Coral reefs are essential to millions of island inhabitants. Yet, coral reefs are threatened by thermal anomalies associated with climate change and by local disturbances that include land-use change, pollution, and the coral-eating sea star Acanthaster solaris. In combination, these disturbances cause coral mortality that reduce the capacity of reefs to produce enough carbonate to keep up with sea-level rise. This study compared the reef-building capacity of shallow-water inner, patch, and outer reefs in the two islands of Pohnpei and Kosrae, Federated States of Micronesia. We identified which reefs were likely to keep up with sea-level rise under different climate-change scenarios, and estimated whether there were differences across habitats in the threshold of percentage coral cover at which net carbonate production becomes negative. We also quantified the influence of A. solaris on carbonate production. Whereas the northwestern outer reefs of Pohnpei and Kosrae had the highest net rates of carbonate production (18.5 and 16.4 kg CaCO3 m-2 yr-1, respectively), the southeastern outer reefs had the lowest rates of carbonate production (1.2–1.3 and 0.7 kg CaCO3 m-2 yr-1, respectively). The patch reefs of Pohnpei had on average higher net carbonate production rates (9.5 kg CaCO3 m-2 yr-1) than the inner reefs of both Pohnpei and Kosrae (7.0 and 7.8 kg CaCO3 m-2 yr-1, respectively). A. solaris were common on Kosrae and caused an average reduction in carbonate production of 0.6 kg CaCO3 m-2 yr-1 on Kosraean reefs. Northern outer reefs are the most likely habitats to keep up with sea-level rise in both Pohnpei and Kosrae. Overall, the inner reefs of Pohnpei and Kosrae need ~ 5.5% more coral cover to generate the same amount of carbonate as outer reefs. Therefore, inner reefs need special protection from land-use change and local pollution to keep pace with sea-level rise under all climate-change scenarios.
Citation: van Woesik R, Cacciapaglia CW (2019) Carbonate production of Micronesian reefs suppressed by thermal anomalies and Acanthaster as sea-level rises. PLoS ONE 14(11): e0224887. https://doi.org/10.1371/journal.pone.0224887
Editor: James R. Guest, Newcastle University, UNITED KINGDOM
Received: May 31, 2019; Accepted: October 23, 2019; Published: November 15, 2019
Copyright: © 2019 van Woesik, Cacciapaglia. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All data and R-code files are available in the Supporting Information files.
Funding: RvW received funding for the research from the National Science Foundation, award NSF OCE-1657633. The funders did not play a role in the study design, data collection, data analysis, decision to publish, and preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Coral reefs are an integral component of global marine ecosystems and are essential to millions of people that benefit from the goods and services that coral reefs provide. For example, coral reefs reduce storm-wave energy by up to 97% , reducing the threat of coastal inundation during severe storms . However, contemporary thermal-stress events, associated with global-climate change, cause coral bleaching and mortality, which can lead to shifts in species dominance [3–6]. These changes have reduced the capacity of coral reefs to accrete carbonate in some localities and keep up with sea-level rise . Here we examine whether the reefs of Pohnpei and Kosrae, Federated States of Micronesia (FSM; Figure A in S1 File), are producing enough carbonate to keep up with sea-level rise, while experiencing thermal stress and local disturbances.
Rates of carbonate production have been studied using geological coring [8–10], hydrochemistry , modeling , and in situ estimates [13–15]. All approaches show considerable variation across ocean basins, with erosional forces only exceeding rates of carbonate production when gross calcification rates are low. In a global synthesis, Vecsei (2004)  showed that carbonate production decreased with depth and was lower on reef flats than in other habitats. These results agree with van Woesik and Cacciapaglia (2018)  who showed major differences in carbonate production among reef habitats in the Republic of Palau and the island of Yap, western FSM, with outer reefs averaging greater carbonate production (i.e., 10 kg CaCO3 m-2 yr-1) than inner reefs (i.e., averaging 7 kg CaCO3 m-2 yr-1). Yet, in less favorable environments, Perry et al. (2013)  estimated that Caribbean reefs have modern net carbonate production rates averaging only 1.5 kg CaCO3 m-2 yr-1, with some reefs showing net negative carbonate budgets. On the shallow fore-reef slopes in the Maldives in the Indian Ocean, Perry and Morgan (2017)  showed that after a thermal-stress event net accretion rates were negative, at -3 kg CaCO3 m-2 yr-1. Although, through the same thermal-stress event, Ryan et al. (2019)  showed evidence that the upper reef crest and reef flats of the same Maldivian reef maintained positive accretion rates. In a study in the Seychelles, Januchowski-Hartley et al. (2017)  showed that carbonate production was dependent on thermal stress, depth, macroalgal presence, wave energy, and the abundance of excavating parrotfish.
Carbonate production rates can be influenced also by other chronic local disturbances. For example, the coral predator Acanthaster solaris has long been known to reduce coral populations when the sea stars are in high densities . While high densities of Acanthaster larvae have been associated with elevated nutrient concentrations through river discharge [21, 22], other aspects of their biology and ecology remain unresolved . Here we examine the influence of A. solaris populations on the capacity of Pohnpeian and Kosraean shallow-water reefs to produce carbonate, and determine the density of A. solaris, relative to the available percentage of live coral cover, beyond which carbonate production is reduced to zero.
Quantifying carbonate production is critical when predicting how coral reef systems will respond to sea-level rise, especially as the rate of sea-level rise is predicted to accelerate rapidly into the future [24–26]. These estimates in carbonate production should also influence conservation targets, especially if inner reefs require more coral cover to produce the same amount of carbonate as outer reefs . Here we examine the coral reefs on two islands, Pohnpei and Kosrae, FSM, and quantify the in situ rates of carbonate production to identify which reefs are likely to keep up with sea-level rise [27, 28] under different climate-change scenarios. We also estimate whether there are differences in the threshold of percentage coral cover, at which net carbonate production becomes negative, across habitats.
Study design and field methods
Twenty-four study sites were randomly selected in each of Pohnpei (6.2°N, 158.2°E) and Kosrae (5.3°N, 162.9°E) FSM using a randomly stratified sampling approach with the package sp  in R . In Pohnpei, reefs were stratified as inner reefs, patch reefs, and outer reefs. In Kosrae, we only stratified the reefs as either inner reefs or outer reefs (because of the lack of patch reefs). Sample size of each strata was determined by calculating the geographic area of each reef type, using the area function from the R package raster , and allocating the number of sites in accordance with the area estimates. Reef surveys focused on the 2–5 meters depth contour to estimate shallow-water carbonate production.
Six, 10 m transects, using a modified line-intercept technique that followed the reef substrate, were used to measure the benthic composition for every centimeter, at each site of the 48 sites [32, 17]. A few meters gap was allocated between the ends of the transects to ensure no overlap of substrate between transects. Corals were recorded to species level, except massive Porites and encrusting Montipora, which were recorded in the field as growth forms. All other organisms along each transect were identified to the highest possible taxonomic resolution. Rugosity was recorded using the planar length of a second transect that spanned across the reef horizontally. Echinoids were recorded within 30 cm on either side of the 10 m tape. The urchins were recorded as Echinometra, Diadema, and ‘Other’, and the diameter of each echinoid test was measured to the nearest 0.5 cm. The abundance of Acanthaster solaris (crown-of-thorns sea star) were recorded within 5 m along each of the six 10 m transects. Herbivorous parrotfishes were videoed and identified to species and their estimated length was recorded to the nearest cm along six transects, each of which was 30 m long by 4 m wide. Care was taken to record the fish-transect videos ahead of the other transects to avoid any disturbance to the fishes.
Net carbonate production (kg CaCO3 m-2 yr-1) was estimated using the following equation: (1) where calcification is the gross carbonate production by reef building organisms at site i ; sedimentation is the contribution of sediment to the reef, where it increases carbonate production rates if sedimentation is low (< 0.05 kg m-2 d-1) [33,34] and then sgn (x) is positive, whereas if terrestrial sedimentation is high sgn (x) is negative because the sediment smothers corals; erosion is the rate of erosion, estimated following van Woesik (2013) ; and Acanthaster is amount of carbonate potentially lost by Acanthaster solaris (i.e., the crown-of-thorns sea star) eating corals. High densities of A. solaris reduce live coral cover , which in turn reduces a reef’s capacity to produce carbonate. Calcification of organisms was calculated as follows: (2) where r is the rugosity of site i averaged across six transects; m is the adjustment coefficient for the morphology of species j at site i (following van Woesik and Cacciapaglia 2018) ; x is the planar percent cover averaged across site i for species j, d is the density (g cm-3) of species j (following ) in site i; g is the vertical growth rate of coral species j (cm yr-1) (after van Woesik and Cacciapaglia 2018 ); 10 is an adjustment constant to convert units back to kg CaCO3 m-2 yr-1; and ca is the contribution of coralline algae to carbonate production, calculated following Perry et al. (2012)  as: (3) where pca is the planar coralline algae cover averaged across six transects at site i, 0.018 is the gross carbonate production (g cm-2 yr-1) estimated using averages from Perry et al. (2012) ; and 10 is an adjustment constant used to convert units from g cm-2 yr-1 to kg m-2 yr-1.
The accretion of reefs can be supplemented by calcareous sedimentation [9,10], or compromised by excessive amounts of terrestrial sedimentation (when > 0.05 kg m-2 d-1) , which causes coral smothering and reduces the rate of carbonate production . The sedimentation rate that was used in Pohnpei and Kosrae was 0.4 kg CaCO3 m-2 yr-1 following estimates from Montaggioni (2005)  and Hubbard (1997) . We witnessed some terrestrial runoff and a high deposition of fine sediment in the southern bay of Kosrae (Utwe Bay), and we therefore introduced a negative sedimentation component to Eq 1, using -0.4 kg CaCO3 m-2 yr-1 , at sites that were downstream of river runoff at that location.
Reef erosion was comprised of three biological components, echinoids or sea urchins, herbivorous fishes, and macroboring organisms. Gross erosional rates were calculated as: (4) where parrotfish is the rate of erosion by herbivorous fish species j at site i; urchin is the rate of biological erosion by sea urchins species j at site i; and macroboring is the erosional forces of macroboring organisms in site i. Erosion by parrotfish was estimated after  using the equation: (5) where vol is the estimated volume of the bites of individual parrotfish n for species j at site i; sp is the scar proportion, or the proportion of bites that leave scars on corals for individual n, of species j at site i; br is the bite rate (bites day-1) of individual n, of species j, at site i; the average density D of corals was calculated at site i based on coral composition; the constant 365 was to convert days into years; and 0.001 was a constant to convert grams into kilograms. Bite volume vol was further defined using the following equation: (6) where length is the length (cm) of each parrotfish n, of species j at site i; the constants were gained using a linear regression of data collected by Ong and Holland (2010) ; the constant one thousand was used to convert cubic millimeters to cubic centimeters. Scar proportion, sp, from Eq 5 was further defined as follows: (7) where length is the length of fish n of species j at site i. The equation was based on a regression using data from Bonaldo and Bellwood (2008)  and Ong and Holland (2010) . Bite rate, br, from Eq 5 was defined as: (8) where brc is a bite rate constant, reeftime is the amount of time fishes spend grazing on reefs, estimated to be 9 hours per day. These constants were estimated by Peter Mumby (personal communication). Length is length of fish n of species j at site i.
Erosion by sea urchins (kg CaCO3 m-2) was estimated after  using the following equation: (9) where Diadema is the erosion caused by a Diadema individual n at site i; Echinometra is the erosion caused by an Echinometra individual n at site i, and Other urchins is the erosion caused by sea urchins that were not Echinometra or Diadema. Diadema was defined as: (10) following an equation by Januchowski-Hartley et al. (2017) , where diameter is the test size (cm) of the individual n at site i. Echinometra, from Eq 9, also follows an equation from Januchowski-Hartley et al. (2017) : (11) where diameter is the test size of individual n within the genus Echinometra at site i. Other urchins in Eq 9 follows an equation from Januchowski-Hartley et al. (2017) , as follows: (12) where diameter is the test size of individual n outside the genus of Echinometra or Diadema at site i. Macroboring organisms were included into Eq 4 to incorporate the erosional forces of boring sponges following the equation: (13) where plamc is the planar cover of the macroboring organisms averaged over at site i; and mec is the constant used to define macroboring erosion, which was set as a conservative 10 kg CaCO3 m-2 yr-1 following Glynn (1997) .
While Glynn (1973)  estimated the densities of Acanthaster that would overwhelm the ability for corals to persist, there have been no studies aimed at quantifying the influence of Acanthaster on carbonate production rates. We used field estimates to evaluate the effect of Acanthaster solaris on carbonate production as follows: (14) where Acanthasteri is the reduction in gross production caused by observed A. solaris at site i; RIi is the rate of coral ingestion at site i (see Eq 16); 50 is a constant to convert observational transect size to m2; and tc.transecti is the per transect consumption rate: (15) where A.spi is the number of A. solaris observed in site i divided by the number of transects; con is the average consumption rate (0.01 m2 d-1) estimated from Keesing and Lucas (1992) ; Densityi is the average density of corals in site i (g cm3); 10 is the constant used to convert the unit to kg m2, and 365 converts days to years. (16) where RIi is the rate of coral ingestion in site i; Ri is the resource density or live coral cover (%) at site i. The handling rate, or how long it takes for a single A. solaris to eat a coral colony, h, was estimated using the average size of coral colonies and the average rate of consumption, which was conservatively estimated to be around 3.5 days. The 4.53 constant is used to rescale RIi, which is resource dependent, to match average coral density with estimated consumption rate; ai is the attack rate, estimated using the speed at which A. solaris can move and the density of corals in the transect following Eq 17: (17) where 12 is a constant for active predatory hours; Ri is the resource density or live coral cover in site i as a percentage, subtracted from 100; and 1E10 was added to convert the value to a non-zero area where corals are present. These values are divided by 10 for the transect length, to determine average distance between corals, and it was assumed the A. solaris had to search the area of a circle with this average distance between corals equaling the diameter of that circle. The area was then divided by the speed at which A. solaris can move, speed, (504 m d-1; Muller et al. 2011 ). It was assumed that A. solaris could only reduce or negate carbonate production in this model, so the effects were subtracted from gross production to a maximum erosional force netting zero gross production.
To convert net carbonate production, from Eq 1, to vertical reef growth (in mm) we used: (18) where Cp is carbonate production (from Eq 1) and alpha is a coefficient estimated as -0.01949 (after van Woesik and Cacciapaglia 2018 ).
We used an additive mixed effects model in a Bayesian framework  to estimate the value of coral cover, for the different habitats, at which net carbonate production became negative, using the following: (19) where G is the net carbonate production at site i; f(coral cover) uses an O’Sullivan spline smoothing function ; Habitat is the fixed effect of interest; a is a random intercept for site; and error is the error term for the residuals. We assumed that no prior information was known and therefore used multivariate normal diffuse and normal diffuse priors . All models were run in R and coded in JAGS  (all the R code is available in S1 Data and at https://github.com/rvanwoesik).
We would also like to thank Eugene Joseph the Director of the Conservation Society of Pohnpei and Andy George the Director of the Kosrae Conservation Society for granting us permission to conduct research on Pohnpei and Kosrae respectively.
Gross carbonate production on Pohnpei averaged 8.2 kg CaCO3 yr-1, and was on average higher on patch reefs (9.1 kg CaCO3 m-2 yr-1) than on outer reefs (7.7 kg CaCO3 m-2 yr-1) and on inner reefs (6.8 kg CaCO3 m-2 yr-1) (Table 1). Net carbonate production rates closely followed rates of gross production (Table 1), although within-habitat differences were considerable (Figs 1 and 2). For example, the outer northwestern reefs of Pohnpei supported the highest rates of net carbonate production (18.5 kg CaCO3 m-2 yr-1), and the lowest rates were recorded on the southeastern outer reefs (1.2–1.3 kg CaCO3 m-2 yr-1), (Table 1 and Fig 2).
Net shallow-water coral-reef carbonate production stratified by (a) island, (b) habitat type (at both Kosrae and Pohnpei inner reefs and outer reefs, and Pohnpei patch reefs), and (c) site (at 24 sites in Pohnpei and 24 sites in Kosrae, Federated States of Micronesia) 2018. The thick horizontal lines are the medians, the box surrounding the medians are the first and third quartiles, the whiskers identify the range of the data, and the circles identify outliers. These data do not include the erosional effects of Acanthaster solaris.
Base elevation map was plotted in R using raw 10 m Digital Elevation Model from https://pae-paha.pacioos.hawaii.edu/thredds/ncss/usgs_dem_10m_pohnpei/dataset.html.
All values are in kg CaCO3 m-2 yr-1. Gross production is the total rate of carbonate production across habitats, excluding erosion rates. Erosion includes parrotfish and urchin erosional forces combined (which does not include Acanthaster solaris erosion). Net production is gross carbonate production minus the erosional estimates and sedimentation inputs. Acanthaster erosion is the erosion caused by A. solaris. Combined reef strata were only available for inner and outer reefs, due to the lack of patch reefs in Kosrae.
Gross carbonate production on Kosrae averaged 7.4 kg CaCO3 m-2 yr-1, although on average the outer and inner reefs did not vary greatly (7.6 and 7.0 kg CaCO3 m-2 yr-1, respectively). Similar to Pohnpei, carbonate production on the outer reefs on Kosrae was variable and was highest on the northern outer reefs (16.4 kg CaCO3 m-2 yr-1), and lowest on southeastern outer reefs (0.7 kg CaCO3 m-2 yr-1) (Table 1 and Fig 3). Net production differed from gross production at sites where sedimentation and erosion were much higher than background rates. This occurred in Utwe Bay in southern Kosrae, where terrestrially derived sediment was much higher than elsewhere (personal observations). Terrestrially derived sediment smothers coral colonies and thereby reduces carbonate production.
Base elevation map was plotted in R using raw 10 m Digital Elevation Model from https://pae-paha.pacioos.hawaii.edu/thredds/ncss/usgs_dem_10m_kosrae/dataset.html.
The reefs of Pohnpei and Kosrae supported similar coral assemblages, although there were some differences in species dominance. The reefs of Pohnpei, particularly the patch and inner reefs, were dominated by Porites rus, Porites cylindrica, and Porites lobata. The outer reefs were dominated by encrusting Montipora and Acropora hyacinthus (Fig 4). These five species contributed 82% of the gross carbonate production on Pohnpei. The reefs of Kosrae were dominated by encrusting Montipora, Porites rus, Goniastrea retiformis, Porites lobata, and Porites lichen (Fig 4). These five species contributed 78% of the total gross carbonate production on Kosrae. Importantly, the inner reefs of both Pohnpei and Kosrae, and the patch reefs of Pohnpei had a higher live-coral-cover threshold than the outer reefs of both islands, although there was considerable uncertainty in the thresholds for the inner reefs (i.e., high 95% credible intervals) (Fig 5).
Cumulative shallow-water coral-reef carbonate production by coral species and other benthic taxa for (a) 24 sites in Pohnpei and (b) 24 sites in Kosrae, Federated States of Micronesia, 2018.
Percent threshold live coral cover (LCC) needed to maintain net positive accretion stratified by shallow-water coral-reef habitat (at both Kosrae and Pohnpei inner reefs and outer reefs, and Pohnpei patch reefs) for (a) 24 sites in Pohnpei and (b) 24 sites in Kosrae, Federated States of Micronesia, 2018. The dots represent the posterior means and the vertical lines represent the 95% credible intervals.
Acanthaster solaris were observed on reefs of both islands although in 2018 populations indicative of an outbreak (>30 hectare-1) were only observed on some of the shallow reefs of Kosrae. Since outer reefs tended to have the highest densities of A. solaris, carbonate production on these outer reefs were most affected (Table 1 and Figs 6 and 7). We re-ran the carbonate production model for both islands to incorporate A. solaris and found that carbonate production was reduced on Kosrae by an average 0.6 kg CaCO3 m-2 yr-1 and on Pohnpei by 0.04 kg CaCO3 m-2 yr-1, across all habitats (Table 1 and Figs 6 and 7). A. solaris densities did not reduce carbonate production to negative values, although at two northwestern sites on Kosrae gross carbonate production was reduced by 80% and 62%, where A. solaris densities were 17 and 14 (per 300 m2) and where coral cover was low. For mitigation purposes, and to sustain a productive reef, we found that A. solaris densities should be kept below a density threshold that is proportional to 7.3% of the relative coral densities (Figure B in S1 File). For example, if a 100 m2 site supports 30% live coral cover, any more than two Acanthaster in that site for one year will likely reduce gross carbonate production to zero.
Plain magenta dots are sites that had no observed A. solaris. Base elevation map was plotted in R using raw 10 m Digital Elevation Model from https://pae-paha.pacioos.hawaii.edu/thredds/ncss/usgs_dem_10m_pohnpei/dataset.html.
Plain magenta dots are sites that had no observed A. solaris. Base elevation map was plotted in R using raw 10 m Digital Elevation Model from https://pae-paha.pacioos.hawaii.edu/thredds/ncss/usgs_dem_10m_kosrae/dataset.html.
Other echinoids also reduce carbonate production, particularly Echinometra mathaii, and related species in high densities . We noticed that large populations of E. mathaii were common on southeastern outer reefs and caused considerable erosion (Figures C and D in S1 File) and are less common on inner and patch reefs. Carbonate erosion by parrotfishes was also high on the southeastern reefs and along western reefs, and was lower elsewhere (Figures E and F in S1 File).
This study aimed to identify the spatial variation of shallow-water carbonate production in Pohnpei and Kosrae, Federated States of Micronesia, to assess which reefs are likely to keep up with sea-level rise, and to determine what role Acanthaster solaris plays in carbonate production. While the leeward, northern and northwestern facing reefs had the highest rates of net carbonate production (16.5–18.5 kg CaCO3 m-2 yr-1), the windward, southeastern facing reefs showed the lowest rates of net carbonate production (0.7–1.3 kg CaCO3 m-2 yr-1). Such high variation in carbonate production along the outer shallow-water reefs is important, especially since habitats with low rates may not have the capacity to keep up with predicted sea-level rise. Based on different greenhouse-gas-emission scenarios, most frequently conveyed as Representative Carbon Pathways (RCPs) 2.6. 4.5, 6.0, and 8.5 Wm-2, the predicted rates of sea-level rise by the year 2100 have been conservatively estimated at 5, 6.5, 6.7, and 9 mm yr-1, respectively . Converting the field estimated rates of carbonate production to vertical rates of reef accretion (following Eq 18) the northwestern shallow-water outer reefs of Pohnpei and the northern shallow-water outer reefs of Kosrae are estimated to vertically accrete at 11.8 mm yr-1 and 11.2 mm yr-1, respectively. These rates of vertical accretion are relatively high for contemporary reefs and exceed the rates of sea-level rise under RCP 8.5. Therefore, if effectively managed, the northern shallow-water outer reefs of Pohnpei and Kosrae will likely have the capacity to keep up with sea-level rise and maintain their essential ecosystem functions.
By contrast, the southeastern shallow-water reefs of Pohnpei and Kosrae, have estimated vertical accretion rates of only 1.3 mm yr-1. Although we did not directly measure sedimentation rates nor did we measure micro-bioerosion rates, our data suggest that the southeastern reefs are not likely to keep up with sea-level rise by the year 2100. Indeed, the projections of our model show that without considering spatial variation, on average, Pohnpei’s and Kosrae’s shallow-water outer reefs fall short of the moderate rates of sea-level rise projected under RCP 4.5 (i.e., 6.2 mm yr-1 accretion and 6.5 mm yr-1 of sea-level rise under RCP 4.5, Figure G in S1 File). Most concerning is that the shallow-water inner reefs of both islands have estimated rates of vertical accretion averaging 5.9 mm yr-1, which is lower than most predicted rates of sea-level rise, even the conservative rates associated with RCP 4.5 by the year 2100 (Figure G in S1 File). Although there is some uncertainty in the live-coral-cover thresholds for the inner reefs of Pohnpei and Kosrae, these inner reefs on average require around 5.5% higher live coral cover than outer reefs to produce the same amount of carbonate (Fig 4). These results provide a strong conservation message that in order for nearshore reefs to have a chance to keep up with sea-level rise, it is critical to mitigate land-use discharge and pollution to nearshore shallow-water reefs. Without conserving these relatively sensitive, nearshore reefs our projections suggest that they would likely drown in the near future. Additionally, mitigating terrestrial runoff may also prevent large, persistent outbreaks of Acanthaster [21, 22].
Increases in the survival of Acanthaster brachiolaria-stage larvae have been associated with river discharge and elevated nutrient concentrations [21, 22]. Therefore, implementing management strategies on small Pacific islands that mitigate terrestrial discharge will not only reduce sedimentation stress on corals but will also effectively suppress chronically dense Acanthaster populations that reduce carbonate production. In Kosrae, A. solaris reduced carbonate production by on average 0.6 kg CaCO3 m-2 yr-1, with a maximum reduction of 5.1 kg CaCO3 m-2 yr-1 when A. solaris were at high densities (> 15 individuals per 300 m2). Although our estimates of the impact of A. solaris on carbonate production are novel, they still comprise a degree of uncertainty because of the assumptions underlying Eqs 14–17. Therefore, future improvements in these estimates can be made by examining these assumptions, which could include an adjustment for coral composition.
The coral species that were the most important contributors to carbonate production on both islands were Porites rus and Porites lobata, particularly on the shallow-water inner reefs. The most important contributors to carbonate production on the shallow-water outer reefs were encrusting Montipora, merulinids, and acroporids. There was a lack of patch reefs and large lagoons in Kosrae, therefore Porites cylindrica was less common on Kosrae than on Pohnpei. Yet, if the shallow-water outer reefs are unable to keep up with sea-level rise and are breached by offshore waves, the patch reefs of Pohnpei and the inner reefs of both islands will likely become more similar in coral composition to that of the outer reefs . Still, whether these reefs, even altered in composition, will be able to produce enough carbonate to keep up with sea-level rise is an open question.
The shallow-water coral-reef carbonate production rates measured on both Pohnpei and Kosrae are lower than the field estimates recorded farther west on Palau and Yap (~2.2 kg CaCO3 m-2 yr-1 less, when averaged among habitat types) . The lower carbonate production rates are most likely a result of reduced Acropora cover caused by the recent thermal-stress events on both Kosrae and Pohnpei in 2016 and 2017 (Peter Houk, pers. comm.). At the same time, similar thermal-stress events were not recorded in Palau and Yap. Thermal-stress events are known to significantly reduce a reefs’ capacity to produce calcium carbonate , and under extreme events can temporarily reduce net accretion to negative values . Although there are some studies on the net ecosystem calcification of coral reefs and the influence of coral bleaching on that process [47–51], more field studies are needed that examine (i) thermal-stress events and the dynamics of carbonate production through those events, and (ii) the rates of recovery of carbonate production from thermal-stress events.
The capacity of coral reefs to keep up with rising sea level is important for coastal residents and is particularly relevant to residents of low-lying islands who cannot move to higher elevations. Historically, healthy coral reefs have kept up with dynamic shifts in sea level through glacial-interglacial periods , yet disturbances to modern reefs are suppressing the capacity of coral reefs to produce enough carbonate  and protect island residents from storm-wave damage. In addition, drowned reefs will not be able to provide goods and services or support fisheries. If the coral species contributing to reef complexity and carbonate production are unable to persist under the stress of climate change then the coral reefs will not keep up with sea-level rise and drown.
S1 File. Seven supporting figures.
These figures include (A) a location map, (B) limit on the number of Acanthaster solaris sea stars, as a proportion of live coral cover (LCC) that a 100 m2 shallow-water coral-reef habitat, (C and D) erosional kriged maps for both islands, and (E and F) erosional maps for parrotfishes for both islands, and (G) net vertical accretion split by strata compared to the rates of sea-level rise under representative concentration pathways 2.6, 4.5, 6.0, and 8.5 W m-2.
In Pohnpei, FSM, we would like to sincerely thank Alloise Malfitani and Valentina Permiakova, from the Pohnpei Dive Club, for their warm friendship and hospitality, together with Jerry Martins, Walden Lohn and Robinson Lowdaur for captaining our research vessel, and who all willingly shared their expert local knowledge. Thanks also go to Mark Figueras and staff from the Mangrove Bay Hotel for hosting us in Pohnpei. We would also like to thank Eugene Joseph, and his marine monitoring team for a collaborative opportunity at the Conservation Society of Pohnpei. In Kosrae, our research was made possible by our marvelous hosts Mark, Maria, and Ochi Stevens and staff from the Pacific Treelodge and Micronesian Ecodivers, together with Carlos J. Cianchini, Inston, and Salik Waguk of Kosrae Association of Tourism Operators, who all willingly shared their intimate knowledge of their island environment and ensured our every need was met. We would also like to thank Andy George and his marine monitoring team at the Kosrae Conservation Society for a collaborative opportunity. Special thanks go to Sandra J van Woesik, Kelly J van Woesik, Liana J van Woesik, and Marina L Fleming for their invaluable voluntary field and lab assistance. We would also like to extend our thanks to Sandra J van Woesik and the anonymous reviewers for their editorial comments on the manuscript. This is contribution number 217 from the Institute for Global Ecology at the Florida Institute of Technology.
- 1. Ferrario F, Beck MW, Storlazzi CD, Micheli F, Shepard CC, Airoldi L. The effectiveness of coral reefs for coastal hazard risk reduction and adaptation. Nature Communications. 2014; 5: 3794. pmid:24825660
- 2. Storlazzi CD, Elias E, Field ME, Presto MK. Numerical modeling of the impact of sea-level rise on fringing coral reef hydrodynamics and sediment transport. Coral Reefs. 2011; 30(1): 83–96.
- 3. Loya Y, Sakai K, Yamazato K, Nakano Y, Sambali H, van Woesik R. Coral bleaching: the winners and the losers. Ecology Letters. 2001; 4: 122–131.
- 4. Hoegh-Guldberg O, Mumby PJ, Hooten AJ, Steneck RS, Greenfield P, Gomez E, et al. Coral Reefs Under Rapid Climate Change and Ocean Acidification. Science. 2007; 318: 1737–1742. pmid:18079392
- 5. Baker AC, Glynn PW, Riegl B. Climate change and coral reef bleaching: An ecological assessment of long-term impacts, recovery trends and future outlook. Estuarine, Coastal and Shelf Science. 2008; 80(4): 435–71.
- 6. Ateweberhan M, McClanahan TR. Relationship between historical sea-surface temperature variability and climate change-induced coral mortality in the western Indian Ocean. Marine Pollution Bulletin. 2010; 60(7): 964–70. pmid:20447661
- 7. Perry CT, Murphy GN, Kench PS, Smithers SG, Edinger EN, Steneck RS, et al. Caribbean-wide decline in carbonate production threatens coral reef growth. Nature communications. 2013; 4: ncomms2409.
- 8. Adey W. Coral reef morphogenesis: a multidimensional model. Science.1978; 202: 831–837. pmid:17752443
- 9. Davies PJ. Reef growth. In, Barnes D.J. (ed) Perspectives on Coral Reefs, Australian Institute of Marine Science, Publisher: Brian Clouston. 1983; 69–106.
- 10. Montaggioni LF. History of Indo-Pacific coral reef systems since the last glaciation: Development patterns and controlling factors. Earth-Science Reviews. 2005; 71: 1–75.
- 11. Smith SV, Kinsey DW. Calcium carbonate production, coral reef growth, and sea level change. Science. 1976; 194(4268): 937–939. pmid:17748553
- 12. Kleypas JA 1997 Modeled estimates of global reef habitat and carbonate production since the Last Glacial Maximum Paleoceanography 1997: 12: 533–545.
- 13. Edinger EN, Limmon GV, Jompa J, Widjatmoko W, Heikoop JM, Risk MJ. Normal coral growth rates on dying reefs: are coral growth rates good indicators of reef health? Marine Pollution Bulletin. 2000; 40: 404–425.
- 14. Perry CT, Edinger EN, Kench PS, Murphy GN, Smithers SG, Steneck RS, et al. Estimating rates of biologically driven coral reef framework production and erosion: a new census-based carbonate budget methodology and applications to the reefs of Bonaire. Coral Reefs. 2012; 31(3): 853–868.
- 15. Ryan EJ, Hanmer K, Kench PS. Massive corals maintain a positive carbonate budget of a Maldivian upper reef platform despite major bleaching event. Scientific reports. 2019; 9(1): 6515. pmid:31019243
- 16. Vecsei A. A new estimate of global reefal carbonate production including the fore-reefs. Global and Planetary Change. 2004; 43(1–2): 1–18.
- 17. van Woesik R, Cacciapaglia CW. Keeping up with sea-level rise: Carbonate production rates in Palau and Yap, western Pacific Ocean. PLOS ONE. 2018; 13(5): e0197077. pmid:29738545
- 18. Perry CT and Morgan KM. Bleaching drives collapse in reef carbonate budgets and reef growth potential on southern Maldives reefs. Scientific Reports. 2017; 7: 40581 pmid:28084450
- 19. Januchowski-Hartley FA, Graham NAJ, Wilson SK, Jennings S, Perry CT. Drivers and predictions of coral reef carbonate budget trajectories. Proceedings of the Royal Society B. 2017; 284: 20162533. pmid:28123092
- 20. Chesher RH. Destruction of Pacific corals by the sea star Acanthaster planci. Science. 1969; 165: 280–283. pmid:17814827
- 21. Birkeland C. The Faustian traits of the crown-of-thorns starfish. American Scientist. 1989;77:154–163.
- 22. Fabricius KE, Okaji K, Death G. Three lines of evidence to link outbreaks of the crown-of-thorns seastar Acanthaster planci to the release of larval food limitation. Coral Reefs. 2010; 29, 593–605.
- 23. Pratchett M, Caballes C, Wilmes J, Matthews S, Mellin C, Sweatman H, et al. Thirty years of research on crown-of-thorns starfish (1986–2016): scientific advances and emerging opportunities. Diversity. 2017; 9(4): 41.
- 24. Vermeer M, Rahmstorf S. Global sea level linked to global temperature. Proceedings of the National Academy of Sciences. 2009; 106(51): 21527–21532.
- 25. IPCC. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (eds. Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J et al.), Cambridge University Press, Cambridge, UK and New York, NY, USA. 2013: 1552 pp.
- 26. Jevrejeva S, Grinsted A, Moore JC. Upper limit for sea level projections by 2100. Environmental Research Letters. 2014; 9, 104008.
- 27. Neumann AC, MacIntyre I. Reef response to sea level rise: keep-up, catch-up or give-up. Proc. 5th International Coral Reef Congress. 1985; 3: 105–110.
- 28. van Woesik R, Golbuu Y, Roff G. Keep up or drown: adjustment of western Pacific coral reefs to sea-level rise in the 21st century. Royal Society Open Science 2. 2015; (150181).
- 29. Pebesma EJ, Bivand RS. Classes and methods for spatial data in R. R News 2005; 5(2).
- 30. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, 2012. ISBN 3-900051-07-0; 2018.
- 31. Hijmans R, van Etten J. Raster: Geographic Data Analysis and Modeling, no. R package version 517. 2014; 2–2.1.
- 32. Perry CT, Spencer T, Kench PS. Carbonate budgets and reef production states: a geomorphic perspective on the ecological phase-shift concept. Coral Reefs. 2008; 27(4): 853–866.
- 33. van Woesik R. Quantifying uncertainty and resilience on coral reefs using a Bayesian approach. Environmental Research Letters. 2013; 8(4): 044051.
- 34. Rogers CS. Sublethal and lethal effects of sediments applied to common Caribbean reef corals in the field. Marine Pollution Bulletin. 1983; 14: 378–382.
- 35. Moran PJ. The Acanthaster phenomenon. Oceanogr. Mar. Biol. Annu. Rev. 1986; 24: 379–480.
- 36. Hubbard DK. Reefs as dynamic systems. In, Birkeland C. (ed) Life and Death of Coral Reefs, Publisher, Chapman and Hall, USA. 1997; 43–67.
- 37. Victor S, Neth L, Golbuu Y, Wolanski E, Richmond RH. Sedimentation in mangroves and coral reefs in a wet tropical island, Pohnpei, Micronesia. Estuarine, Coastal and Shelf Science. 2006; 66(3–4): 409–416.
- 38. Ong L, Holland KN. Bioerosion of coral reefs by two Hawaiian parrotfishes: Species, size differences and fishery implications. Marine Biology. 2010; 157 (6): 1313–1323.
- 39. Bonaldo RM, Bellwood DR. Size-dependent variation in the functional role of the parrotfish Scarus rivulatus on the Great Barrier Reef, Australia. Marine Ecology Progress Series. 2008; 360: 237–244.
- 40. Glynn PW. Bioerosion and coral-reef growth: a dynamic balance. In, Birkeland C (ed) Life and Death of Coral Reefs, Publisher, Chapman and Hall, USA. 1997; 68–95.
- 41. Glynn PW. Acanthaster: effect on coral reef growth in Panama. Science. 1973; 180(4085): 504–506. pmid:17817814
- 42. Keesing JK, Lucas JS. Size-specific locomotion rate and movement pattern of four common Indo-Pacific sea stars (Echinodermata; Asteroidea). Journal of Experimental Marine Biology and Ecology. 1992; 156: 89–104.
- 43. Mueller B, Bos AR, Graf G, Gumanao GS. Size-specific locomotion rate and movement pattern of four common Indo-Pacific sea stars (Echinodermata; Asteroidea). Aquatic Biology. 2011; 12(2): 157–164.
- 44. Zuur AF, Saveliev AA, Ieno EN. A beginner’s guide to generalised additive mixed models with R. Newburgh, United Kingdom.: Highland Statistics Ltd.; 2014; 332 pp.
- 45. Wand MP, Ormerod JT. On semiparametric regression with O’Sullivan penalized splines. Australian & New Zealand Journal of Statistics. 2008; 50(2): 179–198.
- 46. Plummer M. rjags: Bayesian graphical models using MCMC. R package version. 2013; 3(10).
- 47. Engels MS, Fletcher CH, Field ME, Storlazzi CD, Grossman EE, Rooney JJ, et al. Holocene reef accretion: southwest Molokai, Hawaii, USA. Journal of Sedimentary Research. 2004; 74(2): 255–269.
- 48. Kayanne H, Hata H, Kudo S, Yamano H, Watanabe A, Ikeda Y et al. Seasonal and bleaching‐induced changes in coral reef metabolism and CO2 flux. Global Biogeochemical Cycles. 2005; 19(3): GB3015.
- 49. Courtney TA, De Carlo EH, Page HN, Bahr KD, Barro A, Howins N et al. Recovery of reef‐scale calcification following a bleaching event in Kāne'ohe Bay, Hawai'i. Limnology and Oceanography Letters. 2018; 3(1): 1–9.
- 50. McMahon A, Santos IR, Schulz KG, Scott A, Silverman J, Davis KL et al. Coral reef calcification and production after the 2016 bleaching event at Lizard Island, Great Barrier Reef. Journal of Geophysical Research. Oceans. 2019; 124: 4003–4016.
- 51. DeCarlo TM, Cohen AL, Wong GTF, Shiah FK, Lentz SJ, Davis KA et al. Community production modulates coral reef pH and the sensitivity of ecosystem calcification to ocean acidification, J. Geophys. Res. Oceans. 2017; 122.
- 52. Chappell J, Polach H. Post-glacial sea-level rise from a coral record at Huon Peninsula, Papua New Guinea. Nature. 1991; 349: 147–149.
- 53. Perry CT, Alvarez-Filip L, Graham NAJ, Mumby PJ, Wilson SK, Kench PS et al. Loss of coral reef growth capacity to track future increases in sea level. Nature. 2018; 558: 396–400. pmid:29904103