Skeletal growth capacity as a measure of coral species and community resilience

Abstract

Accretion and erosion of scleractinian (stony coral) carbonate skeletons determine whether a colony will increase or decrease in size with potential consequences for ecosystem processes, functions and services. The capacity for skeletal growth can be estimated by comparing a colony’s rate of calcification with its rate of erosion. Calcification depends on the species-specific metabolic activity of living tissue, and erosion depends primarily on the availability and density of barren skeleton, those areas on the colony where polyps have died. Assessment of skeletal growth capacity requires data on calcification rates, erosion rates and both live and barren colony surface area. Rates of calcification and erosion are documented for many Caribbean species and others can be estimated from existing data. Three-dimensional surface area of colonies can be determined from data collected during demographic surveys, which identify species, measure dimensions, and estimate the proportion of live tissue on a colony. Data from demographic surveys conducted in the U.S. Virgin Islands are used to calculate the skeletal growth capacity (GC) as an indicator of coral species and community resilience. Scleractinia are the primary architects of coral reefs, and the gain or loss of skeletal framework is vitally important to reef ecosystem processes that lead to valued goods and services. Estimates of GC reflect stony coral resilience, which is the capacity to recover from disturbances by returning to previous physical and functional levels. GC can also provide insight to the effects of stressors such as ocean acidification, and can inform several management decisions, including restoration site selection and threatened species designation.

1. Introduction

Coral reefs worldwide are degrading from local and global stressors, a condition that is particularly acute in the Caribbean and western Atlantic (Gardner et al., 2003; Jackson et al., 2014). The primary architects of coral reefs are Scleractinia, or stony corals, which construct calcium carbonate (CaCO₃) structures as they grow. Communities of stony corals have undergone dramatic changes during the last several decades, including altered species composition, loss of living coral tissue and diminished topographic complexity. Importantly, large and morphologically complex species are dying or losing the live tissue necessary for future growth and reproduction. These foundational species are being supplanted by smaller colonies that are more tolerant to environmental stressors but provide less reef structure critical to reef functions, processes, and ecosystem services (Burman et al., 2012; González-Barriosa and Lorenzo Álvarez-Filip, 2018; Toth et al., 2019). The consequences of colony degradation and community shift include species homogenization, loss of architectural complexity (reef framework), and the ‘flattening’ of reefs (Green et al., 2008; Alvarez-Filip et al., 2009, 2011; Burman et al., 2012).

The capacity to gain or lose reef structure can be examined through the lens of individual stony coral colonies aggregated to species and community levels (Chave et al., 1972). A colony’s capacity for skeletal growth incorporates the offsetting processes of CaCO₃ accretion and erosion, i.e., the capacity for construction of new framework vs degradation of existing framework (Wild et al., 2011; Graham and Nash 2013; Kuffner et al., 2013; González-Barriosa and Lorenzo Álvarez-Filip, 2018; Lange et al., 2020). Calculation of skeletal growth capacity for a colony requires four pieces of information—the species-specific rates of calcification and erosion and the live and barren coral surface area engaged in each. Calcification can only occur where there is live tissue, so calcification rates are applied to the live coral surface area of a colony. Erosion occurs predominantly on barren skeleton so erosion rates, which are dependent on CaCO₃ density, are applied to the bare or dead surface area of a colony. Combined, these relationships provide a snapshot, or point-in-time estimate, of the offsetting capacities for skeletal gains and losses on a colony. Aggregating the values from individual colonies can provide useful species- and community-level information on stony coral resilience—the capacity to maintain, augment or lose the structural framework of a reef after a disturbance. Implications are consequential for coral reef functional roles that rely on stony coral size, topography, and architecture such as fish habitat and shelter, shoreline protection and carbon sequestration (Moberg and Folke, 1999; Principe et al., 2012; Yee et al., 2014).

Rates of calcification vary across and within species and are influenced by several factors (Chappell, 1980; Glynn, 1997; Mollica et al., 2018. González-Barriosa and Lorenzo Álvarez-Filip (2018) presented calcification rates for 47 Caribbean coral species using published literature on extension rates, skeletal density, and consideration of colony morphology. Rates of erosion are also variable and dependent on multiple factors (Glynn, 1997; Chazottes et al., 2002), including species differences in skeletal density. Some carbonate flux studies have calculated rates of erosion indirectly by combining a field census of eroding organisms with their estimated rates of erosion. A more direct method was to measure in situ erosion of Orbicella spp. colonies over a 17-yr period in the Florida Keys (Kuffner et al., 2019). Their reported average rate of erosion (8.2 kg CaCO₃ m⁻² yr⁻¹) is adapted here, using differences in skeletal density previously summarized by Hughes (1987), to estimate erosion rates of other Caribbean species.

Studies generating reef-wide carbonate budgets have estimated three-dimensional live coral surface area by combining transect rugosity (an estimate of colony heights) with two-dimensional live coral cover measurements (Perry et al., 2012). A more direct quantification is possible using data collected from colony-based, or ‘demographic’ assessment surveys. Initially used by Ginsburg et al. (1996), demographic surveys have been adopted by several reef assessment programs (Kramer, 2003; Fisher et al., 2014; NOAA, 2014; FFWC, 2019; Fisher et al., 2019). Demographic surveys measure at least two colony dimensions (usually height and maximum diameter) and partial mortality (barren skeleton) or its converse, percent live tissue. Applying geometric surrogates to account for species morphological differences (Alcala and Vogt, 1997; Bythell et al., 2001; Fisher, 2007; Courtney et al., 2007), these measurements can be converted into estimates of three-dimensional live and barren (‘dead’) colony surface area for use in calculation of accretion and erosion capacity.

Stony coral skeletal growth capacity (GC) can therefore be calculated from the combination of demographic coral data and previously published calcification and erosion rates. This unique indicator is applied to data from demographic surveys of corals in U.S. Virgin Islands (Fisher et al., 2014) to determine its ability to distinguish capacity for skeletal growth among colonies, species and communities. Although dependent on data from several sources, the indicator provides an easily calculated and easily understood gauge of skeletal growth capacity for different species and scleractinian communities. Its application can be invaluable for informing threatened species designations, siting of restoration projects, evaluating reef resilience and forecasting ecosystem services.

2. Methods

2.1. Formulation

Skeletal growth capacity (GC) was calculated for individual stony coral colonies documented during two field surveys in the U.S. Virgin Islands, and then aggregated for species- and community-level values. GC compares the capacity for calcification by living tissue (live colony surface area, LCSA) versus the capacity for erosion of bare or dead coral skeleton (dead colony surface area, DCSA):

$$\mathrm{GC}=\left(\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{c}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\times \mathrm{L}\mathrm{C}\mathrm{S}\mathrm{A}\right)-\left(\mathrm{e}\mathrm{r}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\times \mathrm{D}\mathrm{C}\mathrm{S}\mathrm{A}\right)$$

where GC is calculated for each colony independently and summed for species and community values. Calcification rates (CR) and erosion rates (ER) are reported in kg CaCO₃ m⁻² coral surface area yr⁻¹ and three-dimensional coral surface areas (LCSA and DCSA) are reported in m². The resulting values estimate the kg CaCO₃ yr⁻¹ of coral skeleton that could be gained or lost based on colony condition at the time of the survey. Derivation of each of the four components is described below and species-specific values for CR and ER, including 6 species listed as threatened under the Endangered Species Act (ESA), are listed in Table 1.

Table 1.

Species values for morphology (M), calcification rates (CR, kg CaCO₃ m⁻² yr⁻¹), skeletal densities (Sk Den, g CaCO₃ cm⁻²) and erosion rates (ER, kg CaCO₃ m⁻² yr⁻¹) used in calculation of GC. M was assigned based on species morphology (flat, hemispheric, globular, or branched) and used to calculate colony surface area. CR was adopted from González-Barriosa and Lorenzo Álvarez-Filip (2018) and ER calculated from Kuffner et al. (2019) using skeletal densities provided by Hughes (1987).

Species	M	CR	Sk Den	ER	Species	M	CR	Sk Den	ER
Acropora cervicornis *	4	19.28	1.315	10.10	Meandrina meandrites	1	1.93	1.639	8.10
Acropora palmata *	4	17.94	1.615	8.23	Montastraea cavernosa	2	6.62	1.60	8.30
Acropora prolifera	4	9.41	1.315	10.10	Mycetophyllia aliciae	1	0.08	1.639	8.10
Agaricia agaricites	1	4.27	1.945	6.83	Mycetophyllia lamarckiana	1	0.08	1.639	8.10
Agaricia fragilis	1	0.10	2.31	5.75	Orbicella annularis *	3	10.71	1.62	8.20
Agaricia humilis	1	5.28	2.09	6.36	Orbicella faveolata *	2	11.49	1.62	8.20
Agaricia lamarcki	1	0.09	2.28	5.83	Orbicella franksi *	2	9.20	1.62	8.20
Cladocera arbuscula	2	6.28	1.639	8.10	Porites astreoides	2	5.78	1.465	9.07
Colpophyllia natans	2	4.47	0.745	17.83	Porites colonensis	1	5.78	1.45	9.16
Dendrogyra cylindrus *	3	12.13	1.639	8.10	Porites divaricata	3	1.45	1.45	9.16
Dichocoenia stokesi	3	4.09	2.17	6.12	Porites furcata	3	1.86	1.05	12.65
Diploria labyrinthiformis	2	5.86	1.57	8.46	Porites porites	3	5.15	1.18	11.26
Eusmilia fastigiata	2	8.19	1.30	10.22	Pseudodiploria clivosa	2	5.13	1.57	8.46
Helioceris cucullata	1	0.09	1.639	8.10	Pseudodiploria strigosa	2	5.23	1.57	8.46
Isophyllastrea rigida	2	3.76	1.639	8.10	Siderastrea siderea	2	6.51	1.61	8.25
Isophyllia sinuosa	2	3.57	1.639	8.10	Solenastrea bournoni	2	14.08	1.639	8.10
Madracis decactis	3	13.70	1.68	7.91	Stephanocoenia intersepta	2	4.05	1.639	8.10
Madracis mirabilis	3	6.28	1.68	7.91

^*Species listed as threatened under the Endangered Species Act.

2.2. Calcification rates

Calcification rates (CR) for 43 Atlantic and Caribbean stony coral species were reported by González-Barriosa and Lorenzo Álvarez-Filip (2018). Included were rates for 32 of the 35 species documented in two U.S. Virgin Island surveys (Fisher et al. 2014, see below). One species, Porites colonensis, was not reported and was assigned the CR for the morphologically similar Porites astreoides. Two other species, Cladocera arbuscula and Madracis mirabilis, were assigned the average of the 32 species that were listed in González-Barriosa and Lorenzo Álvarez-Filip (2018) and were present in the surveys (CR = 6.28 kg m⁻² yr⁻¹, Table 1). Other studies have documented calcification rates, such as Kuffner et al. (2013) for Siderastrea siderea, but a single source was used here both for consistency and because the values were reported for three-dimensional rather than planar colony surface areas.

2.3. Erosion rates

Kuffner et al. (2019) measured the erosion rate (ER) of orbicellid corals over a 17-yr period in the Florida Keys and recorded an average loss of 8.2 kg CaCO₃ m⁻² yr⁻¹. Because erosion is influenced by skeletal density, erosion rates for other species were calculated in relation to the density of Orbicella faveolata (1.62 g cm⁻²; Hughes, 1987). For example, if a species had twice the skeletal density it was assigned half the erosion rate of O. faveolata. Skeletal densities for 16 of the 35 species identified in the two surveys were reported by Hughes (1987). Where more than one value was reported for a species the midpoint was used; where no value was reported a value from the same genus was assigned (10 species); and where no value or like genus was recorded, the average density for all 16 species that were recorded and present in the surveys was assigned (1.649 g cm⁻²) with the resulting erosion rate of 8.1 kg CaCO₃ m⁻² yr⁻¹ (Table 1). For perspective, Hughes (1987) notes that the highest potential density for any coral is 2.94 g cm⁻², which is the density of solid aragonite.

2.4. Three-dimensional LCSA and DCSA

Data from a demographic survey in the U.S. Virgin Islands (Fisher et al., 2014) were used to calculate the three-dimensional (3D) surface area of individual colonies. Surveys documented colonies from shallow coral reefs (1–12 m depth) at St. Croix in 2007 (STX, 51 stations) and at St. Thomas and St. John in 2009 (STJ, 52 stations). In both surveys, a probabilistic survey design was implemented, covering 46.2 km² of sea floor for STX and 31.8 km² for STJ. Surveys documented stony corals occurring in each transect by recording species, height, maximum diameter, and percent of live tissue (%LT) on each colony in a transect. Estimation of %LT is the converse of ‘partial mortality’ measured in similar demographic surveys (Kramer 2003, NOAA 2014) and both work equally well for GC calculations. Different methods have been used to estimate 3D colony surface area (CSA) from colony dimensions (Alcala and Vogt, 1997; Bythell et al., 2001; Fisher et al., 2007; Courtney et al., 2007; Naumann et al., 2009). Here, a hemispheric surrogate was applied such that CSA = Mπr’², where r’ was calculated as the average of height and ½ the diameter, and M was a species-specific morphological factor ranging from 1 to 4 to account for structural complexity; flat colonies M = 1, hemispheric colonies M = 2, lobed and domed colonies M = 3, and branched colonies M = 4 (Fisher et al. 2014, Table 1). Live colony surface area was calculated as LCSA= (CSA)(%LT/100)). Bare skeleton or dead surface area was calculated as DCSA = CSA-LCSA. Any colonies noted in the surveys with 0 %LT (‘standing dead’ colonies) were excluded because they were not routinely reported by all surveyors.

2.5. Growth capacity calculations

GC was calculated on seven Orbicella annularis colonies selected from survey data as examples to illustrate the calculations and resulting ranges of variability. CR and ER values from Table 1 were used for O. annularis and LCSA and DCSA were calculated from CSA and %LT using the assigned morphological factor M = 3. GC was then calculated on a single colony of O. annularis to illustrate the variability in LCSA, DCSA and GC values when %LT was artificially changed from 10 to 100 %. The direction (positive or negative) and value of GC depends on the relation of CR, ER and the percent of live tissue on a colony (%LT). A species benchmark (LT₀) was calculated to identify the %LT for each species where GC = 0. LT₀ represents the percent of live tissue where the capacity for growth and erosion for a species are equal, regardless of colony size. It was calculated from the relationships (CR)(LCSA)=(ER) (DCSA) and CSA = LCSA + DCSA where CSA = 100 % colony surface area. Average %LT for each species in both surveys were compared with the calculated LT₀.

2.6. Application to field data

Using assigned M, ER and CR values (Table 1), GC was calculated for all colonies recorded in two surveys in U.S. Virgin Islands and summed for species-level and community-level GC values. For spatial context, data were normalized to the area sampled. In the STX survey, colonies were documented from 51 – 25.1 m² radial belt transects (1280 m²) and in the STJ survey from 52 – 25 m² linear belt transects (1300 m²). Consequently, the community GC value for each region was divided by 0.00128 km² and 0.0013 km², respectively, to normalize values to one square kilometer of sea floor within each region. Because these were probabilistic surveys, the values are representative of the entire region sampled.

3. Results

3.1. Species calcification and erosion rates

Calcification rates (CR) for different species ranged from 0.08 to 19.28. kg CaCO₃ m⁻² yr⁻¹ (Table 1). Five species, Mycetophyllia lamarckiana, M. aliciae, Helioceris cucullata, Agaricia lamarcki, and A. fragilis had very low CR (0.08–0.1 kg m⁻² yr⁻¹). Erosion rates ranged from 5.75 to 17.83 kg CaCO₃ m⁻² yr⁻¹; the highest was for Colpophyllia natans and all others ranged from 5.75 to 12.65 kg m⁻² yr⁻¹.

3.2. Calculations of GC and LT₀

Seven example Orbicella annularis colonies ranged from GC = −35.71 to 49.46 kg CaCO₃ yr⁻¹ (Table 2) with the transition from positive to negative GC occurring between colonies 4 (50 %LT) and 5 (40 %LT). Artificially reducing from 100 % to 10 % LT for an O. annularis colony altered GC from 49.46 to −29.15 kg CaCO₃ yr⁻¹ (Table 3). For O. annularis LT₀ = 43.4 %, i.e., the percent of live tissue where calcification and erosion are equal.

Table 2.

Coral skeletal growth capacity (Calcification – Erosion) calculated for seven example Orbicella annularis colonies recorded in USVI surveys. For O. annularis, M factor = 3, calcification rate (CR) = 10.7, and erosion rate (ER) = 8.2 kg CaCO₃ m⁻² yr⁻¹. Growth capacity (GC) values are listed in decreasing order to demonstrate the transition from positive to negative GC between colonies 4 (50 %LT) and 5 (40 %LT) and coinciding with a calculated LT₀ = 43.4 % for O. annularis. Radius (r’) used in CSA calculations was the average of Ht and 1/2 Diam.

Colony	Ht (cm)	Diam (cm)	Rad (r’) (cm)	CSA (m2)	%LT (%)	LCSA (m2)	DCSA (m2)	Calcification (kg yr−1)	Erosion (kg yr−1)	GC (kg yr−1)
1	90	100	70.0	4.62	100	4.62	0.00	49.46	0.00	49.46
2	23	16	15.5	0.23	100	0.23	0.00	2.43	0.00	2.43
3	13	22	12.0	0.14	70	0.10	0.04	1.02	0.33	0.68
4	25	30	20.0	0.38	50	0.19	0.19	2.02	1.55	0.47
5	25	36	21.5	0.44	40	0.17	0.26	1.87	2.14	−0.28
6	60	100	55.0	2.85	10	0.29	2.57	3.05	21.04	−17.99
7	90	130	77.5	5.66	10	0.57	5.09	6.06	41.78	−35.71

Ht = height, Diam = diameter, Rad = radius, CSA = colony surface area, %LT = percent live tissue, LCSA and DCSA = living and dead colony surface area.

Table 3.

Coral skeletal growth capacity (GC, kg CaCO₃ yr⁻¹) calculations for a single Orbicella annularis colony (colony 1 in Table 2) with artificial changes in %LT (10–100 %). Calcification and erosion are equal at %LT = 43.36 for O. annularis, i.e., LT₀ = 43.36 %.

%LT	LCSA (m2)	DCSA (m2)	Calcification (kg yr−1)	Erosion (kg yr−1)	GC (kg yr−1)
100	4.62	0.00	49.46	0.00	49.46
75	3.47	1.16	37.11	9.47	27.64
50	2.31	2.31	24.74	18.94	5.80
43.36	2.01	2.61	21.43	21.46	0
25	1.16	3.47	12.37	28.41	−16.04
10	0.46	4.16	4.95	34.10	−29.15

%LT = percent live tissue on a colony, LCSA = live colony surface area, DCSA = dead colony surface area.

3.3. Species LT₀

LT₀ values for the 35 recorded species ranged from 31.44 % (Acropora palmata) to 99.02 % (Mycetophyllia aliciae and M. lamarckiana), with most species (27) exhibiting LT₀ > 50 % and the community average LT₀ = 63.61 % (Table 4). Comparisons revealed several instances (5 each for STX and STJ, bolded values) where the average %LT for a species was less than LT₀. Four species (Agaricia fragilis, Orbicella annularis, Porites divaricata and Porites furcata) had average %LT less than LT₀ in both regions. Orbicella annularis dominated the amount of CSA provided by corals in both regions (Fig. 1) but the average %LT was much lower for STX (20.4 %) relative to STJ (39.7 %). Both values were less than the LT₀ of 43.4 % for O. annularis (Table 4).

Table 4.

Comparison of LT₀ with the average %LT of each species documented in St Croix (STX) and St. John/St. Thomas (STJ). LT₀ = percent of live tissue for a colony where GC = 0, regardless of colony size. Bold values highlight those species where the average %LT is less than LT₀ and the number of colonies are recorded in parentheses.

Species	LT0 (%)	STX Avg %LT (# col)	STJ Avg %LT (# col)
Acropora cervicornis	34.38	80.0 (1)	59.0 (42)
Acropora palmata	31.44	–	75.0 (8)
Acropora prolifera	51.77	–	82.5 (4)
Agaricia agaricites	61.53	92.0 (20)	90.2 (51)
Agaricia fragilis	98.29	95.0 (4)	93.3 (15)
Agaricia humilis	54.62	–	100 (2)
Agaricia lamarcki	98.48	–	100 (3)
Cladocera arbuscula	57.14	90.0 (1)	–
Colpophyllia natans	79.96	69.2 (13)	82.8 (9)
Dendrogyra cylindrus	40.05	–	94.8 (23)
Dichocoenia stokesi	59.95	66.5 (23)	76.5 (13)
Diploria labyrinthiformis	59.08	70.0 (11)	82.5 (88)
Eusmilia fastigiata	55.51	57.5 (6)	67.6 (19)
Helioceris cucullata	98.90	–	90.0 (1)
Isophyllastrea rigida	68.31	–	90.0 (1)
Isophyllia sinuosa	69.42	100 (1)	–
Madracis decactis	36.60	68.9 (19)	87.1 (33)
Madracis mirabilis	56.53	90.0 (2)	80.0 (3)
Meandrina meandrites	80.77	94.5 (31)	92.7 (57)
Montastraea cavernosa	55.64	63.6 (268)	64.2 (169)
Mycetophyllia aliciae	99.02	–	100 (5)
Mycetophyllia lamarckiana	99.02	100 (1)	100 (1)
Orbicella annularis	43.36	20.4 (132)	39.7 (233)
Orbicella faveolata	41.65	53.4 (96)	66.9 (140)
Orbicella franksi	47.13	65.3 (17)	72.8 (88)
Porites astreoides	61.07	87.2 (276)	88.7 (476)
Porites colonensis	61.32	–	88.7 (15)
Porites divaricata	86.34	73.3 (6)	70.0 (14)
Porites furcata	87.18	58.1 (77)	74.8 (65)
Porites porites	68.61	73.5 (63)	73.5 (166)
Pseudodiploria clivosa	62.25	88.8 (59)	83.8 (39)
Pseudodiploria strigosa	61.80	79.3 (270)	88.1 (165)
Siderastrea siderea	55.90	73.2 (257)	77.2 (414)
Solenastrea bournoni	36.53	–	100 (1)
Stephanocoenia intersepta	66.68	90.5 (39)	86.8 (34)
Community Average	63.61	70.3 (1,693)	75.4 (2,397)

Fig. 1.

Comparison of colony surface area (CSA, m²) with live colony surface area (LCSA, m²) for each species recorded at STX and STJ. Dotted line represents LCSA/CSA = 0.50 or 50 % live surface area. Species noted are Dendrogyra cylindrus (Dcyl), Montastraea cavernosa (Mcav), Orbicella annularis (Oann), O. faveolata (Ofav), Porites astreoides (Past), Pseudodiploria strigosa (Pstr) and Siderastrea siderea (Ssid).

3.4. Community GC values

3.4.1. St Croix

Calculation of GC for 25 species at STX and 33 species at STJ (Tables 5 and 6) showed a varying capacity for skeletal growth for different species and coral communities of each region. At STX, 12 of the 25 species and 525 of the 1,693 colonies (31 %) exhibited negative GC (Table 5), contributing to a negative community value of −1,192.77 kg CaCO₃ yr⁻¹ (n = 51 transects; average −23.4; range −383.3 to 36.4; standard deviation 67.4). Strong negative GC were found for Orbicella annularis, O. faveolata, Porites furcata and Montastraea cavernosa. Orbicella annularis had the greatest impact on the community GC (Fig. 2) with a dominant CSA coupled with a low %LT. The strongest positive GC at STX was for Psuedodiploria strigosa followed by Porites astreoides and Pseudodiploria clivosa. Three of the ESA threatened species were not found at STX (Acropora palmata, Dendrogyra cylindrus and Mycetophyllia ferox) but GC for A. cervicornis was slightly positive (2.63), slightly negative for Orbicella franksi (−2.79) and strongly negative for O. annularis (−1047.94) and O. faveolata (−84.81). The negative community value at STX (−1,192.77 kg CaCO₃ yr⁻¹) divided by the area of sea floor sampled (0.00128 km²) yields a potential loss of 931,852 kg CaCO₃ km⁻² yr⁻¹ across the region sampled. The same calculation can be applied to species GC values to estimate potential gains or losses of CaCO₃ for different species in the region.

Table 5.

Coral skeleton growth capacity (GC, kg CaCO₃ yr⁻¹) for 25 species documented at St. Croix (STX) in 2007. Values for live colony surface area (LCSA), dead colony surface area (DCSA), calcification (Calc) and erosion were summed across colonies of each species and GC values are ranked by species from high to low.

Species	∑ LCSA (m2)	∑ DCSA (m2)	∑ Calc (kg yr−1)	∑ Erosion (kg yr−1)	GC (kg yr−1)	GC Rank
Acropora cervicornis	0.16	0.04	3.03	0.40	2.63	6
Agaricia agaracites	0.41	0.06	1.73	0.39	1.35	7
Agaricia fragilis	0.02	0.00	0.00	0.01	−0.01	14
Cladocera arbuscula	0.01	0.00	0.09	0.01	0.08	11
Colpophyllia natans	4.57	1.72	20.42	30.60	−10.18	19
Dichocoenia stokesi	1.34	0.82	5.49	5.00	0.49	10
Diploria labyrinthiformis	1.49	1.18	8.74	9.99	−1.25	15
Eusmilia fastigiata	0.25	0.47	2.05	4.79	−2.74	16
Isophyllia sinuosa	0.02	0.00	0.06	0.00	0.06	12
Madracis decactis	1.47	1.12	20.08	8.84	11.24	4
Madracis mirabilis	0.12	0.02	0.78	0.19	0.59	9
Meandrina meandrites	0.66	0.07	1.27	0.60	0.68	8
Montastraea cavernosa	27.62	30.18	182.85	250.62	−67.77	22
Mycetophyllia lamarckiana	0.01	0.00	0.00	0.00	0.00	13
Orbicella annularis	38.46	178.03	411.87	1459.81	−1047.94	25
Orbicella faveolata	16.43	33.37	188.82	273.62	−84.81	24
Orbicella franksi	1.45	1.96	13.29	16.08	−2.79	17
Porites astreoides	10.25	2.18	59.26	19.77	39.49	2
Porites divaricata	1.13	0.87	1.64	7.94	−6.31	18
Porites furcata	6.08	7.29	11.31	92.27	−80.95	23
Porites porites	4.83	3.29	24.85	36.98	−12.13	20
Pseudodiploria clivosa	6.36	1.50	32.61	12.73	19.88	3
Pseudodiploria strigosa	27.06	9.90	141.53	83.76	57.77	1
Siderastrea siderea	19.94	17.31	129.79	142.82	−13.04	21
Stephanocoenia intersepta	1.07	0.18	4.31	1.43	2.88	5
Community Total	171.19	291.55	1265.88	2458.65	−1192.77

Table 6.

Coral skeleton growth capacity (GC, kg CaCO₃ yr⁻¹) for 33 species documented at St. Thomas and St. Johns (STJ) in 2009. Values for live colony surface area (LCSA), dead colony surface area (DCSA), calcification (Calc) and erosion were summed across colonies of each species and GC values are ranked by species from high to low.

Species	∑ LCSA (m2)	∑ DCSA (m2)	∑ Calc (kg yr−1)	∑ Erosion (kg yr−1)	GC (kg yr−1)	GC Rank
Acropora cervicornis	12.66	13.26	244.17	133.98	110.19	3
Acropora palmata	8.89	3.76	159.56	30.92	128.64	2
Acropora prolifera	0.72	0.13	6.77	1.28	5.49	10
Agaricia agaricites	0.67	0.11	2.86	0.72	2.14	12
Agaricia fragilis	0.14	0.01	0.01	0.06	−0.04	25
Agaricia humilis	0.01	0.00	0.06	0.00	0.06	19
Agaricia lamarcki	0.12	0.00	0.01	0.00	0.01	22
Colpophyllia natans	0.41	0.12	1.85	2.23	−0.38	26
Dendrogyra cylindrus	24.69	4.35	299.46	35.25	264.22	1
Dichocoenia stokesi	0.47	0.47	1.91	2.85	−0.94	27
Diploria labyrinthiformis	5.53	2.20	32.38	18.63	13.74	8
Eusmilia fastigiata	0.36	0.21	2.94	2.11	0.83	16
Helioceris cucullata	0.00	0.00	0.00	0.00	0.00	24
Isophyllastrea rigida	0.01	0.00	0.05	0.01	0.04	20
Madracis decactis	2.06	0.73	28.22	5.77	22.45	7
Madracis mirabilis	0.14	0.03	0.87	0.22	0.65	17
Meandrina meandrites	1.44	0.17	2.78	1.41	1.37	14
Montastraea cavernosa	17.80	19.07	117.85	158.33	−40.48	32
Mycetophyllia aliciae	0.15	0.00	0.01	0.00	0.01	21
Mycetophyllia lamarckiana	0.02	0.00	0.00	0.00	0.00	23
Orbicella annularis	67.96	147.36	727.86	1208.34	−480.48	33
Orbicella faveolata	23.48	29.85	269.77	244.74	25.03	6
Orbicella franksi	13.50	6.89	124.22	56.49	67.73	5
Porites astreoides	21.65	4.04	125.13	36.66	88.47	4
Porites colonensis	0.35	0.07	2.00	0.67	1.33	15
Porites divaricata	0.49	0.28	0.72	2.59	−1.87	29
Porites furcata	5.28	2.44	9.82	30.93	−21.11	31
Porites porites	12.97	5.40	66.78	60.78	5.99	9
Pseudodiploria clivosa	1.59	0.73	8.14	6.18	1.96	13
Pseudodiploria strigosa	9.04	5.74	47.28	48.56	−1.28	28
Siderastrea siderea	31.26	25.09	203.50	207.03	−3.53	30
Solenastrea bournoni	0.01	0.00	0.12	0.00	0.12	18
Stephanocoenia intersepta	0.98	0.21	3.97	1.71	2.26	11
Community Total	264.86	272.72	2491.09	2298.45	192.64

Fig. 2.

Comparison of calcification capacity (kg CaCO₃ yr⁻¹) with erosion capacity (kg CaCO₃ yr⁻¹) for each species recorded at STX and STJ. Dotted line represents calcification capacity - erosion capacity = 0. Points above the line have a greater capacity for calcification and those below the line have a greater capacity for erosion. Species noted are Acropora cervicornis (Acer), Dendrogyra cylindrus (Dcyl), Montastraea cavernosa (Mcav), Orbicella annularis (Oann), O. faveolata (Ofav), Porites astreoides (Past), P. furcata (Pfur), Pseudodiploria strigosa (Pstr) and Siderastrea siderea (Ssid). Axis scales are different for STX and STJ.

3.4.2. St. Thomas and St. John

In contrast, 24 of the 33 stony coral species and 1,805 of the 2,397 colonies (75 %) at STJ exhibited a positive GC (Table 6), resulting in a positive community value of 192.64 kg CaCO₃ yr⁻¹ (n = 53 transects; average 3.7; range −155.7 to 239.5; standard deviation 55.0). The highest positive values were for Dendrogyra cylindrus (264.22 kg yr⁻¹), Acropora palmata (128.64 kg yr⁻¹) and A. cervicornis (110.19 kg yr⁻¹), three of the ESA listed species. Among the remaining listed species Orbicella faveolata and O. franksi were also strongly positive, but O. annularis exhibited the lowest negative GC (−480.48 kg yr⁻¹) for the region. However, this did not impact the community value as much as the negative GC for O. annularis at STX (Fig. 2). The positive community value at STJ (192.64 kg CaCO₃ yr⁻¹) divided by the area of sea floor sampled (0.0013 km²) yields a potential gain of 148,185 kg CaCO₃ km⁻² yr⁻¹ across the region. The same calculation can be applied to species GC values to estimate potential gains or losses of CaCO₃ for different species in the region. A t-test comparing transect GC revealed a significant difference (p = 0.027) between STJ and STX.

4. Discussion

4.1. Indicator utility

Worldwide degradation of reefs has been measured in terms of altered species composition, loss of living coral tissue and reduced topographic complexity (Burman et al., 2012; González-Barriosa and Lorenzo Álvarez-Filip, 2018; Toth et al., 2019). Large and morphologically complex species that contribute substantially to reef structure are disintegrating—the loss of live tissue on these foundational colonies depletes the biological capacity to construct new skeleton while simultaneously creating barren skeleton more vulnerable to erosion. There is an emerging need to assess status and change in reef structure as well as the capacity of the remaining live coral to maintain or restore the skeletal framework crucial to ecosystem integrity. Both assessment needs can be met using demographic (colony-based) surveys that collect data for calculating the 3D surface areas for living and dead portions of colonies. If all species calcified and eroded at the same rate, the relative amount of live and dead surface area would provide a reasonable projection of skeletal growth capacity (GC). A more accurate estimate is made possible, as shown here, using species-specific rates of calcification and calcium carbonate densities. In either case, estimation of GC provides meaningful evidence of the past quality of the reef environment and its future capacity to support coral species and communities.

4.2. Indicator context

The very existence of a coral colony is evidence that environmental conditions were at some time supportive of colony growth and skeletal accretion, indicating a positive GC for at least some duration of the colony’s life. Degradation to a negative GC can only occur when a substantial portion of the skeleton loses living tissue. Environmental conditions that lead to tissue loss are therefore relatively recent or episodic, perhaps resulting from hurricanes or disease (Alvarez-Filip et al., 2011) which could allow for intermittent growth between morbidity events. Once tissue is lost, erosion can occur at any time, even simultaneously with skeletal construction occurring on live portions of the same colony.

A positive GC indicates that under supportive environmental conditions there is a potential for coral tissue to metabolize and accrete CaCO₃, even compensating for erosion losses. A negative GC, on the other hand, indicates that so much live tissue was lost from the colony that remaining tissue is insufficient to replace or offset future erosion losses, even in supportive environmental conditions. Importantly, GC does not predict stony coral accretion or erosion, which are influenced by a variety of future environmental circumstances. Rather, it describes the basic physiological factors (live tissue and vulnerable skeleton) upon which different environmental conditions will act.

4.3. Regional comparisons

Results from the two U.S. Virgin Island regions indicated a capacity for skeletal erosion at STX (−931,852 kg CaCO₃ km⁻² yr⁻¹) and for moderate accretion at STJ (148,185 kg CaCO₃ km⁻² yr⁻¹). The average living tissue (%LT) for the coral communities were not markedly different (70 % vs 75 %), so differences between the two regions are more likely explained by higher %LT at STJ for species with greater surface area such as Orbicella annularis, O. faveolata and O. franksi. Also, populations of Acropora cervicornis, A. palmata and Dendrogyra cylindrus, all larger colonies with positive GC at STJ, were absent from STX. While positive, the result for STJ may still reflect adverse conditions for coral communities; it is only one-tenth the average geologic production of carbonate previously estimated for U.S. Virgin Islands (0.148 vs 1.13 kg CaCO₃ m⁻² yr⁻¹; Hubbard et al. 1990).

4.4. ESA-listed species

All three ESA-listed orbicellid species were negative for GC at STX but only O. annularis was negative at STJ. GC values for O. annularis were the lowest of all species at both regions, and substantially lower at STX, which had roughly half the %LT of O. annularis colonies at STJ. Other listed species A. palmata and D. cylindrus were not reported at STX and there was only one A. cervicornis colony. This contrasted with STJ, where all these species were present and, except for O. annularis, demonstrated positive GC values. Thus, planning for the recovery of these species (ESA §4; Federal Register 2014) would have a greater potential for success at STJ.

4.5. Reef carbonate balance

Several studies have emphasized the importance of carbonate balance as an indicator of reef health (Chave et al., 1972; Scoffin et al., 1980; Perry et al., 2012, 2013; Kuffner et al., 2019; Lange et al., 2020). Estimates of GC are relevant to reef carbonate balance because stony corals are the dominant contributors to overall reef accretion (Chave et al., 1972; Scoffin et al., 1980; Hubbard et al., 1990; Mallela and Perry, 2007; Kuffner et al., 2019). Mallela and Perry (2007), for example, reported that stony corals constituted 97 % of carbonate production on two reefs in Jamaica. In such cases, GC estimates for stony corals could serve as surrogate indicators for reef carbonate balance. This would, however, necessitate including the carbonate contributions of ‘standing dead’ colonies (no live tissue) which were not included in GC calculations because they were sometimes noted but not consistently documented in the surveys. As an example, there were 20 standing dead Acropora palmata noted for STX that, if included in GC calculations, would have further reduced the reef-wide carbonate budget by 3019 kg CaCO₃ yr⁻¹.

4.6. Management value

Calculation of GC can inform a variety of management decisions. Reef structure is vulnerable to acidification from ocean absorption of carbon dioxide (Silverman et al. 2009; Wild et al. 2011; Eyre et al. 2018) which will shift corals toward negative community-level GCs because of changes to calcification and erosion (dissolution) rates. Listing of threatened and endangered species must consider not only the abundance and condition of the species but also the vulnerability to extinction (Federal Register, 2014). A negative species GC, while only a point-in-time calculation, can signal an at-risk species more likely to decline than to recover, such as indicated for O. annularis in these surveys. Also, as noted above, regions with a high positive community-level GC might be more favorable for transplanting and restoring threatened species, even those with a negative species-level GC.

GC can be a valuable estimate of reef resilience, representing the capacity of coral species and communities to recover from the loss of live tissue and deterioration of skeletal structure. However, there are aspects of interpretation to consider in its calculation. Most importantly, GC is calculated for a single point in time and can project, but not predict, future accretion or erosion—multiple environmental variables can alter the trajectory of coral condition. Another consideration relates to the negative GC for O. annularis; this species is known to exhibit barren skeleton along the sides of even healthy colonies, so lower GC values relative to other species do not necessarily mean poor health. This does not alter the fact that only living tissue can generate skeleton and that all barren skeleton, including that which might naturally occur on O. annularis, is vulnerable to erosion.

4.7. Species-specific calcifying and erosion rates

The LT₀, based on the relationship between species-specific calcifying (CR) and erosion rates (ER), provides a simple means to estimate the live tissue for each species needed to sustain the existing skeletal CaCO₃. For these Caribbean species, LT₀ ranged from 31 % (Acropora palmata) to 99 % (Mycetophyllia aliciae and M. lamarckiana). This wide range highlights a substantial variation in CR and ER. For example, very low CR (0.08–0.1 kg CaCO₃ m⁻² yr⁻¹) were reported for Mycetophyllia lamarckiana, M. aliciae, Helioceris cucullata, Agaricia lamarcki, and A. fragilis. These values may seem irregular but could reasonably reflect the slow growth of these species. Regardless, any influence on the community-level GC was marginal since, combined, they represented only ~1 % of the colonies and even less of the colony surface area for both regions. Also, ER for Colpophyllia natans, driven by a low skeletal density, was substantially higher than other species. Yet C. natans were also <1 % of colonies at both regions, so the community-level values were not greatly affected by this potential outlier. Even so, the future utility of GC can be improved with additional species-specific data on skeletal density, CR and ER, and these data would be especially valuable for assessing ESA-listed, proposed and candidate species.

4.8. Reef resilience

The direction and strength of the skeletal GC indicator reflect the point-in-time capacity of stony coral colonies, species, or communities to accrete or erode skeleton, and consequently to provide ecosystem services that rely on reef structural framework, including fish habitat, tourism, shoreline protection, and carbon sequestration. Demonstrated here was the ability of GC calculations to distinguish differences in resilience of coral species and communities, i.e., their capacity to recover and overcome skeletal loss. GC calculations can be used in conjunction with other demographic survey indicators such as diversity, colony density, and live and total surface area to identify healthy, degraded or at-risk coral populations and regions.

Data availability

Data will be posted on EPA’s Environmental Data Gateway