The evolutionary success of reef-building corals is often attributed to photosymbiosis, a mutualistic relationship scleractinian corals developed with zooxanthellae; however, because zooxanthellae are not fossilized, it is difficult (and contentious) to determine whether ancient corals harbored symbionts. In this study, we analyze the δ15N of skeletal organic matrix in a suite of modern and fossil scleractinian corals (zooxanthellate- and azooxanthellate-like) with varying levels of diagenetic alteration. Significantly, we report the first analyses that distinguish shallow-water zooxanthellate and deep-water azooxanthellate fossil corals. Early Miocene (18–20 Ma) corals exhibit the same nitrogen isotopic ratio offset identified in modern corals. These results suggest that the coral organic matrix δ15N proxy can successfully be used to detect photosymbiosis in the fossil record. This proxy will significantly improve our ability to effectively define the evolutionary relationship between photosymbiosis and reef-building through space and time. For example, Late Triassic corals have symbiotic values, which tie photosymbiosis to major coral reef expansion. Furthermore, the early Miocene corals from Indonesia have low δ15N values relative to modern corals, implying that the west Pacific was a nutrient-depleted environment and that oligotrophy may have facilitated the diversification of the reef builders in the Coral Triangle.
Through photosymbiosis, dinoflagellates, called zooxanthellae, live within the tissue of many modern scleractinian corals. Zooxanthellae photosynthesize within the coral tissue providing corals with most of their energy, while the coral hosts in turn live in shallow, clear waters where zooxanthellae have optimal exposure to sunlight for photosynthesis1,2,3. Although both zooxanthellate and azooxanthellate scleractinians exist and can build reefs, a symbiotic relationship with zooxanthellae is an obvious advantage. Zooxanthellae provide corals with photosymbiotic byproducts (e.g., oxygen and glucose), which allow the corals to calcify at an expedited rate, making zooxanthellate corals more efficient reef-builders than azooxanthellate corals in shallow, oligotrophic conditions2, 4, 5. Photosymbiosis is, therefore, considered a key aspect of modern corals and thought to be the main driver of the Triassic expansion and diversification of shallow-water scleractinian corals1, 6,7,8,9,10. Nevertheless, since zooxanthellae live within the soft tissue of corals, they are not directly preserved in the fossil record, making it difficult to determine whether fossil corals had symbionts, when photosymbiosis originated, or how this relationship evolved through time.
Morphological characteristics of coral skeletons (corallite size, growth form, and level of corallite integration11) are frequently used to infer photosymbiosis in the fossil record and are often used in Phanerozoic compilations of coral-zooxanthellae symbiosis8, 11,12,13. For example, zooxanthellate corals tend to be colonial and have small, highly integrated corallites whereas azooxanthellate corals tend to have solitary growth forms and larger polyps11. Additionally, some coral colony morphologies are influenced by the presence of zooxanthellae and are used in the fossil record to demonstrate dependence on sunlight14. A delicate plate-like or platy coral morphology, for example, is common in quiet, deep, poorly-lit waters where corals need to maximize their exposure to sunlight for zooxanthellae photosynthesis14. Morphological characteristics, however, are not definitive and can be misleading when trying to identify whether extinct species, rather than coral assemblages, were symbiotic13, 15. More recently, macroscopic and microscopic growth bands have been used to infer photosymbiosis in corals9, 10. Macroscopic density bands, which are inferred to be annual growth bands, have been used to estimate fossil coral growth rates as a proxy for ancient photosymbiosis because modern zooxanthellate corals can grow faster than their azooxanthellate counterparts9. Alternatively, the regularity of microscopic skeletal growth bands in the fibrous aragonitic fibrous bundles of coral skeletons has also been proposed as a signature of photosymbiosis10, 16, 17. Although both growth band proxies have promise, they can be significantly altered by diagenesis and so can only be used in the most pristine samples10, 15, 17, 18.
Two geochemical proxies for ancient photosymbiosis were proposed to resolve the photosymbiotic assignment of contentious fossil corals: the stable oxygen (δ18 O) and carbon (δ13C) isotope ratios of fossil coral skeletons7 and the stable nitrogen isotope ratio (δ15N) of the coral skeleton-bound organic matrix19. The δ18O and δ 13C ratios of fossil coral skeletons are, again, greatly influenced by diagenesis7, 10, 15, 17, 18. Conversion of aragonite to calcite is common in carbonate systems and can profoundly impact skeletal δ18O15, 18, 20 making the proxy applicable only to a few, well-preserved fossil corals. The coral skeleton-bound organic matrix, on the other hand, may be protected from diagenesis by the skeletal structure of the fossil and should be (at least partially) preserved along with the fossil itself21,22,23,24,25 making the δ15N proxy applicable to a wider range of fossil corals.
Organic matrix δ15N values differ significantly in modern zooxanthellate corals (3–13‰19, 25; all δ15N values here are reported with respect to AIR) relative to modern azooxanthellate corals (9–19‰10, 19, 26). Differences in organic matrix δ15N between zooxanthellate and azooxanthellate corals were originally attributed to the nutritional lifestyle of the coral polyps and the source of nitrogen to the organic matrix of the coral19 but recent studies show that the differences are also controlled by the δ15N of the nitrogen sources to the corals at the site of growth and nutrient cycling in deep versus shallow waters25,26,27. For example, corals living below the euphotic zone feed on suspended particulate organic matter that has relatively high δ15N due to partial decomposition25; as a result, their skeletal-bound organic matrix has higher δ15N values than zooxanthellate corals living in oligotrophic settings. In the modern ocean this leads to a ~7‰ δ15N offset between zooxanthellate and azooxanthellate corals10, 25,26,27.
Since the first application of the δ15N proxy, the only fossil corals tested are well-preserved (aragonitic) Triassic corals from Alakir Çay Valley in Turkey, which produce δ15N values ranging from 2 to 7‰10, 19, 24. Nevertheless, to date, no fossil coral has displayed a high δ15N value (i.e. >9‰), equivalent to modern azooxanthellate corals, which raises a fundamental question about the reliability of the proxy in deep time. Additionally, even though diagenesis should not affect skeleton-bound ratios in pristine corals, no tests have been conducted on the δ15N of fossil corals with varying levels of recrystallization.
In this study, a suite of modern zooxanthellate and azooxanthellate corals as well as fossil zooxanthellate- and azooxanthellate-like corals11 with varying coral morphologies and levels of diagenetic overprint are analyzed for skeletal-bound organic matrix δ15N (Table 1). The key objectives are: (1) to evaluate the applicability of the δ15N proxy19 for ancient photosymbiosis to fossil corals with different morphologies and from varying localities and ages. Zooxanthellate- and azooxanthellate-like fossil samples are expected to plot in different δ15N ranges, as they do for modern corals. (2) Determine the effect of diagenetic alteration on the δ15N of fossil corals. Samples are screened for diagenesis through macroscopic observations, thin section petrography, micro-Raman spectroscopy, and scanning electron microscopy (SEM); one recrystallized sample was run in conjunction with aragonitic ones. When organic matter was preserved within the original, aragonitic skeleton of the coral, it was assumed to be primary. (3) Compare the original δ15N proxy dialysis/combustion method19 to the recently developed persulfate/denitrifier method25. When enough mass was available (i.e. 15–50 g), a duplicate sample was also processed using the dialysis/combustion method to test for procedural precision.
Defining a successful proxy for ancient photosymbiosis is important in invertebrate paleontology, carbonate sedimentology, and coral biology as it will provide critical data regarding the evolutionary link between scleractinian corals and symbionts, as well as, the association of coral photosymbiosis and reef-building.
All of the modern (zooxanthellate and azooxanthellate), Holocene, and Miocene samples, as well as one Triassic sample (Distichomeandra sp., sample 2 [herein noted as Distichomeandra sp. (2)]) are composed of aragonite. These corals display primary aragonite bundles in scanning electron microscopy (SEM) secondary electron (SE) images, aragonite fans in thin sections, and Raman spectra consistent with aragonite28 (Fig. S15), thus we conclude that they have retained their original aragonitic skeletal structure (Fig. 1). Triassic samples Gablonzeria sp. and Distichomeandra sp. sample 1 [herein noted as Distichomeandra sp. (1)] (Table 1) retained their coralline structures, as visible in thin section, but displayed varying levels of diagenesis; both blocky calcite and aragonite bundles were visible on specimens in SEM SE images and thin sections. These samples were, therefore, subjected to a more detailed sample preparation to remove the recrystallized regions prior to analyses (see Methods). The Oligocene coral (Antiguastrea lucasiana; Table 1) is completely recrystallized to calcite and is used to evaluate the impact of diagenesis on coral δ15N (Fig. 1e). For detailed diagenetic screening summaries see supplemental material.
In general, the thin section petrography and Raman spectroscopy of modern, Miocene, and Triassic corals showed that organic material is disseminated throughout the aragonitic coral skeleton or concentrated in the centers of calcification with no obvious contamination. Only one specimen (Oulophyllia sp. AZ5949) had some organic-rich microborings that contained organics (see Fig. S15); so far as it was possible, all borings were removed from coral specimens during sample preparation.
Organic Matrix Nitrogen Isotope Ratios
The skeletal organic matrix δ15N values range from 1.90‰ to 6.38‰ (AIR) for modern zooxanthellate and fossil zooxanthellate-like corals (Fig. 2). Standard deviations range from ± 0.10‰ to ± 1.43‰ (Table 2); sample Distichomeandra (2)-TR2 displays an unusually high standard deviation of ± 1.43‰, but the standard deviations of all fossil samples analyzed using the dialysis/combustion method are lower than the Triassic sample ( ± 2.72‰) analyzed in the original study19. The organic matrix recovered from modern azooxanthellate sample D. dianthus and Miocene samples were too scarce for replicates (Table 2).
Several samples were analyzed using two methods: the dialysis/combustion method18 and the persulfate/denitrifier method25 for comparison (Table 2). Samples analyzed using the newly developed persulfate/denitrifier method have a precision of 0.2‰ and all results except those of Diploria strigosa-TR2 fall within one standard deviation of results from the dialysis/combustion method. The modern azooxanthellate coral Desmophyllum dianthus δ15N value is 9.49‰ using the dialysis/combustion method, and 10.30‰ using the persulfate/denitrifier method (Fig. 2; Table 2) and correlates well with the modern azooxanthellate range (9–18‰19, 25). Miocene samples Oulophyllia sp. and Caryophyllia sp. were too small to be analyzed using the dialysis/combustion method, so they were only analyzed using the persulfate/denitrifier method (Table 2). The δ15N values of deep-water azooxanthellate-like Caryophyllia sp. samples are similar to those of modern azooxanthellate corals, 11.3‰ and 10.7‰, respectively (Figs 2, 3).
Coral Organic Matrix δ15N as Proxy for Photosymbiosis
A systematic isotopic offset of ~7‰ in coral skeletal organic matrix δ15N can be observed between extant zooxanthellate and azooxanthellate corals10, 19 (Figs 2, 3) The δ15N offset can be explained partly as the result of the difference in internal N recycling between symbiotic and asymbiotic corals, and partly by the difference in the δ15N of N source to the corals at the site of coral growth10, 25,26,27. For the first time, we report high δ15N ratios in two small, solitary deep-water Miocene Caryophyllia corals (small solitary forms) similar to those of extant azooxanthellate corals (Figs 2, 3). In contrast, the zooxanthellate-like coral, Acropora papillare, from shallower waters recorded much lower values (Figs 2, 3). These data demonstrate that both zooxanthellate and azooxanthellate fossil corals can be distinguished in the fossil record and indicate that the organic matrix δ15N has the potential to serve as a successful proxy for photosymbiosis10, 19.
Results from Triassic corals from Turkey (1.9–5.2‰; Tables 1, 2) correlate well with previously reported results from the same region10, 11, 20, 24. These results corroborate the hypothesis that the δ15N of Norian N sources to the Alakir Çay Valley region was low10 (Figs 2, 3), indicating that the region was oligotrophic similar to the modern western and subtropical North Atlantic10, 27 with a rich diversity of symbiotic corals, including solitary, phaceloid, cerioid, meandroid, and thamnasterioid morphologies10, 11, 20, 24.
Recrystallization of fossil corals from aragonite to calcite has been suggested to reduce the amount of organic matrix21,22,23, but not the nitrogen isotope ratio24. As a preliminary test, we analyzed a completely recrystallized Oligocene coral, Antiguastrea lucasiana (Tables 1, 2; Fig. 1e), which we expected to have a high δ15N if diagenesis has a significant impact. This coral, which is zooxanthellate based on its morphology11, has δ15N values (4.06 ± 0.26‰; Table 2; Fig. 2) consistent with a modern zooxanthellate coral (Fig. 2). This result suggests that the δ15N of the coral organic matrix is likely still isolated from the environment during mineralogical phase transition, causing a minimal impact on the δ15N values. Preserved insoluble organic matrices have been reported from Pliocene29, Cretaceous30, and Triassic scleractinian corals20, 31 suggesting that the δ15N results presented herein most likely represent true organic matrix values and are not diagenetic. Decay of coral organic matrix, however, may lead to lower organic matrix concentrations impacting the precision of the δ15N measurements21, 24, 32,33,34. A comprehensive, detailed analysis of the possible biases or possibilities for contamination in ancient organically bound nitrogen archives is beyond the scope of this project, but, overall, the robustness of the δ15N proxy indicates that it is more widely applicable as a test for ancient photosymbiosis than the δ18O/ δ13C isotope proxy of Stanley and Swart7.
In order to thoroughly test the impact of diagenesis on this proxy, one would need to use more precise analytical techniques and diagenetic screenings of several specimens with a diagenetic gradient; ideally, one part of their skeleton would be clearly altered (calcitic) and another would be pristine (aragonitic). Thus, if the δ15N signatures were identical in these regions, one could conclude that this proxy would be applicable to a wide range of fossil corals, including those that are not perfectly preserved.
We performed a direct comparison of the original dialysis/combustion method19 and the newly developed persulfate/denitrifier method25 for analyzing organic matrix δ15N in coral skeletons. One significant advantage of the persulfate/denitrifier method is that the sample size requirement is >100 times smaller than the dialysis/combustion method; only 5-10mg of coral carbonate is required for each analysis. This difference in sample size has huge significance for using museum specimens, small corals, or partially altered corals. This method also allows sampling to be more precise so that microborings or other small-scale contaminations can be avoided, which is not always the case with the dialysis/combustion method. Furthermore, in this study the dialysis/combustion resulted in large standard deviations (Table 2) whereas the persulfate/denitrifier method has a precision of 0.2‰27.
When enough mass was available (15–50 g) coral samples were processed twice to test the procedural precision of the dialysis/combustion method and were separated in two trials “TR-1” and “TR-2” respectively (Fig. 2). Only the trials of Favia fragum and Gablonzeria sp. (1) differed by more than 0.5‰ (Table 2). Although the variability between trials may be cause for concern, organic matrix δ15N values of modern corals are known to vary over time by up to 3‰35. Duplicate samples consist of different parts of the same coral, and so each trial may represent different times of growth; therefore, the difference between the means of the trials is likely an artifact of the large sample size required for the dialysis/combustion method (i.e. 15–50 g19).
Results from the persulfate/denitrifier method are within one standard deviation of results from the dialysis/combustion method, except results for Diploria strigosa-TR2 (Table 2). This difference may be caused by the fact that the dialysis/combustion method uses dialysis to collect the organic matrix after dissolution of the coral skeleton. Through dialysis, a water-soluble fraction of the organic matrix is lost. In contrast, the persulfate/denitrifier method analyzes the bulk organic matrix including both the soluble and insoluble organic matrix. This fundamental difference in actual analyte could result in differences in the final δ15N values19. A thorough study of the differences in nitrogen composition of aggregate and soluble organic matrix in fossil corals would be needed to resolve the influence of soluble organic matrix on δ15N values. In sum, we suggest that the persulfate/denitrifier method should be the favored technique as it uses less material, can be more precise in the sampling, and results in the analysis of both soluble and insoluble organic matrix.
Implications for Fossil Coral Paleobiology
The origin of scleractinian photosymbiosis is debated; some ancestral state reconstructions suggest that the first Scleractinia were zooxanthellate36, whereas others suggest that scleractinians originated as azooxanthellate6, 37. Morphological similarities between modern scleractinians and some tabulate corals (i.e. platy morphology) indicate that algal-coral symbiosis may have existed as early as the Paleozoic11, 38, but some authors conclude that Paleozoic corals lacked symbionts39. Scleractinians became the dominant reef-builders during the Late Triassic and the similar evolutionary radiation of dinoflagellates and Scleractinia strongly suggest that photosymbiosis evolved during this time1, 6, 40, 41. The Triassic fossil corals analyzed herein (Norian age, see Table 1) yielded zooxanthellate-like δ15N values (1.9–5.2‰; Figs 2, 3); these results, which are in agreement with the symbiotic assessments of Triassic corals from previous studies7, 9, 10, 17, 19, 24 (Fig. 3) corroborate the hypothesis that the evolutionary success of Scleractinia as reef-builders is linked to photosymbiosis, as evidenced by the Late Triassic reef bloom6, 12, 42.
Triassic corals could not have been symbiotic with modern zooxanthellae of the genus Symbiodinium 1, 3, 43, 44. Molecular clock analyses suggest that this genus originated in the early Eocene (~50 Ma), and that the majority of extant lineages diversified during the mid-Miocene45 (~15 Ma). If Triassic corals were symbiotic with a different group of dinoflagellates [1, 40], or if Triassic corals did not receive as much nitrogen from their symbionts as modern corals (i.e., if Triassic symbiosis was not as efficient as modern symbiosis), the internal N recycling between the coral host and symbiont would be different from modern corals. Discrepancies in this symbiotic relationship may account for coral skeletal-bound δ15N variability, which could cause the distinction between azooxanthellate and zooxanthellate fossil corals to be less clear. For example, N recycling efficiency has been shown to cause up to 2‰ variations in modern corals27. A larger dataset of fossil corals from different localities and time periods is needed to refine the confidence intervals as they apply to the fossil record, particularly as there is some overlap in the range of zooxanthellate and azooxanthellae δ15N values in modern corals10, 25, 27 (Fig. 3).
A better understanding of how zooxanthellate and azooxanthellate corals interacted through their evolutionary history will also provide data about how these corals reacted to environmental perturbations; for instance, a comprehensive assessment of the presence/absence of photosymbiosis in deep time is now possible with the δ15N proxy. Ancient analogues of coral reef success, demise, and symbiosis can help ecologists predict how scleractinians will be affected by the current carbon cycle perturbation created by anthropogenic carbon dioxide emissions.
Nitrogen Cycling in Ancient Oceans
The coral δ15N proxy provides not only information on photosymbiosis, but also additional information about the marine nitrogen cycle in the past because coral δ15N is affected by regional nitrogen sources to corals25, 27 (Fig. 3). In the modern ocean, most zooxanthellate corals live in oligotrophic environments1, 2, 19, the cause of which has been attributed to the ecological advantage of coral-zooxanthellae symbiosis with more efficient internal nutrient cycling1, 2, 10, 19. A key implication of the Triassic coral results is that the early development and diversification of major fossil reefs happened in oligotrophic oceans1, 10.
The modern Coral Triangle is characterized by exceptional ecological diversity46,47,48; however, little is known about its early environment and development46,47,48. We measured the δ15N of several well-preserved early Miocene corals from Indonesia collected as part of the THROUGHFLOW Project48. Both shallow-water (Acropora papillare) and deep-water corals (Caryophyllia sp.) from adjacent sites of similar Miocene age (18–20 Ma48) were analyzed. The Acropora papillare has a δ15N of ~3‰ whereas the Caryophyllia corals have a δ15N of ~11‰ (Table 2; Fig. 3). Despite the small number of samples analyzed (n = 3), this pair of shallow-water and deep-water corals from adjacent sites provides important information about the symbiosis and environment of the Coral Triangle in the early Miocene. First, the Acropora sample has a low δ15N value, suggesting that this coral was symbiotic10, 19 (Figs 2, 3). Second, the δ15N values of both Acropora and the Caryophyllia are among the lowest relative to their modern range respectively (Fig. 3); the calculated δ15N of regional nitrate supplied to the euphotic zone was 2.5‰ (Fig. 3), which probably indicate active N fixation in the ambient ocean and a nutrient-depleted environment. Combined, these δ15N results indicate that the early Miocene Coral Triangle was a nutrient-depleted environment containing symbiotic corals. It is, thus, possible that the nutrient-depleted environment provided an ecological advantage to symbiotic corals and encouraged rapid diversification within the Coral Triangle.
The following corals were analyzed in this study: two modern (<100 years old) zooxanthellate corals, one modern azooxanthellate coral, one zooxanthellate-like Holocene coral, three Miocene corals (two zooxanthellate-like and one azooxanthellate-like), one Oligocene zooxanthellate-like coral, and three Triassic zooxanthellate-like corals (Table 1). The Triassic samples used in this study, Gablonzeria sp. and Distichomeandra sp., were collected from the same locality as Triassic samples previously studied using the δ15N photosymbiosis proxy (Alakir Cay Valley, Turkey)10, 19, 24; additional analyses of Triassic fossil corals from Austria can be found in supplemental data. The material analyzed comprises a range of coral morphologies and growth forms; fossil material includes both zooxanthellate and azooxanthellate-like lifestyles inferred for corallite size, growth form, and level of corallite integration11 (Table 1).
To assess the diagenetic overprint of the coral samples several techniques were employed including microscopic assessments of polished slabs, thin section petrography, and scanning electron microscopy (SEM), and laser Raman spectroscopy. Specimens were cut perpendicular to corallite growth for thin sections and macroscopic observations. SEM SE images were used to reveal the microstructural composition of the coral skeletons because aragonite consists of needle-like structures, but calcite is blocky in appearance. Samples were coated in gold and photographed in high vacuum in a JEOL JSM-6490LV SEM. Micro-laser Raman spectroscopy was performed using the Bruker 785 nm red laser system at Rensselaer Polytechnic Institute (RPI) on thin sections and grain mounts of carbonate skeletal material and occluded organics in fossil specimens (see supplemental data). Spectra were generated on a 50 um spot, with 5 co-additions of 4 scans each and corrected background; reference spectra on an aragonite standard were performed for comparison. The diagenetic screenings of Triassic samples Gablonzeria and Distichomeandra (2) revealed infills that were removed prior to analyses (Figs S10, S12).
Sample preparation for nitrogen isotope analyses was conducted following the original dialysis/combustion method19 (see Fig. S1). When enough mass was available (i.e. 15–50 g), a duplicate sample was also processed to test for procedural contamination. The original and replicate samples are labeled “TR1” and “TR2” respectively. Desmophyllium dianthus, and the Miocene samples did not yield enough mass for replicates.
All specimens were soaked in sodium hypochlorite (commercial bleach, 8.25%) overnight, rinsed with deionized water, and dried at 50 °C overnight. Dried samples were then crushed to a fine powder in a ceramic mortar and pestle. The powder was treated with 2 M Sodium Hydroxide (NaOH) at 80 °C for 15 minutes to remove any residual extrinsic organic material. The samples were rinsed with deionized water through a vacuum filtration system and the cleaned powders were dried in an oven overnight at 50 °C and weighed. The skeletal powders were then decalcified by addition of 4M Hydrochloric acid (HCI). Acid was added until the powder dissolved and the evolution of bubbles ceased. Samples were moisturized with Milli-Q water prior to acid addition to prevent excessive bubble formation. A magnetic mixer was used to stir each solution on a stir plate. The decalcified samples were then neutralized by the addition of 4M NaOH using a glass pipette to achieve a neutral pH of 7. During neutralization, approximate pH values were inferred by applying droplets of each solution onto pH strips. After neutralization, samples were dialyzed using Spectra/Por bags with a molecular-mass cutoff of 3.5 kDa. The Spectra/Por bags were placed in glass containers filled with Milli-Q water that could withhold four times the volume of the sample. The Milli-Q water was changed every 3–4 hours, for a total of four times a day and then left overnight. The dialyzed material was transferred into 24 oz. Nasco sterile Whirl-pak bags. The bags were chosen due to their ability to hold liquid without leakage and their sterility to avoid contamination. The bags were then frozen at −20 °C and lyophilized using a Virtis Benchtop SLC freeze dryer.
δ15N values were measured using an Elementar Vario Isotope Select Elemental Analyzer connected to a Micromass Isotope Ratio Mass Spectrometer (EA-IRMS) at RPI. In house reference materials as well as IAEA600, SRM1547, and NBS-22 were used as isotope calibration standards. Each freeze-dried sample was divided into triplicates and transferred into tin cups, and were run in blocks bracketed by standards and blanks. Nitrogen isotope ratios were corrected relative to that of the IAEA 600 standard and reported in delta notation in units of parts per thousand or per mil (‰ vs. AIR). Long-term precision on IAEA standards and in-house reference materials on the continuous flow IRMS is 0.2‰ for δ15N and 0.03‰ for δ13C.
Several samples were analyzed using the newly developed persulfate/denitrifer method at Princeton University. Briefly, the protocol uses 5–10 mg of coral skeleton powder, which is cleaned oxidatively with concentrated sodium hypochlorite solution to remove any external N contamination. After cleaning, the sample is dissolved by addition of 4 M HCl. The resulting organic matter is oxidized into nitrate using a basic potassium persulfate solution and then converted bacterially into nitrous oxide, which can be measured for δ15N by gas chromatography-isotope ratio mass spectrometry (for detailed procedure see Wang et al.25).
For materials requests and correspondence please contact Chiara Tornabene: firstname.lastname@example.org or Rowan C. Martindale: email@example.com.
Stanley, G. D. & van de Schootbrugge, B. The evolution of the coral-algal symbiosis in Coral Bleaching: patterns, processes, causes and consequences (eds. van Oppen, M. J. H. & Lough, J. M.) 7–19 (Springer, 2009).
Tambutté, S. et al. Coral biomineralization: From the gene to the environment. J. Exp. Mar. Biol. Ecol. 408, 58–78 (2011).
Davy, S. K., Allemand, D. & Weis, V. M. Cell biology of cnidarian-dinoflagellate symbiosis. Microbiol. Mol. Biol. R. 76, 229–261 (2012).
Goreau, T. F. & Goreau, N. I. The physiology of skeleton formation in corals. II. Calcium deposition by hermatypic corals under various conditions in the reef. Biol. Bull. 117, 239–250 (1959).
Allemand, D. et al. Biomineralisation in reef-building corals: From molecular mechanisms to environmental control. Comptes Rendus-Palevol 3, 453–467 (2004).
Stanley, G. D. Early history of scleractinian corals and its geological consequences. Geology 9, 507–511 (1981).
Stanley, G. D. & Swart, P. K. Evolution of the coral-zooxanthellae symbiosis during the Triassic; A geochemical approach. Paleobiology 21, 179–199 (1995).
Rosen, B. R. Algal symbiosis, and the collapse and recovery of reef communities: Lazarus corals across the K-T boundary 164-180 (Cambridge University Press 2000).
Stanley, G. D. & Hemle, K. P. Middle Triassic coral growth bands and their implication for photosymbiosis. Palaios 25, 754–763 (2010).
Frankowiak, K. et al. Photosymbiosis and the expansion of shallow-water corals. Sci Adv. 2, e1601122 (2016).
Coates, A. G. & Jackson, J. B. C. Clonal growth, algal symbiosis, and reef formation by corals. Paleobiology 13, 363–378 (1987).
Kiessling, W. Reef expansion during the Triassic: Spread of photosymbiosis balancing climatic cooling. Palaeogeogr. Palaeoclimatol. Palaeoecol. 290, 11–19 (2010).
Kiessling, W. & Kocsis, Á. T. Biodiversity dynamics and environmental occupancy of fossil azooxanthellate and zooxanthellate scleractinian corals. Paleobiology 41, 402–414 (2015).
Rosen, B. R. & Aillud et al. Platy coral assemblages: 200 million years of functional stability in response to the limiting effects of light and turbidity. ICRS 9(1), 255–264 (2000).
Stanton, R. J. Nutrient models for the development and location of ancient reefs. Geo. Alp 3, 191–206 (2006).
Stolarski, J. Three-dimensional micro- and nanostructural characteristics of the scleractinian coral skeleton: A biocalcification proxy. Acta Palaeontol. Pol. 48, 497–530 (2003).
Frankowiak, K. et al. Fine-scale skeletal banding can distinguish symbiotic from asymbiotic species among modern and fossil scleractinian corals. PloS One 11, e0147066 (2016).
Reuter, M., Brachert, T. C. & Kroeger, K. F. Diagenesis of growth bands in fossil scleractinian corals: Identification and modes of preservation. Facies 51, 146–159 (2005).
Muscatine, L. et al. Stable isotopes (δ13C and δ15N) of organic matrix from coral skeleton. Proc. Natl. Acad. Sci. USA 102, 1525–1530 (2005).
Frankowiak, K., Mazur, M., Gothmann, A. M. & Stolarski, J. Diagenetic alteration of Triassic coral from the aragonite Konservat-Lagerstätte in Alakir Çay, Turkey: Implications for geochemical measurements. Palaios 28, 333–342 (2013).
Whitelaw, M. J., Batts, B. D., Murray-Wallace, C. V. & Mcrae, C. R. Diagenesis of the organic matrix in Anadara trapezia during the Late Quaternary: preliminary findings. Proc. Linn. Soc. N.S.W. 123, 225–234 (2001).
Perrin, C. & Smith, D. Decay of skeletal organic matrices and early diagenesis in coral skeletons. Comptes Rendus Palevol 6, 253–260 (2007).
Perrin, C. & Smith, D. C. Earliest steps of diagenesis in living scleractinian corals: Evidence from ultrastructural pattern and Raman spectroscopy. J. Sediment. Res. 77, 495–507 (2007).
Peters, P. Detecting photosymbiosis in Triassic corals using stable isotopes (δ13C, δ18O and δ15N) (Universiteit Utrecht Bachelor Thesis) (2014).
Wang, X. T. et al. Influence of open ocean nitrogen supply on the skeletal δ15N of modern shallow-water scleractinian corals. Earth Planet. Sci. Lett. 441, 125–132 (2016).
Wang, X. T. et al. Isotopic composition of carbonate-bound organic nitrogen in deep-sea scleractinian corals: A new window into past biogeochemical change. Earth Planet. Sci. Lett. 400, 243–250 (2014).
Wang, X. T. et al. Isotopic composition of skeleton-bound organic nitrogen in reef-building symbiotic corals: A new method and proxy evaluation at Bermuda. Geochim. Cosmochim. Acta 148, 179–190 (2015).
Herman, R. G., Bogdan, C. E., Sommer, A. J. & Simpson, D. R. Discrimination among carbonate minerals by Raman spectroscopy using the laser microprobe. Appl. Spectrosc. 41, 437–440 (1987).
Yamazaki, A., Watanabe, T., Takahata, N., Sano, Y. & Tsunogai, U. Nitrogen isotopes in intra-crystal coralline aragonites. Chem. Geol. 351, 276–280 (2013).
Sorauf, J. E. Skeletal microstructure, geochemistry, and organic remnants in cretaceous scleractinian corals: Santonian Gosau beds of Gosau, Austria. J. Paleontol. 73, 1029–1041 (1999).
Gautret, P., Cuif, J. P. & Stolarski, J. Organic components of the skeleton of scleractinian corals: evidence from in situ acridine orange staining. Acta Palaeontol. Pol. 45, 107–118 (2000).
Bahlmann, E. et al. Performance evaluation of nitrogen isotope ratio determination in marine and lacustrine sediments: An inter-laboratory comparison. Org. Geochem. 41, 3–12 (2010).
Young, S. D. Organic material from scleractinian coral skeletons I. Variation in composition between several species. Comp. Biochem. Physiol. B, Biochem. Mol. Biol. 40, 113–120 (1971).
Robbins, L. L. & Ostrom, P. H. Molecular isotopic and biochemical evidence of the origin and diagenesis of shell organic material. Geology 23, 345–348 (1995).
Ren, H. et al. 21st-century rise in anthropogenic nitrogen deposition on a remote coral reef. Science 356 (6339):749-752 (2017).
Barbeitos, M. S., Romano, S. L., Lasker, H. R. & Ayala, F. J. Repeated loss of coloniality and symbiosis in scleractinian corals. Proc. Natl. Acad. Sci. USA 107, 11877–11882 (2010).
Kitahara, M. V., Cairns, S. D., Stolarski, J., Blair, D. & Miller, D. J. A comprehensive phylogenetic analysis of the Scleractinia (Cnidaria, Anthozoa) based on mitochondrial CO1 sequence data. PloS One 5, e11490 (2010).
Zapalski, M. K. Evidence of photosymbiosis in Paleozoic tabulate corals. Proc. R. Soc. B. 281, 20132663 (2014).
Scrutton, C. The palaeozoic corals. II: structure, variation and palaeoecology. Proc. Yorkshire Geol. Soc. 52, 1–57 (1998).
Palliani, R. B. & Riding, J. B. Subdivision of the dinoflagellate cyst family Suessiaceae and discussion of its evolution. Journal of Micropalaeon. 19, 133–137 (2000).
Flügel, E. Triassic reef patterns, in Phanerozoic Reef Patterns, Kiessling, W., Flügel, E., Golonka, J. Eds. SEPM Special Publication, 391–463 (2000).
Coffroth, M. A. & Santos, S. R. Genetic diversity of symbiotic dinoflagellates in the genus Symbiodinium. Protist 156, 19–34 (2005).
Stat, M., Carter, D. & Hoegh-Guldberg, O. The evolutionary history of Symbiodinium and scleractinian hosts-Symbiosis, diversity, and the effect of climate change. Perspect. Plant Ecol. Evol. Syst. 8, 23–43 (2006).
Fransolet, D., Roberty, S. & Plumier, J. Establishment of endosymbiosis: The case of cnidarians and Symbiodinium. J. Exp. Mar. Biol. Ecol. 420–421, 1–7 (2012).
Pochon, X., Montoya-Burgos, J. I., Stadelmann, B. & Pawlowski, J. Molecular phylogeny, evolutionary rates, and divergence timing of the symbiotic dinoflagellate genus Symbiodinium. Mol.Phylogenet. Evol. 38, 20–30 (2006).
Johnson, K. G., Hasibuan, F., Müller, W. & Todd, J. A. Biotic and environmental origins of the southeast asian marine biodiversity hotspot: The Throughflow project. Palaios 30, 1–6 (2015).
Santodomingo, N. et al. A diverse patch reef from turbid habitats in the middle Miocene (east Kalimantan, Indonesia). Palaios, 30, 128–149 (2015).
Renema, W. et al. Ages of Miocene fossil localities in the northern Kutai basin (East Kalimantan, Indonesia). Palaios 30, 26–39 (2015).
We would like to thank all the experts who offered valuable input for this research project: Drs. Ewa Roniewicz and Ann Budd for taxonomic identifications, Daniel Sigman and Terrence Quinn for access to laboratories, Timothy Shanahan and Daniel Breecker for access to equipment. We also gratefully acknowledge the Geologische Bundesanstalt Wien (Irene Zorn), the Naturhistorisches Museum Wien (Alexander Lukeneder), and the Natural History Museum London (Kenneth Johnson) for loaning the Miocene, Oligocene, and Triassic samples, as well as Leopold Krystyn for the donation of Triassic samples from Turkey and Jess Adkins for the modern azooxanthellate sample. We also thank three anonymous reviewers whose suggestions greatly improved this manuscript. Lastly, the associations that funded this research are gratefully acknowledged: the Muehlberger Field Scholarship Fund, the Paleontological Society of Austin, and the University of Texas at Austin.
The authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
About this article
Cite this article
Tornabene, C., Martindale, R.C., Wang, X.T. et al. Detecting Photosymbiosis in Fossil Scleractinian Corals. Sci Rep 7, 9465 (2017). https://doi.org/10.1038/s41598-017-09008-4
Global Change Biology (2020)
Review of Palaeobotany and Palynology (2020)
Stratigraphic distribution and paleoecological significance of Early Jurassic (Pliensbachian-Toarcian) lithiotid-coral reefal deposits from the Central High Atlas of Morocco
Palaeogeography, Palaeoclimatology, Palaeoecology (2019)
Earth-Science Reviews (2018)