD, Stolarski J, Blair D, Miller DJ (2010) A comprehensive phylogenetic analysis of the Scleractinia (Cnidaria, Anthozoa) based on mitochondrial CO1 sequence data. PLoS ONE 5:e11490. doi:10.1371/journal.pone.0011490 Kohler KE, Gill SM (2006) Coral point count with excel extensions (CPCe): a visual basic program for the determination of coral and substrate coverage using random point count methodology. Comput Geosci 32:1259–1269 Kolinski SP, Cox EF (2003) An update on modes and timing of gamete and planula release in Hawaiian scleractinian corals with implications for conservation and management. Pac Sci 57:17– 27 Kutner MH, Nachtsheim CJ, Neter J, Li W (2005) Applied linear statistical models, 5th edn. McGraw-Hill & Irwin, Boston Lo´pez-Pe´rez RA, Mora-Pe´rez MG, Leyte-Morales GE (2007) Coral (Anthozoa: Scleractinia) recruitment at Bahı´as de Huatulco, western Mexico: implications for coral community structure and dynamics. Pac Sci 61:355–369 Luna JG (ed) (1968) Manual of histologic staining methods of the armed forces institute of pathology, 3rd edn. McGraw-Hill, New York Mangubhai S, Harrison PL (2009) Extended breeding seasons and asynchronous spawning among equatorial reef corals in Kenya. Mar Ecol Prog Ser 374:305–310 Manly BJ (2007) Randomization, bootstrap and Monte Carlo methods in biology, 3rd edn. Chapman & Hall/CRC, Boca Raton Mora-Pe´rez MG (2005) Biologı´a reproductiva del coral Porites panamensis Verrill 1866 (Anthozoa: Scleractinia) en Bahı´a de La Paz, Baja California Sur, Me´xico. MSc thesis, Centro

123

Interdisciplinario de Ciencias Marinas, Instituto Polite´cnico Nacional, Me´xico Podesta´ GP, Glynn PW (2001) The 1997–1998 El Nin˜o event in Panama´ and Gala´pagos: an update of thermal stress indices relative to coral bleaching. Bull Mar Sci 69:43–59 Qian SS (2010) Environmental and ecological statistics with R. Chapman & Hall/CRC, Boca Raton Reyes-Bonilla H (2001) Effects of the 1997–1998 El Nin˜o-Southern Oscillation on coral communities of the Gulf of California, Mexico. Bull Mar Sci 69:251–266 Reyes-Bonilla H (2002) Check list of valid names and synonyms of stony corals (Anthozoa: Scleractinia) from the eastern Pacific. J Nat Hist 36:1–13 Richmond RH, Hunter CL (1990) Reproduction and recruitment of corals: comparisons among the Caribbean, the tropical Pacific, and the Red Sea. Mar Ecol Prog Ser 60:185–203 Robinson G (1985) Influence of the 1982–1983 El Nin˜o on Gala´pagos marine life. In: Robinson G, del Pino EM (eds) El Nin˜o in the Gala´pagos Islands: the 1982–1983 event. Fundacio´n Charles Darwin para las Islas Gala´pagos, Quito, pp 153–190 Rodrı´guez-Troncoso AP (2006) Ciclo reproductivo de tres especies formadoras del arrecife La Entrega, Oaxaca, Me´xico. MSc thesis, Universidad Auto´noma de Baja California, Mexico Rodrı´guez-Troncoso AP, Carpizo-Ituarte E, Leyte-Morales GE, ChiBarraga´n G, Tapia-Va´zquez O (2011) Sexual reproduction of three coral species from the Mexican south Pacific. Mar Biol. doi:10.1007/s00227-001-1765-9 Shlesinger Y, Loya J (1985) Coral community reproductive patterns: Red Sea versus the Great Barrier Reef. Science 228:1333–1335 Shlesinger Y, Goulet TL, Loya Y (1998) Reproductive patterns of scleractinian corals in the northern Red Sea. Mar Biol 132:691– 701 Smith DB (1991) The reproduction and recruitment of Porites panamensis Verrill at Uva Island, Pacific Panama´. MS thesis, Univ. Miami, FL Stefani F, Benzoni F, Pichon M, Cancelliere C, Galli P (2008) A multidisciplinary approach to the definition of species boundaries in branching species of the coral genus Psammocora (Cnidaria, Scleractinia). Zool Scr 37:71–91 Szmant AM (1986) Reproductive ecology of Caribbean reef corals. Coral Reefs 5:43–54 ´ ngel B, Zapata FA, Herna´ndez H, Jime´nez JM (2001) Coral Vargas-A and coral reef responses to the 1997–1998 El Nin˜o event on the Pacific coast of Colombia. Bull Mar Sci 69:111–132 Vizcaı´no-Ochoa VE (2003) Biologı´a reproductiva de tres especies de corales formadores de arrecifes en Bahı´a Banderas, Me´xico. Ms thesis, Univ Auton Baja Calif Wellington GM, Strong AE, Merlen G (2001) The 1997–1998 event in Panama´ and Gala´pagos: an update of thermal stress indices relative to coral bleaching. Bull Mar Sci 69:27–42 Wells JW (1956) Scleractinia. In: Moore RC (ed) Treatise on invertebrate palaeontology. Pt F Coelenterata. Geol Soc Am, University of Kansas Press, Lawrence, pp 328–444 Wilson JR, Harrison PL (2003) Spawning patterns of scleractinian corals at the Solitary Islands–a high latitude coral community in eastern Australia. Mar Ecol Prog Ser 260:115–123 Zar JH (1999) Biostatistical analysis, 4th edn. Prentice Hall, Upper Saddle River

Article Title:  Reef coral reproduction in the equatorial eastern Pacific: Costa Rica, Panamá, and the  Galápagos Islands (Ecuador). VII. Siderastreidae, Psammocora stellata and Psammocorapro fundacella    Journal Title:  Marine Biology, Springer‐Verlag    Authors:  P. W. Glynn, S. B. Colley, J. L. Maté, I. B. Baums, J. S. Feingold, J. Cortés, H. M. Guzman, J. A.  Afflerbach, V. W. Brandtneris, J. S. Ault      P.W. Glynn, J.A. Afflerbach, V. W. Brandtneris, J.S. Ault  Division of Marine Biology and Fisheries  Rosenstiel School of Marine and Atmospheric Science  University of Miami, 4600 Rickenbacker Causeway  Miami, FL 33149, USA  e‐mail: [email protected]    S.B. Colley  Louisiana Applied Coastal Engineering & Science Division  Office of Coastal Protection and Restoration  450 Laurel St., Ste. 1200  Baton Rouge, LA 70801    J.L. Maté, H.M. Guzmán  Smithsonian Tropical Research Institute  PO Box 0843‐03092, Balboa, Ancon, Republic of Panamá    I.B. Baums  Department of Biology  The Pennsylvania State University  208 Mueller Laboratory  University Park, PA 16802    J.S. Feingold  Nova Southeastern University Oceanographic Center  8000 North Ocean Drive, Dania Beach, FL 33004, USA    J. Cortés  Centro de Investigación en Ciencias del Mar y Limnología (CIMAR), and Escuela de Biología, Ciudad de  Investigación, Universidad de Costa Rica, San Pedro, 11051‐2060 San José, Costa Rica    Content Description:  This electronic appendix contains additional information on gamete descriptions  and development and the statistical analyses employed in the assessment of Psammocorafecundity.   Photographs of live colonies are also included. 

Appendix 1  Psammocora stellata colonies on loose rubble substrate at Uva Island reef, Gulf of Chiriquí,  Panamá (4 m depth, 7 March 2012).  Larger colony is approximately 10 cm in length across horizontal  axis of photograph.   

  Appendix 2  Psammocorapro fundacella colonies attached to basalt substrate at Uva Island, Coral  Collection Point (300m N of reef), Gulf of Chiriquí, Panamá (5 m depth, 8 March 2012).  Larger colony is  approximately 10 cm in length along vertical axis of photograph.   

 

Appendix 3  Psammocora stellata, Psammocorapro fundacella.  Summary of sampling information from  major study sites in Costa Rica, Panamá, and Galápagos Islands.  Years sampled: when separated by an  en dash, all intervening years also sampled.    Psammocora stellata  No. months  No.  No.  Study sites  Years  collected  collections  colonies  Costa Rica          Caño Island  1985—89  11  25  78   

 

 

 

 

Panamá  Gulf of Chiriquí  Uva Island 

    1985—90, 92,  97, 98, 2004 

    12 

    41 

    154 

 

 

 

 

 

    1985, 89, 90,  7  92—95, 97, 98 

  18 

  93 

 

 

 

 

1990—91 

12 

22 

95 

 

 

 

 

 

Ecuador  Galápagos Islands  Itabaca, Santa Cruz,  Punta Estrada, Baltra,  Academy Bay, Floreana 

    1985—90 

    11 

    34 

    93 

Gulf of Panamá  Saboga   

Taboga 

 

 

 

 

 

Grand Total    Psammocoraprofundacella 

 

 

[140] 

[513] 

Study sites 

Years 

Costa Rica  Caño Island 

  1985, 86, 88,  89, 91 

No. months  collected    7 

No.  collections    11 

No.  colonies    28 

 

 

 

 

 

Panamá  Gulf of Chiriquí  Uva Island 

    1997—98 

    4 

    9 

    59 

 

Gulf of Panamá  Saboga 

 

 

 

 

  1992, 93, 95 

  3 

  4 

  13 

 

 

 

 

 

Ecuador  Galápagos Islands  Itabaca, Santa Cruz,  Punta Estrada, Baltra,  Academy Bay, Floreana 

    1986—92, 97 

    11 

    27 

    88 

 

 

 

 

 

Grand Total   

 

 

[51] 

[188] 

Appendix 4  Bootstrap Method for Estimating Standard Errors  In the bootstrap, the original sample was assumed to represent the population from which it was  drawn, so re‐samples from this sample population represent approximations of what we would get if we  took many samples from the population.  The response variable(s) y = y1 , K , y n  are considered  independent data points, from which we can recompute a statistic of interest θ ( y1 , K , y n ) .  On  average, about 2/3 of data points from the original data set will be included in a bootstrap sample (Qian  2010).  This step is repeated B times to obtain B boostrap samples.  Corresponding to each bootstrap  sample y *b , the statistics  θ y *b  can be calculated, such as the estimated bootstrap standard error: 

( ) ∑ (θ (y ) − θ )   B

 

seˆ =

b =1

*b

*

(B − 1)

The bootstrap distribution of a statistic based on many re‐samples represents an approximation of the  sampling distribution of the statistic.  From those re‐samples, accuracy measures such as standard errors  and confidence intervals were estimated and used to draw conclusions about the data.    The bootstrap procedure provides the best and most robust approximation of the confidence interval  for the mean response.  The mean and standard error values in the table are identical to the raw data  statistics based on the particular sample size shown in parenthesis.  This obviates the need to report  both original sample and bootstrapped estimates.   

Histogram of boot.csova

0 500 9

10

11

12

13

14

15

1.2

1.3

1.4

1.5

1.6

1.7

Histogram of boot.lsova

Histogram of boot.total

0

500

Frequency

500

1000 1500

boot.csova

1000 1500

boot.mesen

0

Frequency

1500

Frequency

1500 500 0

Frequency

2500

Histogram of boot.mesen

6

7

8

9

10 11 12 13

boot.lsova

100

150

200

250

boot.total

      Appendix 5  Psammocora stellata, Caño Island.  Computational distributions for the individual variables  required to compute item #7 (number of stage IV ova per polyp = boot.total).  Item #7 is the  compounded distribution of items #3 x #4 x #5 = mesen x csova x lsova.     

   

1500 500 0

Frequency

Histogram of boot.polyp

12.0

12.5

13.0

13.5

14.0

14.5

15.0

boot.polyp

1500 500 0

Frequency

Histogram of boot.total

100

150

200

250

boot.total

1000 400 0

Frequency

Histogram of boot.total2

1500

2000

2500 boot.total2

3000

3500

    Appendix 6  Psammocora stellata, Caño Island.  Computational distributions for the individual variables  required to compute item #8 (number of stage IV ova per cm2 of colony = boot.total2).  Item #8 is the  compounded distribution of items #6 x #7 = polyps x boot.total = #8.  Note high skewness in output  response variable (#8) in lower panel.         

    Appendix 7  Statistical diagnostics showing the efficacy of distributional transformation to normality for  computed Psammocora stellata fecundity estimates (#8, number of stage IV ova per cm2 of colony) from  Caño Island, Costa Rica: (upper panels) qqnorm and;(lower panels) Box‐Cox transformed response  variable.   

Appendix 8Optimal Box‐Cox transformation parameters  λ  to convert the original computed response  ~ variable Y, where  Y = Y λ , to achieve normality of the bootstrapped probability distribution required  for various statistical testing.      Untransformed p  Transformed p  Psammocora stellata  Box‐Cox  λ   Caño Island  0.1573  2.003 e‐15  0.7974  Uva Island  0.2657  4.300 e‐08  0.9111  Gulf of Panamá  0.0589  1.042 e‐07  0.4623  Galápagos Islands  ‐0.2315  2.200 e‐16  0.8329  Psammocoraprofundacella        Caño Island  0.4100  9.928 e‐13  0.3426  Uva Island  0.8910  0.0676  0.1344  Galápagos Islands  ‐0.0777  2.200 e‐16  0.1846 

Appendix 9Gamete Development  Described colors are a result of staining with modified Heidenhain’s aniline‐blue.  Swollen mesenteries  indicated early reproductive activity in female colonies.  Clusters of enlarged interstitial cells were the  precursors of Stage I oocytes.  They were spherical and small with a red nucleolus surrounded by white  cytoplasm.  Stage I oocytes ranged in size from about 9 to 24μm (Appendix 10a, b).  Although usually  round to oval, olive leaf‐shaped oocytes were often present and surrounded by a thin, blue mesoglea.   Later, the leaf form was obscured by growth in all directions.  The color of the oocytes depended heavily  on the amount of stain absorbed.  If the tissue appeared orange, Stage I oocytes were brown‐gray in  color.  If the tissue stained reddish, oocytes appeared to have a brownish‐rose tone.  A high degree of  blue stain led to blue‐gray Stage I oocytes.   Stage I oocytes often appeared translucent with the  cytoplasm consisting of fine granules.  In early Stage I oocytes, the nucleus was frequently not  distinguishable from the cytoplasm in color or texture.  Later the nucleus became detectable as white,  round, free of granules and with a large bright red nucleolus.  Stage I oocytes occurred as a cluster of  cells in a mesentery on both sides of the mesoglea or intermingled with older oocyte stages.    Stage II oocytes generally ranged in size from 24 to 65μm (Appendix 10a, b).  They were round  to oval in shape, however, when approaching Stage III the oocytes became deformed owing to space  constraints in the mesenteries.  At such times, the oocyte appeared as a teardrop.  Like Stage I oocytes,  Stage II oocytes appeared translucent or lightly stained.  If the tissue stained red, the oocytes appeared  beige and became more reddish in color when approaching Stage III.  If the tissue was only slightly  stained, the oocytes appeared grayish‐blue.  During this stage, the cytoplasm contained numerous  vesicles, often imparting a grainy texture ranging from white to red and pink.  The nucleus, located in  the center or at one side, remained white and empty in appearance.  The nuclear membrane was well  defined, and a nucleolus was generally visible, appearing as a perfectly round, bright red sphere.  Stage II  oocytes were often adhering to older oocytes, resulting in distorted shapes.  The mesoglea was clearly  visible at this stage as a blue line surrounding the oocyte whose boundary appeared to be a thin black  line.    Stage III oocytes ranged from about 60 to 100μm in diameter, occasionally as small as 45μm and  as large as 120μm (Appendix 10a—d).  Uncrowded oocytes were round to oval, but rectangular at high  densities.  Stages III and IV were always much darker in color than I and II oocytes, mostly pink to bright  red.  Early Stage III oocytes sometimes stained pinker while Stage IV ova were more brick red.  The  grains in the cytoplasm were now more evenly sized and colored.  If the staining was extensive, red  grains were visible.  The nucleus was always centered when visible.  The nuclear membrane was still well  defined, but began to appear reticulated. Multiple nucleoli (up to 10) were occasionally present with  one large nucleolus surrounded by smaller satellite nucleoli.  The mesoglea in Stage III oocytes and Stage  IV ova was normally not detectable.  The defining characteristic of a Stage III oocyte was its thick border.   This dense border, up to 3μm thick, always formed around the oocyte perimeter and stained dark  brown, red or pink.  Harrison and Wallace (1990) described a similar process, involving the formation of  a cortical layer inside the vitelline membrane.  The cortical layer was usually just a shade darker than the  cytoplasm.    Stage IV ova were the largest in diameter, ranging from 90 to 135 μm (Appendix 10a—d).  They  displayed a variety of forms, from oval, rectangular to teardrop shaped or irregular due to space  limitation in the mesentery.  Ova coloration was similar to that of Stage III oocytes.  The cytoplasm was  by now transformed into yolk, consisting of fine, evenly‐spaced granules with a smooth appearance.  In  late Stage IV ova, the nucleus, if visible, was always present at one side of the ovum.  The texture of the  nucleus was always finer than that of the yolk, staining gray and easily discernible against the dark‐red  yolk.  Nucleoli were occasionally detectable, sometimes at the immediate edge of the nucleus.  As in  Stage III oocytes, the mesoglea, which still surrounds the ovum, is so thin as to be undetectable.  The  dense border, which started to thicken in Stage III oocytes, was well developed as a dark rim around the 

ovum and ranged in size from about 3 to 6 μm.  Spent ovaries were observed in several histological  samples of Psammocora stellata from Caño Island, Uva Island and the Galápagos Islands, and in one  colony of Psammocoraprofundacella collected at Caño Island (Appendix 11).    The earliest Stage I spermaries were a cluster of spermatogonia that were between 4 and 24μm  in diameter.  Four to 8 irregularly shaped cells could be distinguished within each spermary.  Stage I  spermaries contained the largest cells, which were translucent even though the nucleus stained  somewhat darker (silvery, grayish) than the cytoplasm.  The nucleolus was often visible as a red or  yellow, sometimes black dot.  The spermary as a whole was often spherical.  The mesoglea generally  surrounded and thereby defined the Stage I spermary whose contour was sometimes a thin black line.    Stage II spermaries were usually between 24 and 45μm in diameter. The cell size of Stage II  spermaries decreased compared to Stage I spermaries, however, their numbers increased to around 40  cells.  Approaching Stage III, the primary spermatocytes became more evenly spherical.  Stage II  spermaries were still translucent with a grayish nucleus and a red nucleolus.  The spermaries were  spherical and still appeared well organized.  A visible, blue mesoglea often surrounded them.         Stage III spermaries were consistently larger than 39μm and attained up to 64μm in diameter  (Appendix 10e).  The cells were smaller than Stage II cells and more densely packed.  Early Stage III  spermaries were still gray in color, the nucleus was no longer well defined, and the red nucleolus began  to disappear.  Mid‐Stage III spermaries were dark red, and late Stage III spermaries were bright, light  red.  If staining was light, only Stage IV spermaries appeared dark magenta lilac in color, and all other  stages appeared colorless.  In this case, distinction between the earlier stages was based on the size of  the spermary, the shape of the spermatocytes and the appearance of the nucleus and nucleolus.  At  Caño Island, Stage III spermaries frequently stained yellow to orange.  Early Stage III spermaries were  often oval‐shaped at all study sites.  A characteristic lumen began to form in the center of late Stage III  spermaries.  The spermaries themselves were only weakly defined by a thin boundary.  A blue mesoglea  was often not visible.    Stage IV spermaries ranged from 64 to 120μm in diameter, and were dark red to dark lilac  (Appendix 10e, f).  Yellow or golden sperm tails were visible in late Stage IV spermaries.  Dark red sperm  heads appeared in the transition of Stage III to IV spermaries with cell division occurring from the central  lumen outwards.  In late Stage IV spermaries, all sperm heads were oriented in one direction with the  tails laid together in a bundle forming an oval‐shaped bouquet.  Stage III spermaries were more loosely  packed than in Stage IV and the secondary spermatocytes were much larger than the spermatozoa.   Spawned spermaries had a dispersed appearance due to the presence of a few residual sperm heads.   Several sperm heads, lacking tails, were usually present outside the spermary.  The mesenterial tissues  formerly surrounding the mature gonads were now in the process of shrinking.  Both mesenteries and  mesoglea were disrupted, appearing shredded.  Occasionally mucus was observed in gonadal tissues  before and after spawning.       

  Appendix 10Psammocora stellata, Psammocoraprofundacella.  Oocyte and spermatocyte stages.  (a), (b)  Stages I—IV oocytes, P. stellata, Caño Island, Costa Rica, 2 November 1986.  (c) Stages III and IV oocytes,  P.  stellata,  Caño  Island,  Costa  Rica,  30  November  1985.    (d)  Stages  II—IV  oocytes,  P.  profundacella,Itabaca  Canal,  Galápagos  Islands,  16  March  1991.    (e)  Stages  III  and  IV  spermaries,  P.  stellata, Stage IV spermary (left) illustrates the bouquet arrangement of mature spermatozoa, Floreana  Island,  Galápagos  Islands,  3  May  1989.  (f)  Stage  IV  spermaries,  P.  profundacella,  Academy  Bay,  Galápagos Islands, 3 June 1990.   Images (c) and (d) are longitudinal sections of polyps, the remainder  are cross‐sections.  g: gastrodermis; gvc: gastrovascular cavity; m: mesentery; t: tails of spermatozoa     

Appendix 11  Psammocora stellata, Psammocorapro fundacella.  Spawned gonads observed in  histological preparations in relation to lunar phase (lunar days 1 and 15 = new and full moon  respectively).    Psammocora stellata          Calendar date  Lunar day  n  Sex Location  November 2, 1986  1  3  ♀ Caño I  March 2, 1990  6  2  ♂ Uva I  June 9, 1987  12  4  ♂ Santa Cruz I  May 12, 1987  13  1  ♀ Uva I  May 3, 1985  13  1  ♂ Uva I  February 16, 1992  13  1  ♀ Uva I  February 13, 1987  14  1  ♂ Pta Estrada  October 15, 1997  15  1  ♀ Uva I  February 26, 1989  20  1  ♀ Uva I  February 25, 1992  22  1  ♂ Saboga I  December 6, 1985  23  1  ♀ Galápagos Is November 26, 1986  23  1  ♀ Santa Cruz I  May 23, 1990  28  1  ♂ Taboga I            Psammocoraprofundacella          February 16, 1991  2  1  ♀ Caño I 

Appendix 12  Psammocora stellata.   Bootstrapped mean ± standard error estimates where [n] are base sample sizes of mature (Stage IV) ova,  numbers of mesenteries polyp‐1, numbers of ova polyp‐1 and cm2 colony surface.  Due to limited sample sizes mean and variance (SEM)  estimates noted were computed from 10,000 bootstrapped sample observations.  All measurements were from histological sections except for  polyps cm‐2, which were counted from undecalcified samples.  These fecundity estimates are ranges based on lower and upper values of SEM.   Ovum volume was calculated from mean ovum diameter (±SEM) using formula for a sphere; mean number of ova polyp‐1 was calculated as  product of mean number of mesenteries polyp‐1, and mean number of ova in cross sections (cs) and longitudinal sections (ls) of mesenteries;  mean number of ova cm‐2 was calculated as product of mean number of polyps cm‐2 and mean number of ova polyp‐1.    (a) Fecundity estimates of Psammocora stellata for Gulf of Panamá were computed by calculating a grand mean value of ls‐mesenteries  from other three localities. Mature ova abundances were computed from this estimated value.    Attribute 

Location 

  (1) Ovum diameter  (µm) 

Caño Island, Costa Rica  98.79+ 1.31  [212] 

Uva Island, Panamá  92.23+ 0.86  [427] 

Gulf of Panamá  93.26 + 2.83  [40] 

Galápagos Islands  94.66+ 1.55  [100] 

(5.61 + 0.22) x 10‐4 [212] 

(4.58 + 0.14) x 10‐4  [427] 

(4.71+ 0.42) x 10‐4 [40] 

(4.82 + 0.26) x 10‐4 [100] 

11.47 + 0.79  [11] 

9.88 + 0.39  [32] 

11.13 + 0.31  [16] 

10.75 + 0.29  [36] 

1.48 + 0.06  [87] 

1.44 + 0.08  [70] 

1.42 + 0.09  [40] 

9.18 + 1.04  [6] 

4.67 + 0.31  [6] 

1.46 + 0.08  [61]  a 7.26 + 0.59   (27) 

13.50 + 0.56  [4] 

12.87 + 0.53  [20] 

11.22+ 0.37  [22] 

14.00 + 0.68  [22] 

  153.93 + 22.10  {89.05 – 282.11} 

  66.50 + 6.07  {48.38 – 90.05} 

117.91 + 12.02  {81.15 – 171.16} 

  115.50 + 15.27  {72.69 – 180.22} 

  2104.83 + 312.15  {1162.70 – 3879.06} 

  855.69 + 86.23  {605.23 – 1226.48} 

1323.32 + 141.43  {898.25 – 1884.72} 

  1617.14 + 228.26  {976.65 – 2640.46} 

 

(2) Ovum volume  (µm3)   

(3) Number of mesenteries  (polyp‐1)   

(4) Number of ova  (cs‐1 of mesentery)   

(5) Number of ova  (ls‐1 of mesentery) 

7.54 + 0.84  [15] 

 

(6) Number of polyps  (cm‐2 of colony)   

(7) No. of stage IV ova (= #3 x #4 x #5)  (polyp‐1)  Range  

(8) No. of stage IV ova (= #6 x #7)  (cm‐2 of colony)  Range

Appendix 13  Psammocoraprofundacella.  Bootstrapped mean ± standard error estimates where [n] are base sample sizes of mature (Stage IV)  ova, numbers of mesenteries polyp‐1, numbers of ova polyp‐1 and cm2 colony surface.  See caption in Appendix 12 for computational details.    (a) Fecundity estimates of P.  profundacella for Uva Island, Panamá were computed by calculating a grand mean value of ls‐mesenteries  from other two localities.  Mature ova abundances were computed from this estimated value.    Attribute  Location    (1) Ovum diameter  (µm) 

Caño Island, Costa Rica  90.51 + 1.93  [45] 

Uva Island, Panamá  94.32 + 1.41  [121] 

(4.12 + 2.70) x 10‐4 [45] 

(4.76 + 0.22) x 10‐4  [121] 

9.41 + 0.84  [5] 

9.85 + 0.21  [32] 

1.62 + 0.12  [34] 

1.42 + 0.10  [38] 

 

Galápagos Islands  94.66 + 1.41  [105]   

(2) Ovum volume   (µm3)   

(4.75 + 0.21) x 10‐4 [105]   

(3) Number of mesenteries  (polyp‐1)   

9.24 + 0.38  [13]   

(4) Number of ova  (cs‐1 of mesentery)   

(5) Number of ova  (ls‐1 of mes) 

a 

14.28 + 1.55  [4] 

11.98 + 1.25 (15) 

15.58 + 1.12  [5] 

19.42 + 0.93  [19] 

 

1.37 + 0.12  [27]   

11.20 + 1.55  [11]   

(6) Number of polyps  (cm‐2 of colony)   

15.50 + 0.88  [11]   

(7) No. of stage IV ova (= #3 x #4 x #5)  (polyp‐1)  Range

  217.55 + 34.29  {108.53 – 361.16} 

  141.71 + 23.89  {71.75 – 255.79} 

167.60 + 21.89  {94.71 – 291.02} 

 

 

(8) No. of stage IV ova (= #6 x #7)  (cm‐2 of colony)  Range

  3390.65+ 590.20  {1678.32 – 5764.50} 

3253.99 + 450.08  {1759.65 – 5782.08} 

  2196.29 + 391.42  {1138.19 – 3939.58} 

Appendix 14  Psammocora stellata, Psammocorapro fundacella.  Boxplots of bootstrapped distribution  of mean ± standard error estimates of fecundity at multiple eastern Pacific locations ordered by latitude.   Each box denotes interquartile range, which contains values between the 25th and 75th percentile;  horizontal line inside box denotes median.  Two ‘whiskers’ on either side of box show adjacent extreme  values.   

  Appendix 15  Psammocora stellata.  Measures of change of population abundances in Costa Rica, Panamá and Galápagos Islands.  Devil’s Crown area in 1976  revised upwards by 16% from Glynn (1994) based on high precision area measurements with CPCe (Kohler and Gill 2006)    Location                  Costa Rica  Measurement  1984  1987  1996  1999  2002  Years Source  Caño Island                  80m2 area  % live coral cover            15  Guzmán& Cortés 2001      5 – 9m depth  0  0  0  0  —        9 – 14m depth  0  0.73 + 0.18  2.27 + 0.49  0.18 + 0.08  —       

 

20m2 area     

total number colonies  5 – 9m depth  9 – 14m depth 

 

 

Cocos Island  Pacheco  Chatham  Presidio  Panamá  Uva Island  288m2 area    Galápagos Iss  Devil’s Crown  Live patches   

 

 

 

 

 

  0  0 

  1  50 

  0  53 

  0  19 

  —  — 

 

 

 

 

% live coral cover      3 – 18m depth  —  0.21 + 0.12  3 – 18m depth  —  0.74 + 0.33  9 – 24m depth  —  0.57 + 0.07  Measurement  1983  1993        x  number colonies m‐2      2 – 3m depth  0  0  Measurement  1976  1983        area analysis (m2)      a 0  3 – 6m depth  2,295 Xarifa Is. Española        34m2 area   % live coral cover        0.5 – 3m depth  —  —  34m2 area  x  number colonies m‐2        0.5 – 3m depth  —  —       

 

  —  —  —  1997      0.04  1993      0      —    — 

 

15       

 

Guzmán& Cortés 2001         

    15  Guzmán& Cortés 2007  —  0.12 + 0.03      —  0.23 + 0.05      —  1.34 + 0.21      2000  2002  Years Source              19  Glynn unpub.  2.05 + 1.00  10.59 + 6.49      2004  2011  Years Source          28  Glynn & Wellington 1983;   —  735    Glynn 1994; Feingold unpub.             7  Brown & Feingold unpub.  9.93 + 0.90  14.57 + 1.08        7  Brown & Feingold unpub.  64.4 + 2.53  153.0 + 10.30    

 

  Appendix 16  View of largest Psammocora stellata patch at west side of Devil’s Crown reef, Floreana  Island, Galápagos Islands, 3 m depth (10 June 2007).  Larger colonies in foreground are about 10 cm in  diameter.