Linkages between coral assemblages and coral proxies of terrestrial exposure along a cross-shelf gradient on the southern Great Barrier Reef

Coral core records, combined with measurements of coral community structure, were used to assess the long-term impact of multiple environmental stressors on reef assemblages along an environmental gradient. Multiple proxies (luminescent lines, Ba/Ca, δ¹⁵N) that reflect different environmental conditions (freshwater discharge, sediment delivery to the nearshore, nutrient availability and transformations) were measured in Porites coral cores collected from nearshore reefs at increasing distance from the intensively agricultural region of Mackay (Queensland, Australia). The corals provide a record (1968–2002) of the frequency and intensity of exposure to terrestrial runoff and fertilizer-derived nitrogen and were used to assess how the present-day coral community composition may have been influenced by flood-related disturbance. Reefs closest to the mainland (5–32 km offshore) were characterized by low hard coral cover (≤10%), with no significant differences among locations. Distinct annual luminescent lines and elevated Ba/Ca values (4.98 ± 0.63 μmol mol⁻¹; mean ± SD) in the most inshore corals (Round Top Island; 5 km offshore) indicated chronic, sub-annual exposure to freshwater and resuspended terrestrial sediment that may have historically prevented reef formation. By contrast, corals from Keswick Island (32 km offshore) indicated episodic, high-magnitude exposure to Pioneer River discharge during extreme flood events (e.g., 1974, 1991), with strongly luminescent lines and substantially enriched coral skeletal δ¹⁵N (12–14‰). The reef assemblages at Keswick and St. Bees islands were categorically different from all other locations, with high fleshy macroalgal cover (80.1 ± 7.2% and 62.7 ± 7.1%, respective mean ± SE) overgrowing dead reef matrix. Coral records from Scawfell Island (51 km offshore) indicated little exposure to Pioneer catchment influence: all locations from Scawfell and further offshore had total hard and soft coral cover comparable to largely undisturbed nearshore to middle shelf reefs of the southern Great Barrier Reef.     

Geomorphology and community structure of Middle Reef, central Great Barrier Reef, Australia: an inner-shelf turbid zone reef subject to episodic mortality events

Middle Reef is an inshore turbid zone reef located 4 km offshore from Townsville, Queensland, Australia. The reef consists of four current-aligned, interconnected reef patches that have reached sea level and formed reef flats. It is regularly exposed to high turbidity (up to 50 mg l⁻¹) generated by wave-driven sediment resuspension or by episodic flood plumes. Middle Reef has a high mean hard coral cover (>39%), relatively low mean macro-algal cover (<15%) and a coral community comprising at least 81 hard coral species. Cluster analysis differentiated six benthic communities which were mapped onto the geomorphological structure of the reef to reveal a spatially patchy community mosaic that reflects hydrodynamic and sediment redistribution processes. Coral cover data collected annually from windward slope transects since 1993 show that coral cover has increased over the last ~15 years despite a history of episodic mortality events. Although episodic mortality may be interpreted as an indication of marginality, over decadal timescales, Middle Reef has recovered rapidly following mortality events and is clearly a resilient coral reef.     

Cold-water event of January 2010 results in catastrophic benthic mortality on patch reefs in the Florida Keys

The Florida Keys are periodically exposed to extreme cold-water events that can have pronounced effects on coral reef community structure. In January 2010, the Florida Keys experienced one of the coldest 12-day periods on record, during which water temperatures decreased below the lethal limit for many tropical reef taxa for several consecutive days. This study provides a quantitative assessment of the scleractinian mortality and acute changes to benthic cover at four patch reefs in the middle and upper Keys that coincided with this cold-water event. Significant decreases in benthic cover of scleractinian corals, gorgonians, sponges, and macroalgae were observed between summer 2009 and February 2010. Gorgonian cover declined from 25.6 ± 4.6% (mean ± SE) to 13.3 ± 2.7%, scleractinian cover from 17.6 ± 1.4% to 10.7 ± 0.9%, macroalgal cover from 8.2 ± 5.2% to 0.7 ± 0.3%, and sponge cover from 3.8 ± 1.4% to 2.3 ± 1.2%. Scleractinian mortality varied across sites depending upon the duration of lethal temperatures and the community composition. Montastraea annularis complex cover was reduced from 4.4 ± 2.4% to 0.6 ± 0.2%, and 93% of all colonies surveyed suffered complete or partial mortality. Complete or partial mortality was also observed in >50% of all Porites astreoides and Montastraea cavernosa colonies and resulted in a significant reduction in cover. When compared with historical accounts of cold-water-induced mortality, our results suggest that the 2010 winter mortality was one of the most severe on record. The level of coral mortality on patch reefs is of particular concern because corals in these habitats had previously demonstrated resistance against stressors (e.g., disease and warm-water bleaching) that had negatively affected corals in other habitats in the Florida Keys during recent decades.     

Benthic structure and cryptic mortality in a Caribbean mesophotic coral reef bank system, the Hind Bank Marine Conservation District, U.S. Virgin Islands

Coral reef banks may form an important component of mesophotic coral ecosystems (MCEs) in the Caribbean, but remain poorly explored relative to shallower reefs and mesophotic habitats on slopes and walls. Consequently, the processes structuring mesophotic coral reef communities are not well understood, particularly the role of disturbance. A large and regionally important mesophotic system, the Hind Bank Marine Conservation District (MCD), St. Thomas, USVI, was systematically surveyed. Data were used to construct a comprehensive benthic habitat map for the MCD, describe the abiotic and biotic components of the benthos among habitats, and investigate patterns of coral health among habitats. Two-thirds of the MCD (23.6 km²) was found to be dense coral reef (Coral Cover = 24.1%) dominated by the Montastraea annularis species complex. Coral reef ecosystems were topographically complex, but could be classified into distinct habitat types, including high coral banks (35.8% of the MCD) and two large novel coral reef habitat types corresponding to an extremely flat basin (18%) and a highly rugose hillock basin (6.5%), containing thousands of coral knolls (2–10 m high). An extreme disease event with undescribed signs of mortality occurred on 47% of coral reefs and reached a high prevalence in affected areas (42.4% ± 6.3 SE, N = 26). The disease was significantly clustered in the basin habitats of the western MCD (global Moran’s I = 0.32, P < 0.01). Observations of the spatial pattern suggested that the driver was specific to the basin habitats and may have been caused by a coherent abiotic event.     

Recent changes in coral assemblages of a South African coral reef, with recommendations for long-term monitoring

Two-mile reef, Sodwana, South Africa is an unusual coral reef, being situated on a submerged fossilized sand dune and being very southerly (27°54). It is a popular Scuba diving venue receiving about 100000 dives year^–1. The line-intercept transect method, as recommended by the global coral reef monitoring network (GCRMN), was used to determine soft coral, hard coral and other benthos percentage cover. Physical coral damage, disease and bleaching were also recorded. Results were compared with those of B. Riegl (1993 – unpublished PhD thesis) 5 to 7 years earlier. The reef appears to be ecologically and highly dynamic. In the interim, there has been an increase in living benthos cover of 22.3% but also an increase in coral bleaching from 0% in 1993 to 1% in 1998. Physical damage, despite the large number of dives on the reef was minimal (1.52%), although it appears as if coral diseases may be increasing. The 20-m transects recommended by GCRMN are too long for this highly rugose reef with its distinct ridges and gullies. It is recommended that benthos cover, coral damage, bleaching and disease should be monitored annually using 40 5-m transects on the reef ridges and 40 5-m transects on the reef slopes.     

Ocean acidification and calcifying reef organisms: a mesocosm investigation

A long-term (10 months) controlled experiment was conducted to test the impact of increased partial pressure of carbon dioxide (pCO₂) on common calcifying coral reef organisms. The experiment was conducted in replicate continuous flow coral reef mesocosms flushed with unfiltered sea water from Kaneohe Bay, Oahu, Hawaii. Mesocosms were located in full sunlight and experienced diurnal and seasonal fluctuations in temperature and sea water chemistry characteristic of the adjacent reef flat. Treatment mesocosms were manipulated to simulate an increase in pCO₂ to levels expected in this century [midday pCO₂ levels exceeding control mesocosms by 365 ± 130 μatm (mean ± sd)]. Acidification had a profound impact on the development and growth of crustose coralline algae (CCA) populations. During the experiment, CCA developed 25% cover in the control mesocosms and only 4% in the acidified mesocosms, representing an 86% relative reduction. Free-living associations of CCA known as rhodoliths living in the control mesocosms grew at a rate of 0.6 g buoyant weight year⁻¹ while those in the acidified experimental treatment decreased in weight at a rate of 0.9 g buoyant weight year⁻¹, representing a 250% difference. CCA play an important role in the growth and stabilization of carbonate reefs, so future changes of this magnitude could greatly impact coral reefs throughout the world. Coral calcification decreased between 15% and 20% under acidified conditions. Linear extension decreased by 14% under acidified conditions in one experiment. Larvae of the coral Pocillopora damicornis were able to recruit under the acidified conditions. In addition, there was no significant difference in production of gametes by the coral Montipora capitata after 6 months of exposure to the treatments.     

Relationships between butterflyfish (Chaetodontidae) feeding rates and coral consumption on the Great Barrier Reef

This study explored differences in the feeding rate among 20 species of coral reef butterflyfishes (Chaetodontidae) from Lizard Island, Great Barrier Reef. Feeding rate, measured as bites per minute (b.p.m.), varied between 2.98 ± 0.65 and 12.29 ± 0.27 (mean ± SE) according to species and was positively related to the proportional consumption of coral (r ² = 0.40, n = 20, P < 0.01), independent of phylogeny (standardised independent contrasts r ² = 0.29, n = 19, P < 0.05). All species fed actively throughout the day, with obligate corallivores having a higher feeding rate at all times than either facultative corallivores or non-corallivores. The feeding rate of the obligate corallivores was also highest during the middle of the day. For eight of the species for which data was available, there was a positive correlation between bite rate and competitive dominance (r = 0.71, P < 0.05). Chaetodon ephippium was the only species for which the feeding rate of pairs was higher than for solitary individuals.     

Identification of a bacterial pathogen associated with Porites white patch syndrome in the Western Indian Ocean

Porites white patch syndrome (PWPS) is a coral disease recently described in the Western Indian Ocean. This study aimed to isolate and identify potential pathogens associated with PWPS utilizing both culture and nonculture screening techniques and inoculation trials. A total of 14 bacterial strains (those dominant in disease lesions, absent or rare in healthy tissues and considered potential pathogens in a previous study) were cultured and used to experimentally inoculate otherwise healthy individuals in an attempt to fulfil Henle–Koch's postulates. However, only one (P180R), identified as closely related (99–100% sequence identity based on 1.4 kb 16S RNA sequence) to Vibrio tubiashii, elicited signs of disease in tank experiments. Following experimental infection (which resulted in a 90% infection rate), the pathogen was also successfully re‐isolated from the diseased tissues and re‐inoculated in healthy corals colonies, therefore fulfilling the final stages of Henle–Koch's postulates. Finally, we report that PWPS appears to be a temperature‐dependent disease, with significantly higher tissue loss (anova : d.f. = 2, F = 39.77, P < 0.01) occurring at 30 °C [1.45 ± 0.85 cm² per day (mean ± SE)] compared to ambient temperatures of 28 and 26 °C (0.73 ± 0.80 cm² per day (mean ± SE) and 0.51 ± 0.50 cm² per day (mean ± SE), respectively).     

Recruitment of scleractinians onto the skeletons of corals killed by black band disease

To determine what happens to scleractinian corals that have been killed by black band disease (BBD), massive corals with BBD were monitored for 11 years on a shallow reef (<10 m depth) in St. John, US Virgin Islands. Small quadrats (0.039 m²) were used to compare the rates of scleractinian recruitment to the skeletons of corals killed by either BBD or physical disturbance (Hurricane Hugo 1989). Coral recruitment was also quantified on the adjacent fringing reef using larger quadrats (0.25 m²) to detect possible biases associated with using small, permanent quadrats to assess recruitment to BBD-killed corals. Of 28 tagged colonies with BBD in 1988, 43% were lost to Hurricane Hugo in 1989, 7% were lost to unknown causes between 1991 and 1992, and 14 were monitored annually for 11 years; of these, 71% were dead and still in their original growth position in 1998. Between 1988 and 1997, corals recruited to the BBD-killed surfaces at a rate of 1.1 ± 0.3 recruits · 0.039 m⁻² · decade⁻¹ (mean ± SE, n = 14), although mortality reduced the density to 0.3 ± 0.2 recruits · 0.039 m⁻² by 1997. The rate of recruitment and the taxonomic composition of the coral recruits to BBD-killed corals were indistinguishable statistically from those to corals killed by Hurricane Hugo. This demonstrates that BBD creates space that is functionally the same as other dead coral surfaces in providing a substratum for coral recruitment. However, because coral recruits are dispersed widely, clumped in distribution and temporally variable in density on the fringing reef as a whole, it is unlikely that they will be found on monitored coral colonies that have been killed by BBD. While this hypothesis is consistent with the higher density of recruits on the fringing reef compared with BBD-killed corals, further studies are required to investigate alternative explanations such as the role of substratum age in favoring recruitment to surfaces other than those killed recently by BBD. Accepted: 26 August 1999     

Long-term persistence of low coral cover and abundance on a disturbed coral reef flat in the northern Red Sea

Levels of coral cover and abundance on a coral reef flat in Eilat (Israeli Red Sea) were estimated in 2001 by surveying nineteen 10-m transects, and compared to the levels reported in the same area between 1966 and 1973. Lower values compared to 1966 levels are evident, and there has been only a modest recovery following a catastrophic low tide that killed a large proportion of the corals in 1970. Percent cover of soft and stony corals (16.1%) was less than half of that reported for 1969 (35%), when a sharp decrease in coral abundance had already been observed. The total number of soft and stony coral colonies observed was 300, compared to 541 in 1966. In contrast to 1966, when half of the transects surveyed contained more than 30 coral colonies, no transects with this number of corals were observed. The cover of seven of the most common stony coral species was 841 cm, which is twice the coral cover of that in 1973, but only 22% of the 1969 level. Millepora dichotoma, an abundant species before 1970, has almost disappeared, and the soft coral Litophyton, abundant in 1972, was not observed. Anthropogenic nutrient enrichment is apparently among the causes for the lack of coral recovery in the studied reef flat. Reefs located further away from sources of pollution have recovered quickly after natural and anthropogenic disturbances and have retained their coral abundance and diversity.     

Foraminiferal assemblage and reef check census in coral reef health monitoring of East Brazilian margin

Human activity is changing environmental conditions on a global scale. Among the ecosystems that are affected by human activities, coral reefs are among the most prominent. In Brazil, the coral reefs of the Corumbau Marine Extractive Reserve (CMER) and Abrolhos National Marine Park (ANMP) in Bahia state have some of the highest coral cover in the South Atlantic Ocean. Hard coral cover, algal cover, and foraminiferal population distribution patterns were used to assess the coral reef benthic environments, and define a background that can be used in worldwide comparisons in future studies. To compare these two monitoring approaches in different coral reef environments, relative frequency data for occurrence of hard coral and algal cover, using point-intercept transects as proposed by the Reef Check protocol, and foraminiferal samples were collected from Corumbau (nearshore) and Abrolhos (offshore) in April 2005. The foraminiferal assemblage was evaluated using the FORAM index (FI — Foraminifera in Reef Assessment and Monitoring), which provides a numeric diagnosis of suitability of benthic habitat to support calcifying organisms that host algal symbionts, originally developed for Caribbean reef areas. Coral cover in the surveyed areas, both in Corumbau and in Abrolhos, ranged from 13% to 37%, while high foraminiferal diversities (H') were found in all stations. Dominance of symbiont-bearing taxa of Amphistegina lessonii and Archaias angulatus only occurred at two shallow stations, Mato Verde and Siriba, both in Abrolhos, where FI > 4.00. Stations located in Corumbau and Abrolhos had FI values < 4.00. Q-mode cluster analysis showed that foraminifers have specific preferences for physical conditions, especially hydrodynamics and light availability, which influence the FI index. Although coral cover in these areas can be considered good by regional standards, foraminifer analysis showed that the benthic system was unfavorable for symbiont-bearing foraminiferal species at most stations. This discrepancy reveals that the FI must be used with caution in areas other than the northwestern Atlantic and Caribbean where it was developed, and that some coral species can thrive in muddier conditions than can most symbiont-bearing foraminifers.     

Tunicamycin-Induced Cell Death in the Trigeminal Ganglion is Suppressed by Nerve Growth Factor in the Mouse Embryo

The effect of nerve growth factor (NGF) on tunicamycin (Tm)-treated neurons in the trigeminal ganglion was investigated by use of caspase-3 immunohistochemistry. In intact embryos at embryonic day 16.5, only a few caspase-3-immunoreactivity were detected in the ganglion neurons. Mean ± SE of the density of the immunoreactivity was 0.22 ± 0.03%. In contrast, the number of the immunoreactive neurons was increased at 24 h after injection of 0.5 μg Tm in 1 μl of 0.05 N NaOH solution into mouse embryos at embryonic day 15.5. The density of immunoreactivity was also increased (mean ± SE = 1.44 ± 0.11%) compared to intact and 0.05 N NaOH-treated embryos (mean ± SE = 0.35 ± 0.03%). The Tm treatment caused increase of the number of trigeminal neurons representing apoptotic profiles (intact, mean ± SE = 79.3 ± 8.5; 0.05 N NaOH, mean ± SE = 132 ± 11.5; 0.5 μg Tm, mean ± SE = 370.2 ± 64.8). In addition, NGF significantly prevented the increase of density of the immunoreactivity (mean ± SE = 0.54 ± 0.16%) and the number of apoptotic cells (mean ± SE = 146.2 ± 11.3). Saline application (without NGF) had no effect on Tm-induced increase of the immunoreactivity (mean ± SE = 1.78 ± 0.23%) or the apoptotic profiles (mean ± SE = 431.9 ± 80.5). These results indicate that Tm-induced cell death in the trigeminal ganglion is suppressed by NGF in the mouse embryo.     

Competitive interactions between corals and Trididemnum solidum on Mexican Caribbean reefs

The ascidian Trididemnum solidum competes for space on Caribbean reefs and is capable of overgrowing live scleractinian corals. From 2006 to 2009, we monitored over 30,000 coral colonies and quantified competitive interactions with this ascidian at four reef sites along the Mexican Caribbean. The total number of competitive interactions increased in time, but the mean percentage of coral colonies involved in interactions remained lower than 1% in all reefs. Bottom cover by T. solidum was also low (mean < 0.5%) in all reef sites in all sampling years. We conclude that during the temporal scope of our study, the overall potential effect of T. solidum on the dynamics of Mexican Caribbean coral populations was minimal.     

Larval behaviour and settlement cues of a brooding coral reef sponge

Patterns of larval release, dispersal and settlement in sponges are poorly understood despite their significance in explaining adult ecology. Time of release, swimming speeds, phototaxis and vertical migration were quantified for larvae of the dictyoceratid sponge Coscinoderma matthewsi. The influence of cues associated with biofilms and coral rubble on larval settlement and metamorphosis was also measured. C. matthewsi is a brooding sponge and releases tufted parenchymellae larvae during the day. Upon release, larvae (>90%) have no phototactic response, maintaining their position at the water surface for 80 min ± 0 (mean ± SE) regardless of a light cue (natural daylight) before exhibiting negative phototaxis. At 28 h post-release, the majority of larvae (94.7% ± 6.1) exposed to light from the surface migrated to the bottom and assumed a demersal phase. Without light, larvae occupied the surface for up to 28 h post-release (89.3% ± 1.8) before migrating to the bottom. Larvae did not settle gregariously and began to settle and metamorphose after 28 h post-release without a cue. Settlement and metamorphosis were faster in the presence of a biofilm (settlement = 15.0% ± 8.7 and metamorphosis = 12.5% ± 9.5 at 28 h post-release), while the addition of coral rubble accelerated metamorphosis further (settlement = 10.0% ± 4.1 and metamorphosis = 27.5% ± 10.3 at 28 h post-release) compared to controls (sterile surfaces) (settlement = 0% and metamorphosis = 0% at 28 h post-release). However, both biofilms and coral rubble decrease total metamorphosis (control = 92.5% ± 4.8, biofilms = 67.5% ± 7.5 and coral rubble = 55.0% ± 13.2) due to mortality after 76 h post-release.     

Fine-scale time series surveys reveal new insights into spatio-temporal trends in coral cover (2002–2018), of a coral reef on the Southern Great Barrier Reef

Reef monitoring programmes often focus on limited sites, predominantly on reef slope areas, which do not capture compositional variability across zones. This study assessed spatial and temporal changes in hard coral cover at four hierarchical spatial scales. ~ 55,000, geo-referenced photoquadrats were collected annually from 2002 to 2018 and analysed using artificial intelligence for 31 sites across reef flat and reef slope zones on Heron Reef, Southern Great Barrier Reef, Australia. Trends in hard coral cover were examined at three spatial scales: (1) “reef scale”, all data; (2) “geomorphic zone scale”—north/south reef slope, inner/outer reef flat; and (3) “site scale”—31 sites. Coral cover trajectories were also examined at: (4) “sub-site scale”—sub-division of sites into 567 sub-sites, to estimate variability in coral cover trajectories via spatial statistical modelling. At reef scale coral cover increased over time to 25.6 ± 0.4 SE % in 2018 but did not recover following disturbances caused by disease (2004–2008), cyclonic conditions (2009) or severe storms (2015) to the observed pre-disturbance level (44.0 ± 0.7 SE %) seen in 2004. At geomorphic zone scale, the reef slope had significantly higher coral cover than the reef flat. Trends of decline and increase were visible in the slope zones, and the southern slope recovered to pre-decline levels. Variable coral cover trends were visible at site scale. Furthermore, sub-site spatial modelling captured eight years of coral recovery that occurred at different times and magnitudes across the four geomorphic zones, effectively estimating variability in the trajectory of the reef’s coral community. Derived spatial predictions for the entire reef show patchy coral recovery, particularly on the southern slope, and that recovery hotspots are distributed across the reef. These findings suggest that to fully understand and interpret coral decline or recovery on a reef, more accurate assessment can be achieved by examining sites distributed within different geomorphic zones to capture variation in exposure, depth and consolidation.

     

Decadal‐scale rates of reef erosion following El Niño‐related mass coral mortality

As the frequency and intensity of coral mortality events increase under climate change, understanding how declines in coral cover may affect the bioerosion of reef frameworks is of increasing importance. Here, we explore decadal‐scale rates of bioerosion of the framework building coral Orbicella annularis by grazing parrotfish following the 1997/1998 El Niño‐related mass mortality event at Long Cay, Belize. Using high‐precision U‐Th dating and CT scan analysis, we quantified in situ rates of external bioerosion over a 13‐year period (1998–2011). Based upon the error‐weighted average U‐Th age of dead O. annularis skeletons, we estimate the average external bioerosion between 1998 and 2011 as 0.92 ± 0.55 cm depth. Empirical observations of herbivore foraging, and a nonlinear numerical response of parrotfish to an increase in food availability, were used to create a model of external bioerosion at Long Cay. Model estimates of external bioerosion were in close agreement with U‐Th estimates (0.85 ± 0.09 cm). The model was then used to quantify how rates of external bioerosion changed across a gradient of coral mortality (i.e., from few corals experiencing mortality following coral bleaching to complete mortality). Our results indicate that external bioerosion is remarkably robust to declines in coral cover, with no significant relationship predicted between the rate of external bioerosion and the proportion of O. annularis that died in the 1998 bleaching event. The outcome was robust because the reduction in grazing intensity that follows coral mortality was compensated for by a positive numerical response of parrotfish to an increase in food availability. Our model estimates further indicate that for an O. annularis‐dominated reef to maintain a positive state of reef accretion, a necessity for sustained ecosystem function, live cover of O. annularis must not drop below a ~5–10% threshold of cover.     

Bottlenecks to coral recovery in the Seychelles

Processes that affect recovery of coral assemblages require investigation because coral reefs are experiencing a diverse array of more frequent disturbances. Potential bottlenecks to coral recovery include limited larval supply, low rates of settlement, and high mortality of new recruits or juvenile corals. We investigated spatial variation in local abundance of scleractinian corals in the Seychelles at three distinct life history stages (recruits, juveniles, and adults) on reefs with differing benthic conditions. Following widespread coral loss due to the 1998 bleaching event, some reefs are recovering (i.e., relatively high scleractinian coral cover: ‘coral-dominated’), some reefs have low cover of living macrobenthos and unconsolidated rubble substrates (‘rubble-dominated’), and some reefs have high cover of macroalgae (‘macroalgal-dominated’). Rates of coral recruitment to artificial settlement tiles were similar across all reef conditions, suggesting that larval supply does not explain differential coral recovery across the three reef types. However, acroporid recruits were absent on macroalgal-dominated reefs (0.0 ± 0.0 recruits tile^?1) in comparison to coral-dominated reefs (5.2 ± 1.6 recruits tile^?1). Juvenile coral colony density was significantly lower on macroalgal-dominated reefs (2.4 ± 1.1 colonies m^?2), compared to coral-dominated reefs (16.8 ± 2.4 m^?2) and rubble-dominated reefs (33.1 ± 7.3 m^?2), suggesting that macroalgal-dominated reefs have either a bottleneck to successful settlement on the natural substrates or a high post-settlement mortality bottleneck. Rubble-dominated reefs had very low cover of adult corals (10.0 ± 1.7 %) compared to coral-dominated reefs (33.4 ± 3.6 %) despite no statistical difference in their juvenile coral densities. A bottleneck caused by low juvenile colony survivorship on unconsolidated rubble-dominated reefs is possible, or alternatively, recruitment to rubble-dominated reefs has only recently begun. This study identified bottlenecks to recovery of coral assemblages that varied depending on post-disturbance habitat condition.     

Do fluctuating temperature environments elevate coral thermal tolerance?

In reef corals, much research has focused on the capacity of corals to acclimatize and/or adapt to different thermal environments, but the majority of work has focused on distinctions in mean temperature. Across small spatial scales, distinctions in daily temperature variation are common, but the role of such environmental variation in setting coral thermal tolerances has received little attention. Here, we take advantage of back-reef pools in American Samoa that differ in thermal variation to investigate the effects of thermally fluctuating environments on coral thermal tolerance. We experimentally heat-stressed Acropora hyacinthus from a thermally moderate lagoon pool (temp range 26.5–33.3°C) and from a more thermally variable pool that naturally experiences 2–3 h high temperature events during summer low tides (temp range 25.0–35°C). We compared mortality and photosystem II photochemical efficiency of colony fragments exposed to ambient temperatures (median: 28.0°C) or elevated temperatures (median: 31.5°C). In the heated treatment, moderate pool corals showed nearly 50% mortality whether they hosted heat-sensitive (49.2 ± 6.5% SE; C2) or heat-resistant (47.0 ± 11.2% SE; D) symbionts. However, variable pool corals, all of which hosted heat-resistant symbionts, survived well, showing low mortalities (16.6 ± 8.8% SE) statistically indistinguishable from controls held at ambient temperatures (5.1–8.3 ± 3.3–8.3% SE). Similarly, moderate pool corals hosting heat-sensitive algae showed rapid rates of decline in algal photosystem II photochemical efficiency in the elevated temperature treatment (slope = −0.04 day⁻¹ ± 0.007 SE); moderate pool corals hosting heat-resistant algae showed intermediate levels of decline (slope = −0.039 day⁻¹ ± 0.007 SE); and variable pool corals hosting heat-resistant algae showed the least decline (slope = −0.028 day⁻¹ ± 0.004 SE). High gene flow among pools suggests that these differences probably reflect coral acclimatization not local genetic adaptation. Our results suggest that previous exposure to an environmentally variable microhabitat adds substantially to coral–algal thermal tolerance, beyond that provided by heat-resistant symbionts alone.     

Partial Mortality in Porites Corals: Variation among Philippine Reefs

Partial mortality or tissue necrosis was quantified in the massive scleractinian coral Porites at three sites in The Philippines (Bolinao, NW Luzon; Puerto Galera, Mindoro; and El Nido, N Palawan). Overall, 15 ± 1 (mean ± 1 standard error, 642 replicates) percent of colony area was dead, mean colony area was 1135 plusmn; 127 cm², and lesion density was 1.7 ± 0.1 dm^—2. Total live coral cover varied between 20 and 63% in belt transects, and Porites and Acropora cover were inversely correlated. ANOVA models incorporating effects of site, colony size, sedimentation rates, wave exposure and depth were highly significant but explained only a small proportion of the variation observed in lesion density and percent dead area (respectively 8 and 2%). Lesion density was found to vary significantly with site (contributed 29% to this explained variance), decrease with increasing colony area (33%), and increase with increasing sedimentation (23%) and wave exposure (14%). Colony size was significantly explained by the factor site (contributing 61% to the total 29% explained variance) and depth (34%), with the smallest colonies being observed in Bolinao and the largest in El Nido. Densities of lesions were highest in Bolinao, intermediate in Puerto Galera, and lowest in El Nido. This pattern is parallel to intensity of human reef exploitation and opposite to that in colony size, live coral cover and Acropora cover. Since only a small part of the observed variance in partial mortality estimators was explained by the ANOVAs, other factors not quantified here must have been more important (e.g. disease incidence, predation, human exploitation).