Light and photoacclimatization drive distinct differences between shallow and mesophotic coral communities

Abstract

The ecological distribution of coral species from shallow to mesophotic reefs is dependent on light, which varies drastically among local environments. Current definitions of mesophotic coral ecosystems primarily rely on a 30‐m recreational SCUBA boundary to define the upper limits of the community; however, this boundary does not consider local conditions and physiological adaptions of coral species. Using in situ benthic imagery and chlorophyll fluorescence measurements, we examined species distribution and community similarity, as well as photoacclimatization of two common depth‐generalist species (Montastraea cavernosa and Porites astreoides) across shallow to mesophotic reef zones at Little Cayman Island. Photoquadrat image analysis revealed a significant shift in coral species assemblages between 25 and 35 m, which was accompanied by a 30% drop in available surface light, suggesting light is a key driver of coral community composition. Patterns of photoacclimatization across depths differed significantly between the two coral species, with available surface light and the quantum yield of photochemistry in photosystem II found to be significant determinants of each species' abundance. These results provide valuable baseline data on coral community composition across a broad depth gradient in Little Cayman that can contribute to a growing body of evidence to set an upper boundary of mesophotic reefs based on light availability and photoacclimatization potential of depth‐generalist species.

INTRODUCTION

Shallow coral reef ecosystems are under increasing anthropogenic stress due to changes in temperature, eutrophication, and disease (Hoegh-Guldberg et al., 2017). As these threats continue to evolve, deeper reef ecosystems, termed mesophotic coral ecosystems (MCEs), are of interest as they often do not experience the same increases in temperature and associated bleaching as their shallow water counterparts (e.g., Bongaerts et al., 2010; Gould et al., 2021; Hinderstein et al., 2010). MCEs are characterized by large changes in downwelling solar irradiance, which promotes communities of low-light-adapted photosymbiotic scleractinian corals, macroalgae, and sponges (Hinderstein et al., 2010; Lesser et al., 2009). Historically, shallow reefs have experienced declines, whereas MCEs have had a relatively stable coral population over time (Bak et al., 2005). Thus, MCEs are often suggested to increase the resilience of coral reefs overall by acting as a refuge to shallow reefs (Bongaerts et al., 2010; Glynn, 1996; Goodbody-Gringley et al., 2021; Sturm et al., 2021). The idea of MCEs as a refuge, however, has been questioned due to increased frequencies of disturbances and anthropogenic impacts at mesophotic depth (Rocha et al., 2018). Yet, despite the critical need to protect coral reefs under projected global changes and the potential role of deeper reefs as a refuge, MCEs remain understudied, primarily due to their inaccessibility. Therefore, a critical gap exists in our understanding of the fundamental ecology driving mesophotic coral communities, impeding the ability to protect the resilience of these important reef systems.

Recently, several studies have addressed the need to determine ecological factors (i.e., light) that drive changes in community composition and improve the definition of MCEs (Edmunds et al., 2018; Lesser et al., 2018). The widely accepted 30–40 m upper boundary between shallow and mesophotic reefs is based on recreational SCUBA limits, but there may be some evidence that community shifts also occur at or around this depth (Pyle & Copus, 2019). The lower boundary of MCEs is generally characterized as 1% of available surface light, as this is the minimum light necessary for photosynthesis (Lesser et al., 2009). In the Gulf of Eilat/Aqaba, Red Sea, community shifts were determined at 40 and 70 m (Tamir et al., 2019) compared with those at 15 and 40 m in Utila, Honduras (Laverick et al., 2017), suggesting that a transition zone from shallow to mesophotic reefs may occur at 30–40 m depth across geographic regions where hermatypic coral reefs persist. Additionally, light explained more variation in coral community structure than depth for both shallow and mesophotic genera in the Gulf of Eilat/Aqaba (Laverick et al., 2020; Tamir et al., 2019). Light, among other factors, is therefore thought to explain community shifts and morphological and physiological differences documented for corals across shallow to mesophotic depth (e.g., Goodbody-Gringley & Waletich, 2018).

Several species that exist along a broad depth distribution exhibit patterns of adaptation and/or acclimatization to deeper environmental conditions, such as changes in skeletal morphology and physiology, including nutrient acquisition, calcification, and photosynthesis (Bruno & Edmunds, 1997; Einbinder et al., 2009; Goodbody-Gringley et al., 2015; Goodbody-Gringley & Waletich, 2018; Gould et al., 2021; Malik et al., 2021; Martinez et al., 2020; Mass et al., 2007). In the Bahamas, the algal symbionts in Montastraea cavernosa were found to exhibit similar increases in photosynthetic efficiency with depth measured by an increase in the quantum yield of photosystem II (PSII) fluorescence (Lesser et al., 2010). In the Gulf of Eilat, the algal symbionts in the depth-generalist coral Stylophora pistillata exhibited changes in the harvesting and utilization of light in response to depth, specifically with increased efficiency of photosynthesis (Martinez et al., 2020; Mass et al., 2007). Under low-light conditions, this coral species may increase light absorption through increases in algal density (Titlyanov et al., 2001) or algal cell pigments (Falkowski & Dubinsky, 1981). One technique used to examine light-induced changes is chlorophyll variable fluorescence, which provides a wide range of measurements for fluorescent and photosynthetic parameters of an organism (Falkowski et al., 2004). Specifically, the quantum yield of photochemistry in PSII (F_v/F_m) captures the availability of open PSII reaction centers and the efficiency of the reaction center (Genty et al., 1989). As light is a limiting factor and determinant of coral distribution, changes in photosynthetic efficiency, or potential photoacclimatization, are critical to understand species distribution.

The aims of this study were to (1) characterize the coral community and assess photoacclimatization in target species across depths in Little Cayman, a remote Caribbean Island; (2) identify the transition depth from shallow to mesophotic reef ecosystems; and (3) determine how light availability and photophysiology affect species distributions. Specifically, we used benthic quadrat images, in situ chlorophyll fluorometry, and photosynthetically active radiation (PAR) to understand the environmental and physical characteristics driving differences between shallow and mesophotic reef coral communities. In addition, we examined photobiology metrics in two common Caribbean species with varying depth distributions, M. cavernosa and Porites astreoides, to identify species-specific differences in acclimatization to low-light conditions associated with depth. Taken together, these data highlight the influence of light in determining patterns of community composition across depth, identify a shallow–mesophotic transition depth that aligns with other regions, and determine the attributes of photobiology that drive differences in the abundance of two key coral species.

METHODS Study sites

Little Cayman is the smallest of the Cayman Islands at 17 × 2 km and is located 120 km northeast of Grand Cayman in the Caribbean Sea. The island has minimal anthropogenic impacts with less than 200 permanent residents and has enforced two no-take marine parks since the mid-1980s (Manfrino et al., 2013). This study focused on three sites in Little Cayman: Dynamite Drop (19°65.534′ N, 80°09.583′ W), Martha's Finyard (19°66.48′ N, 80°11.122′ W), and Crystal Palace (19°70.283′ N, 80°05.094′ W) (Figure 1a). In Little Cayman, the nearshore environment is primarily sandy bottom, hardpan, or seagrass beds that extend from the coast to roughly 10–15 m depths where the reef forms in a spur and groove pattern and extends at relatively constant depth to the reef wall, which is a vertical sheer drop to depths exceeding 1 km. Sampling depths at the selected sites were therefore dictated by the topography of the reef slope where the shallowest depth sampled represented the highest/shallowest point on the reef. The deeper end of the sampling, however, was limited by technical diving capabilities of the team.

FIGURE 1. (a) Map of survey sites in Little Cayman. Benthic photoquadrats were collected at all three sites, while chlorophyll fluorescence measurements were collected at Dynamite Drop only. (b) Photograph of diver with quadrat frame and camera. (c) Image of diver collecting fluorescence induction and relaxation (FIRe) readings from a Porites astreoides colony at 35 m depth on a closed-circuit rebreather.

Benthic community characterization

Photoquadrats were collected by divers using closed-circuit rebreathers (Hollis Prism 2, Hollis Rebreathers, Salt Lake City, UT, USA) from 7 April to 12 April 2021. A 10-m line was haphazardly placed on the reef at the three sites over four depths (15, 25, 35, and 45 m), maintaining constant depth for the full 10 m length. Photographs were taken approximately every meter on either side of the line using a 1-m² polyvinyl chloride pipe quadrat (n = 53 for 15, 25, and 35 m; and n = 58 for 45 m) with a Sony A7R III camera with Sea & Sea YS-D2 strobes mounted above. The camera was mounted to a permanent quadrat frame ensuring a constant distance of 1 m above the center of the quadrat, which was placed perpendicular to the reef resulting in the same angle of each photograph (Figure 1b). In each quadrat, all scleractinian coral colonies were identified to the species level and the surface area was measured using the freehand tool in the ImageJ software (https://imagej.nih.gov/ij/). Additionally, coral abundance per quadrat was calculated as the percent contribution of the total surface area of each species to overall hard coral cover. The percent cover of functional groups categorized as crustose coralline algae (CCA), turf algae, macroalgae, soft corals, sponges, sand/rubbles, or others was visually estimated within each quadrat, while the total measured surface area of all scleractinian corals in each quadrat was used for the category of hard corals.

Light physiology and environment

On 24 June 2021, in situ fluorometry and PAR (400–700 nm) data were collected at Dynamite Drop (19°65.534′ N, 80°09.583′ W). Two candidate species, M. cavernosa and P. astreoides, were selected based on differing depth distributions, life-history strategies, and potential photoacclimatization potential, and analyzed in situ for active fluorescence (Laverick et al., 2017). M. cavernosa is a gonochoric broadcast spawning depth generalist, with a distribution ranging from 0 to 5 m down to 80 to 85 m based on local conditions (Fricke & Meischner, 1985; Laverick et al., 2017), that exhibits significant plasticity in colony morphology, rates of respiration, and primary productivity (Goodbody-Gringley et al., 2015; Goodbody-Gringley & Waletich, 2018; Lesser et al., 2010). P. astreoides, on the contrary, is a hermaphroditic brooding species with a distribution ranging from 0 to 5 m down to 40 to 50 m (Fricke & Meischner, 1985; Laverick et al., 2017) that has shown high photograph repair capacity (Hennige et al., 2011). Historically, P. astreoides was considered a winner on coral reefs due to “weedy” life-history characteristics including high rates of recruitment (Darling et al., 2012; Green et al., 2008), yet recent studies suggest the species may be in decline (Edmunds et al., 2021).

Active fluorescence was measured using a diving-fluorescence induction and relaxation (FIRe) fluorometer (Figure 1c). The FIRe technique, as described by Gorbunov and Falkowski (2005), is the successor of the fast repetition rate fluorometer, which was first used on corals by Gorbunov et al. (2001) and Lesser and Gorbunov (2001). Phase 1 of the FIRe technique uses a strong short pulse of 100-ms duration (single turnover flash) to cumulatively saturate PSII and measures the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$), the functional absorption cross section of PSII (σ_PSII′; A²), and the connectivity parameter (p) that determines the probability of excitation energy transfer between individual photosynthetic units. Phase 2 applies weak modulated light to record the relaxation kinetics of fluorescence yield (500 ms), which provides the maximum photosynthetic rate (P_max; electron s⁻¹ PSII⁻¹). As such, areas of a coral that are pre-acclimated to low light (e.g., shaded) will provide more accurate measurements for physiology in dark/low-light state (e.g., F_v/F_m), while high-light acclimated (e.g., exposed to sunlight) areas of the colony will correctly reflect high-light-adapted parameters (e.g., maximum photosynthetic turnover rates). To account for intra-colony differences in light exposure, five readings were collected per colony (one reading on each cardinal direction and one reading on the top, center of the coral) and averaged to reflect the photophysiology of the entire colony in the analysis. Ten colonies of each species were targeted at each depth (15, 25, 35, and 45 m); however, only eight colonies of M. cavernosa were measured at 45 m (total n = 38) due to the time constraints of technical diving, and no P. astreoides colonies were measured at 45 m (total n = 30) due to the rarity of the species at this depth.

FIRe readings were collected 2 h before and after solar noon on the same day to keep the light environment as similar as possible. On the day of sampling, skies were overcast with times of low-light and sunnier conditions in the afternoon. Corresponding PAR measurements were taken while divers were collecting FIRe readings and within 20 min of solar noon to characterize the highest light availability. PAR was recorded 10 times per second for 2 min at the surface (~1 m) and each target depth (15, 25, 35, and 45 m; LI-COR LI-1500). The average percentage of surface PAR was calculated for each depth and time of day.

Statistical analysis

Coral abundance and community structure were compared among depths and sites with the nonparametric Kruskal–Wallis test as the residuals were not normally distributed, even after transformation (Kruskal & Wallis, 1952). There was no effect of site (Appendix S1: Table S1), and thus, further analyses compared coral abundance and community structure across depths, with each site/depth combination acting as a single replicate (n = 3 per depth). We assessed coral community using nonmetric multidimensional scaling and permutation-based multivariate analysis of variance (PERMANOVA; Anderson, 2001; Anderson & Walsh, 2013), using Bray–Curtis distances. The statistical assumption of homogenous dispersion was met for PERMANOVA. To assess group-level differences between shallow (15 and 25 m) and mesophotic (35 and 45 m) communities, we calculated Bonferroni-corrected p values (Martinez Arbizu, 2019). The data from quadrat functional groups and FIRe residuals were not normally distributed, even after transformation. We proceeded with nonparametric analysis using the Kruskal–Wallis test and post hoc pairwise comparisons using the Dunn's test with the Bonferroni correction (Dunn, 1964). A pairwise correlation matrix was used to compare biological and physical parameters including functional group cover, hard coral species richness and diversity, depth, and PAR. This analysis was used for data from Dynamite Drop as that was the only location with light data and species diversity for each quadrat was calculated using the Shannon–Weaver index. Two pairwise correlation matrices were used to compare physical and photosynthesis parameters, including PAR, the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$), functional absorption cross section of PSII (σ_PSII′), maximum photosynthetic rate (P_max), and the connectivity parameter (p), as well as coral abundance of M. cavernosa and P. astreoides. Matrix comparisons were examined with a generalized linear model (GLM) with the Poisson distribution, which yielded a normal distribution based on the visual analysis of normal quantile plots. Model results were visualized with principal component analysis using the maximum-likelihood factor analysis. All statistical analyses and data visualization were done in the R version 3.6.2 (R Core Team, 2019) with RStudio 1.2.1335 (RStudio Team, 2021).

RESULTS Benthic community characterization

A total of 37 coral species were identified (Appendix S1: Table S2), and coral abundance of the top eight contributing species shows clear trends in species composition changes across depths (Figure 2a). As depth increases, there is an increase in Siderastrea siderea, Stephanocoenia intersepta, and Agaricia spp., with a shift from dominantly Agaricia agaricites to Agaricia lamarcki. Orbicella spp. decrease with depth, and Porites spp., with a clear depth limit of P. astreoides at 35 m. M. cavernosa is relatively uniform across depths with an increase in abundance at 25 m. A PERMANOVA of depth shows a significant difference in coral community by depth (F_1,10 = 6.497, p < 0.001; Appendix S1: Table S3a). A pairwise comparison shows a significant difference between shallow (15 and 25 m) and mesophotic (35 and 45 m) coral communities based on species abundance (p < 0.01; Appendix S1: Table S3b), which is also seen visually (Figure 2b).

FIGURE 2. (a) Coral abundance, species contribution to coral cover, of the top eight species for each depth group and (b) nonmetric multidimensional scaling (NMDS) of coral communities across depths based on coral abundance (stress value = 0.05). For full species names, refer to Appendix S1: Table S2.

Functional groups varied significantly by depth with general increases in CCA and sponges and decreases in hard corals and macroalgae (Figure 3). CCA was significantly lower at 15 m than at all other depths (p < 0.05, Dunn's test; Figure 3a; Appendix S1: Table S4). Hard coral cover significantly decreased as depth increased (H_1,3 = 86.391, p < 0.001, Kruskal–Wallis test; Figure 3b; Appendix S1: Table S4). Specifically, hard coral cover at 25 and 35 m was comparable, but percent cover was significantly higher at 15 m and lower at 45 m. Macroalgal cover was generally uniform, except for a significant decrease at 45 m (p < 0.05, Dunn's test; Figure 3c; Appendix S1: Table S4). Lastly, sponge cover at 25 m and deeper was significantly higher than the percent cover at 15 m (p < 0.05, Dunn's test; Figure 3d; Appendix S1: Table S4).

FIGURE 3. Functional group percent cover across depths for major contributing groups: (a) crustose coralline algae (CCA), (b) hard corals (c), macroalgae, and (d) sponges. Letters indicate Dunn's test significant difference between depth groups.

Light physiology and environment

On the day of FIRe reading, PAR ranged from 34.04 ± 2.3 to 311.12 ± 22.46 μmol m⁻² s⁻¹, with the lowest reading in the morning at 45 m and the highest during solar noon at the surface (Figure 4; Appendix S1: Table S5). Available surface light at 15, 25, 35, and 45 m ranged from 37% to 83%, 25% to 60%, 18% to 31%, and 11% to 20%, respectively (Appendix S1: Table S5). Higher PAR levels were recorded in the afternoon when skies started to clear.

FIGURE 4. Photosynthetically active radiation (PAR; μmol m−2 s−1) during morning, solar noon, and afternoon on the day of photosynthesis metrics. Measurements were taken at four depths (15, 25, 35, and 45 m), and 95% confidence intervals are indicated with shading.

Significant positive correlations were found between PAR and hard coral cover (r = 0.7, p < 0.001; Figure 5), hard coral richness (r = 0.6, p < 0.001; Figure 5), and hard coral diversity (r = 0.4, p < 0.001; Figure 5). Significant negative correlations were found between depth and hard coral cover (r = 0.7, p < 0.001; Figure 5), hard coral richness (r = 0.6, p < 0.001; Figure 5), and hard coral diversity (r = 0.4, p < 0.001; Figure 5).

FIGURE 5. Correlation matrix for comparisons between depth (in meters), photosynthetically active radiation (PAR; μmol m−2 s−1), sponge cover, macroalgal cover, hard coral cover, hard coral species richness, and hard coral species diversity (Shannon–Weaver index). Positive correlations are represented with warm colors (red), and negative correlations are represented with cool colors (blue). The strength of correlations (r) is printed for significant correlations and indicated by the intensity of the colors, where lighter colors represent a weaker correlation, and dark colors represent strong correlations. CCA, crustose coralline algae.

All metrics of chlorophyll fluorescence differed significantly between M. cavernosa and P. astreoides (p < 0.001, Kruskal–Wallis test; Figure 6; Appendix S1: Table S6). Within M. cavernosa measurements, the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$) was significantly lower at 15 m than at all other depths (p < 0.05, Dunn's test; Figure 6a; Appendix S1: Table S7). The functional absorption cross section of PSII (σ_PSII′) was significantly higher in M. cavernosa colonies at 45 m than at all other depths. Finally, the maximum photosynthetic rate (P_max) significantly decreased with depth (p < 0.05, Dunn's test; Figure 6c; Appendix S1: Table S7), and there was no significant difference in the connectivity parameter (p) in M. cavernosa. Within P. astreoides measurements, the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$) was significantly lower at 15 m than at 25 and 35 m (p < 0.05, Dunn's test; Figure 6a; Appendix S1: Table S8). The maximum photosynthetic rate (P_max) was significantly lower at 35 m than at 25 m, and the connectivity parameter was significantly higher at 15 m than at 25 m (p < 0.05, Dunn's test; Figure 6c; Appendix S1: Table S8). There was no significant difference in the functional absorption cross section of PSII (σ_PSII′) in P. astreoides across depths.

FIGURE 6. Kinetic-based chlorophyll fluorescence measurements of (a) quantum yield of photochemistry in photosystem II (PSII; (Fv′/Fm′$$ {F}_v^{\prime }/{F}_m^{\prime } $$), (b) functional absorption cross section of PSII (σPSII′; A2), (c) maximum photosynthetic rate (Pmax; electron s−1 PSII−1), and (d) connectivity parameter (p) using the diving-fluorescence induction and relaxation (FIRe) fluorometer. Letters indicate Dunn's test significant difference between depth groups for each species. Abbreviations: MCAV, Montastraea cavernosa; PAST, Porites astreoides.

GLM with the Poisson distribution for M. cavernosa shows a significant effect of the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$), the PAR percent at depth, and the functional absorption cross section of PSII (σ_PSII′) on coral abundance (Figure 7a, Table 1a). Results of GLM for P. astreoides show a significant effect of the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$) and the PAR percent at depth on coral abundance (Figure 7b, Table 1b).

FIGURE 7. Factanal principal components analysis (PCA) plot of relative effect of the quantum yield of photochemistry in photosystem II (PSII; Fv′/Fm′$$ {F}_v^{\prime }/{F}_m^{\prime } $$), functional absorption cross section of PSII (σPSII′; A2), maximum photosynthetic rate (Pmax; electron s−1 PSII−1), connectivity parameter (p), and photosynthetically active radiation (PAR) on coral abundance (percent contribution to coral cover) for (a) Montastraea cavernosa and (b) Porites astreoides. Variables with a significant effect are indicated with an asterisk (*p [less than] 0.05; ***p [less than] 0.001) as determined by generalized linear model with the Poisson distribution.

TABLE 1 Summary of generalized linear model results for factors contributing to coral abundance of (a) Montastraea cavernosa and (b) Porites astreoides.

Source	Estimate	SE	T	p
(a) Coral abundance of M. cavernosa
Intercept	−0.923	0.210	−4.392	<0.001
Quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$)	1.424	0.296	4.817	<0.001
Functional absorption cross section of PSII (σPSII′)	0.001	0.000	4.456	<0.001
Maximum photosynthetic rate (Pmax)	−0.001	0.000	−1.928	0.063
PAR percent	0.346	0.028	12.350	<0.001
Connectivity parameter (p)	−0.147	0.188	−0.782	0.440
(b) Coral abundance of P. astreoides
Intercept	0.002	0.066	0.037	0.971
Quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$)	0.244	0.104	2.348	0.027
Functional absorption cross section of PSII (σPSII′)	0.000	0.000	−0.754	0.458
Maximum photosynthetic rate (Pmax)	0.000	0.000	−1.283	0.212
PAR percent	−0.107	0.013	−8.360	<0.001
Connectivity parameter (p)	0.011	0.068	0.165	0.871

Note: Bold indicates variables with a statistically significant effect. Abbreviations: PAR, photosynthetically active radiation; PSII, photosystem II. DISCUSSION

This study examined coral abundance and factors driving the ecological distribution of species from shallow to mesophotic depths in Little Cayman, Cayman Islands, documenting clear shifts in species distribution between depths with a significant change in overall community structure between 25 and 35 m. Changes in coral species richness and diversity are explained by the availability of light at depth. Further analyses of photophysiological parameters indicate significant differences in the utilization of light by M. cavernosa and P. astreoides. Additionally, we show that light availability and photophysiology help explain the distribution of M. cavernosa and P. astreoides across a depth gradient.

Coral abundance varied significantly across depths with notable changes in the distribution of species abundance. While percent cover was generally lower than that in other studies in Little Cayman (Manfrino et al., 2013), the methodology used here examined two-dimensional area as opposed to three-dimensional area. Specifically, with increasing depth, we found a higher contribution of plating corals, particularly A. lamarcki, which is known to exhibit this morphology between 20 and 55 m (Figure 2a; Laverick et al., 2019). Additionally, we saw an increase in abundance of depth-generalist species such as S. siderea, S. intersepta, and M. decactis at 45 m, which is consistent with other studies from the Caribbean (Appeldoorn et al., 2019; Kahng et al., 2010). This pattern of coral community composition reveals a break between reefs shallower than 25 m and those deeper than 35 m that aligns with the current definition of MCEs based primarily on recreational SCUBA limits (Pyle & Copus, 2019). While this definition is not based on local ecology, there is support that, in general, community shifts occur around 30 m (Pyle & Copus, 2019). In this study, the break in community composition between 25 and 35 m is consistent with changes in available light found at these depths. In the afternoon, when the water was the clearest and cloud cover was low, the percent of available surface PAR decreased by ~30% between 25 and 35 m, representing the largest difference in available light between any of the depth zones examined (Figure 4; Appendix S1: Table S5). Additionally, there were significant correlations between light and coral species richness and diversity, where decreases in light, and increases in depth, are accompanied by decreases in species richness and diversity (Figure 5). Photophysiological results in this study also align with previously documented evidence that changes in light availability result in shifts to community composition across depths (Laverick et al., 2017, 2020; Tamir et al., 2019) and that light explains changes in depth distribution and morphology of coral species more than depth alone (Tamir et al., 2019).

M. cavernosa and P. astreoides exhibited significantly different photophysiological metrics, which also varied significantly by depth within each species (Figure 6; Appendix S1: Tables S6–S8). Both species showed a significant decline in the maximum photosynthetic rate (P_max) over depths; however, the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$) differed significantly between species, where P. astreoides exhibited higher $$$ {F}_v^{\prime }/{F}_m^{\prime } $$$ values than M. cavernosa across all depths. This finding is consistent with other studies of $$$ {F}_v^{\prime }/{F}_m^{\prime } $$$ that found P. astreoides consistently had significantly higher F_v/F_m than all other species (Warner et al., 2006). At shallow depths in Belize, P. astreoides was also found to exhibit significantly lower excitation pressure over PSII than other species, indicating less reaction centers closing as a photoinhibition mechanism and suggesting that higher $$$ {F}_v^{\prime }/{F}_m^{\prime } $$$ values for P. astreoides may be a result of adaptive photoinhibition (Warner et al., 2006). Such a response to light would be advantageous under high irradiance but would not benefit P. astreoides in lower light environments, such as in MCEs, which may explain the maximum depth distribution of 35 m found in Little Cayman.

While P. astreoides was not found below 35 m in Little Cayman, M. cavernosa was documented at 45 m and likely persists much deeper as it was observed from 3 to 100 m in other parts of the Caribbean (Lesser et al., 2010; Reed, 1985). The ecological distribution of a species is a function of its strategy for obtaining energy within a given environment and its capacity to utilize available light through photosynthesis (Falkowski et al., 1990). Thus, one possible explanation for the persistence of the species at great depth despite the significant decline in the maximum photosynthetic rate (P_max) may be a potential switch from autotrophy to heterotrophy. While this study did not measure relative dependance on autotrophy versus heterotrophy, M. cavernosa is documented to modulate mechanisms for obtaining necessary energy based on environmental conditions (Lesser et al., 2010). For example, previous studies on M. cavernosa indicate a distinct change in energy acquisition between 46 and 61 m, where shallower corals relied predominately on autotrophy but transitioned to heavier reliance on heterotrophy with increased depth (Lesser et al., 2010). While examining metrics of heterotrophy were outside the scope of this study, we found a decline in the maximum rate of photosynthesis (P_max) for M. cavernosa with increasing depth, which suggests a reduced capacity to obtain all of its energy requirements from autotrophy alone and thus implies a potential transition to heterotrophy at depth may enable the species to persist across a broad depth gradient. Alternatively, previous studies have also suggested that efficiency in photochemical conversion of available light may increase with depth as low light reduces nonphotochemical quenching (Lesser & Gorbunov, 2001). The decline in P_max with increased depth for M. cavernosa was found to be coupled with less deviation, as well as increases in $$$ {F}_v^{\prime }/{F}_m^{\prime } $$$ and the function absorption cross section of PSII (σ_PSII′), indicating increased efficiency in photochemical conversion that may also enable the species to persist at depth. These results would be further supported by examining the distribution of Symbiodiniaceae species associated with M. cavernosa and P. astreoides across depths; however, permits for sample collection could not be obtained due to recent outbreaks of stony coral tissue loss disease on Grand Cayman.

Although photophysiological metrics differed between species, both species expressed similar responses to light. In particular, our model comparisons found that for both M. cavernosa and P. astreoides, available surface light (PAR) was the main factor influencing their abundance across depth (Figure 7, Table 1). This is supported by the strong correlation between light and coral species richness and diversity (Figure 5), and is in line with the well-established understanding that light is a major driver of coral community composition across depth (Falkowski et al., 1990). In addition to light, however, the quantum yield of photochemistry in PSII ($$$ {F}_v^{\prime }/{F}_m^{\prime } $$$) was also found to be a significant factor across depth (Figure 7; Appendix S1: Tables S7 and S8). Thus, the results of this study suggest that photoacclimatization potential likely explains the differences in depth distribution of M. cavernosa compared with those of P. astreoides.

Recent work shows strong evidence that mesophotic coral species, including M. cavernosa and P. astreoides, play an influential role in maintaining populations of shallow reefs (Goodbody-Gringley et al., 2021; Sturm et al., 2021). Understanding the fundamental drivers of ecology in mesophotic corals, as well as the physiological light parameters of individual species, increases our ability to predict their distribution and identify potential areas that may serve as refuges, thereby increasing the resilience of global coral reef ecosystems. Such information is imperative as coral reefs continue to decline and are faced with increasing frequency and severity of disturbances and anthropogenic stressors (e.g., Bongaerts et al., 2010; Glynn, 1996; Hoegh-Guldberg et al., 2017).

AUTHOR CONTRIBUTIONS

Gaby E. Carpenter, Gretchen Goodbody-Gringley, and Tali Mass designed the experiment. Gaby E. Carpenter, Alex D. Chequer, Sabrina Weber, and Gretchen Goodbody-Gringley performed the underwater data collection. Gaby E. Carpenter and Gretchen Goodbody-Gringley wrote the manuscript. All authors improved and revised the manuscript, and gave approval for the manuscript.

ACKNOWLEDGMENTS

The authors thank M. Gorbunov for extensive advice and training on the use of the FIRe and the analysis of the resulting data. This work was funded by the United States National Science Foundation (NSF) and the United States—Israel Binational Science Foundation (BSF) (NSF number 1937770 to Gretchen Goodbody-Gringley; BSF number 2019653 to Tali Mass).

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

Data and code (Carpenter et al., 2022) are available from Zenodo: https://doi.org/10.5281/zenodo.6518135.