Spatial and compositional pattern in mineralogy
The dominant mineral phase in Panggang is aragonite (Table 1). Its concentration varies little between facies (Table 1, Fig. 3), with slightly increased values close to areas with dense coral cover on the reef front (Fig. 4). The aragonitic red algae genera Galaxaura and Liagora also thrive on the Kepulauan Seribu reef slope (Atmadja 1977), likely also contributing to the high aragonite concentrations. HMC and LMC occur in fairly similar amounts in all facies (Table 1, Fig. 3), with slightly increased values of LMC in the mollusc wackestone facies of the inner lagoon and some isolated points on the western reef flat (Fig. 4). Corals as the main aragonite producers are less abundant in this area, while less light dependent carbonate producers, such as molluscs and echinoderms, are more common (Table 1). The highest LMC content is found in the mollusc wackestone facies (Table 1), which is known to dominate in the inner lagoon where coral grainstone and coral packstone/grainstone facies are absent (Utami et al. 2018). High LMC content in the inner lagoon is likely due to the contribution from molluscs, the most important component in the mollusc wackestone facies (Table 1). Bivalves with calcitic or mixed aragonite-calcite shells in Kepulauan Seribu belong to the families Ostreidae (LMC), Pectinidae (LMC + ARA) and Pinnidae (LMC + ARA) (van der Meij et al. 2009; Ueda 2020). Five individual bivalve shells were measured using XRD. All of them have a mixed aragonite–calcite shell containing on average 90% aragonite and 10% LMC. Locally, LMC is also abundant in the reef flat (Fig. 4), likely due to contributions from bivalves and larger benthic foraminifers (e.g. Amphistegina and Elphidium, Table 4) (Blackmon and Todd 1959) which are abundant in the coral grainstone facies (Table 1). SEM analysis shows that calcareous nannoplankton, as a potential LMC source, is near absent in the sediment.
No clear correlation exists between HMC and individual biogenic grain types (Table 3). This is caused by the varied contribution of several different biota to the overall HMC content to the sediment. Red algae, such as Amphiroa, that are typically associated with sea grass in Kepulauan Seribu (Atmadja 1977), contribute HMC to the coral grainstone facies on the reef flat (Table 1). Echinoderms are more important in the deeper water coral-mollusc packstone facies and mollusc wackestone facies (Table 1) compared to other facies. Increased HMC contents in the inner lagoon (Fig. 4) can therefore be interpreted as a higher contribution from echinoderms. Mg content is generally higher in coralline algae compared to other organisms with HMC skeleton (Chave 1954). This is consistent with the observation that the lowest Mg contents in HMC are found in the mollusc wackestone facies where coralline algae are rare (Table 1, Fig. 3).
The mineralogy of five fine fraction (< 125 µm) subsamples from the coral-mollusc packstone and mollusc wackestone facies was analyzed using XRD. The aragonite content in these fine fraction subsamples is 6% lower compared to their respective bulk samples (Table 2). In contrast, their HMC content is increased by 4% (Table 2). It has been proposed that fish can produce significant amounts of HMC mud (Salter et al. 2012). Most of the important (> 5% abundance) fish families in the area of Kepulauan Seribu (Madduppa 2013) produce Mg-rich amorphous calcium carbonate. This carbonate phase commonly dissolves within hours and therefore is unlikely to be preserved as sediment (Salter et al. 2017). Only one family (Apogonidae) produces mud-sized HMC ellipsoids with Mg content of around 30–45% (Salter et al. 2018). However, the typical crystal morphologies have not been observed during SEM investigations in our samples. It is, therefore, unlikely that fish-derived Mg calcite contributes significantly to the increased HMC values in the fine fraction. In contrast, SEM investigations confirm that the fine fraction is largely a product of skeletal breakdown (Fig. 2). Carbonate components that occur only in the mud to fine sand size (< 125 µm) fraction, such as tunicate spiculae and needles from Padina thalli, are exclusively aragonitic. Laboratory studies suggest that abrasion rates of corals are much slower compared to dominant HMC components, such as coralline algae and echinoderms (Milliman et al. 1974). Increased HMC content in the fine fraction (Table 2) therefore could be related to preferential breakdown of HMC components. However, it remains unclear in how far these laboratory results can be applied to natural settings. Quantitative data on the composition of different grains size fractions from isolated carbonate platforms of Belize were presented by Gischler and Zingeler (2002). Our calculations based on this data indicate that coral fragments are actually increased in abundance relative to echinoderm and red algae fragments in the fine (< 125 µm) fraction as compared to the larger grain size fraction. This is opposite to what would be expected by the laboratory results on mechanical abrasion (Milliman et al. 1974), possibly indicating that other processes, such as bioerosion, are more important in reef settings. The preferential abrasion of HMC components therefore seems not to be the reason for the increased HMC content in the fine fraction. A known source of mud-sized HMC are red algae growing on sea grass (Land 1970; Perry et al. 2019). Patches of sea grasses are common on the Seribu islands and are associated with reduced grain sizes in the sediment, likely due to their baffling effect (Utami et al. 2018). They also might contribute to the increased HMC contents in the fine fraction. However, a comparison of the mineralogical composition between locations with dense and sparse sea grass cover on the sand apron of Panggang does not show a significant difference with respect to HMC content (Mann–Whitney U test, p = 0.91). Due to the abundance of sea grass, our dataset does not include samples from locations without sea grass.
Alternatively, micritization in northern Belize increases the HMC content of skeletal grains, which later are preferentially abraded and contribute to carbonate muds. Cementation in micro-bioerosion traces is an important mechanism for micritization (Reid and Macintyre 2000). High Mg calcite cement has been observed, besides other mineral phases, in endolithic microborings in corals (Schroeder 1972). Bioerosion and/or micritization are ubiquitous in reefs affected by terrestrial runoff or upwelling in SE Asia (Tomascik et al. 1997; Wilson 2012; Madden and Wilson 2012) including the Seribu Islands (Fig. 2) (Park et al. 1992, 2010). Both processes are most intense at the outer surface of carbonate grains. The higher surface-to-volume ratio of smaller grains could therefore lead to stronger alteration of its mineralogy compared to larger grains. Micritization and bioerosion could therefore have contributed to the increased HMC content in the fine fraction of Panggang (Table 2).
Small amounts of siliciclastic minerals are found in Kepulauan Seribu (Table 1). Quartz content increases with water depth (Table 3), in the inner lagoon and reef front. Quartz and smectite are strongly concentrated in the fine fraction (Table 2), indicating that quartz grains will settle from suspension in low-energy environments but are easily winnowed by waves and currents in high-energy environments. Smectite occurs only in > 6 m water depth within the inner lagoon (Fig. 4), where it likely settles below the wave base. The fact that siliciclastic minerals are found in almost all samples (Fig. 4) shows that Panggang is influenced by terrestrial runoff (Fig. 1). A possible source for the siliciclastics could be Sumatra, since Kepulauan Seribu is relatively isolated from a direct influence from Java and Kalimantan by the regional current system (Tomascik et al. 1997; Jordan 1998). Partially due to this siliciclastic influence, the depth of abundant coral growth in Kepulauan Seribu (0–18 m, Jordan 1998) is typical for equatorial carbonates in SE Asia, but very shallow compared to other regions (0–100 m) with clearer water and lower nutrient input (Wilson 2008).
Spatial patterns in stable isotope composition of sediment
Delineating facies based on their δ13C and δ18O values is not possible due to a lack of a significance difference between carbonate facies (Kruskal–Wallis test, p = 0.17 and p = 0.15, respectively). However, the average stable isotope values of the coral grainstone facies (δ13C = 0.0, δ18O = − 4.4) is lower compared to the mollusc wackestone facies (δ13C = 0.4, δ18O = − 4.0), suggesting some facies control (Fig. 6). Coral abundance shows a significantly negative correlation to δ13C and δ18O values (Table 3). This results from a strong kinetic control on isotope fractionation in corals (McConnaughey 2003). Coral skeletons therefore show a depletion of heavy isotopes and consequently a strong offset from equilibrium values towards more negative isotope ratios (Fig. 6). Corals exert a strong influence on the bulk isotope composition, since they are the predominant skeletal material and major aragonite producer (Fig. 6). Marine mollusc shells, as the second most important sedimentary component (Table 3), do not show kinetic effects in their isotopic composition (McConnaughey and Gillikin 2008). Their δ18O values are often close to equilibrium values (Grossman and Ku 1986) and therefore more positive compared to corals (Fig. 6) (Aharon 1991). Carbon isotope signatures in molluscs are complex, with near equilibrium values e.g. in giant clams (Aharon 1991) but often several ‰ lighter due to metabolic carbon contribution (McConnaughey and Gillikin 2008). In reefal carbonates, molluscs are typically less depleted in 13C compared to cooccurring corals (Fig. 6) (González and Lohmann 1985). This results in a positive (negative) correlation (Table 3) between oxygen and carbon isotopes and molluscs (corals). Higher stable isotope values in the inner lagoon and sand apron of Panggang (Fig. 7) result from the increased abundance of molluscs in mollusc wackestone and coral packstone/grainstone facies (Table 1) and coincides with the overall lowest aragonite value (Fig. 4). The most negative δ13C and δ18O values occur at the reef front in coral grainstone and coral packstone/grainstone facies where coral cover is densest (Fig. 7). However, the compositional differences between individual facies types (Table 1) are not strong enough to result in separate isotope fields for each facies (Fig. 6).
The most important group of purely calcitic components are benthic foraminifers. Many shallow water benthic foraminifers from reefal environments precipitate their skeleton close to oxygen isotope equilibrium (± 1‰) (Saraswati et al. 2004; Maeda et al. 2017), although slightly stronger depletion of 18O was observed as well (Wefer and Berger 1991). Carbon isotopes typically are lighter compared to equilibrium values with common 13C depletion on the order of 1–3‰ for most miliolids and 2–5‰ for rotaliids (Wefer et al. 1981; Saraswati et al. 2004). This is consistent with lower δ13C values in rotaliids (Calcarina, Amphistegina) compared to miliolids (Sorites) in the study area (Fig. 6). Benthic foraminifers from reefal environments therefore are characterized by generally more positive stable isotope values compared to cooccurring corals (Fig. 6), but can show partial overlap in their δ13C ranges (González and Lohmann 1985; Gischler et al. 2012), especially if rotaliids are abundant. The observed negative correlation between aragonite content and stable isotope values (Table 3) likely results from this isotopic difference between corals and benthic foraminifers as the main aragonite and calcite producers, respectively. The fact that some molluscs contribute calcite to the sediment likely enhances this effect.
The positive correlation between stable isotopes and fines is a spurious relationship (Table 3), resulting from the strong positive (negative) correlation with molluscs (corals). This is supported by the observation that the fine fraction samples show no consistent isotopic difference to the bulk samples (Table 3). This similarity indicates that the fine fraction is genetically related to the coarse fraction and is formed by skeletal breakdown.
Comparison with other carbonate systems
To facilitate the comparison of mineralogical and stable isotope signatures between humid equatorial, arid (sub)tropical, and temperate carbonate systems we have grouped them geographically (Figs. 8 and 10). Carbonates from each geographical group share common characteristics and are deposited under similar environmental conditions (Fig. 9, Table 5). Carbonate systems from the tropical western Atlantic (TWA) and Persian Gulf represent mostly (sub)tropical carbonates (sensu Wilson 2002, 2012) characterized by at least seasonally arid conditions. These carbonates often contain facies with non-skeletal grains and are deposited in normal marine to hypersaline waters (Table 5). Exceptions are Florida Bay and Belize, which are characterized by riverine input leading to strong gradients in salinity ranging from brackish to hypersaline and absence or low abundance of non-skeletal grains. The carbonates from the Indo-Pacific Warm Pool (IPWP) (Figs. 8, 9, 10, Table 5) are deposited on isolated carbonate platforms under the highest average temperatures (> 28 °C) and high seasonal precipitation, but outside the direct influence of river runoff. Non-skeletal grains are uncommon. High latitude systems (Figs. 8, 9, 10, Table 5) are typical temperate water Heterozoan carbonates (Wilson 2012).
The δ13C signature of carbonates from Kepulauan Seribu is considerably more negative compared to most other platform and reef systems (Fig. 8). The carbon isotopic composition of marine carbonates is mainly influenced by the δ13C of the dissolved inorganic carbon (DIC) in water (Smith and Kroopnick 1981), the mineralogy (Rubinson and Clayton 1969; Romanek et al. 1992) and so called “vital effects” in carbonate secreting organisms (e.g. kinetic and metabolic effects) (McConnaughey 2003; McConnaughey and Gillikin 2008). Because of these vital effects skeletal carbonates are typically depleted in δ13C compared to inorganically precipitated carbonates from the same environment (Swart et al. 2009; Gischler et al. 2012). Non-skeletal grains and inorganically precipitated muds are much more common in modern platforms and reefs from the tropical, western Atlantic (TWA) and Persian Gulf compared to carbonate systems from the Indo-Pacific Warm Pool or New Zealand (Table 5) (Milliman et al. 1974; Gischler 2011; Gallagher et al. 2018). Most carbonates from the TWA and Persian Gulf therefore show broader δ13C ranges encompassing more positive δ13C values (Fig. 8).
Due to mineralogical effects, inorganic aragonite is enriched in δ13C by about 1.8 ‰ compared to coprecipitated calcite (Rubinson and Clayton 1969; Romanek et al. 1992). This mineralogical effect should favor a positive correlation between aragonite content and δ13C. However, the actual relation depends on the source and therefore isotopic signature of the aragonite. Aragonitic non-skeletal grains and green algae (Fig. 8) are characterized by more positive δ13C values compared to most calcitic skeletal grains, resulting in a positive correlation between both parameters (Swart et al. 2009; Gischler et al. 2012). Corals on the other hand are more depleted in δ13C compared to most calcite secreting organisms (Fig. 6), which explains the negative correlation between aragonite and δ13C in the coral dominated Kepulauan Seribu (Table 3) and the lower δ13C compared to the relatively calcite-rich cool water carbonates (Fig. 8). However, the δ13C signature of carbonate from Kepulauan Seribu is also more negative compared to most other coral dominated reef facies from the IPWP (Fig. 8).
Carbon isotope data from a Porites coral from Bidadari Island (Fig. 1) in the Java Sea, is on average more depleted compared to Porites corals from the Seychelles and Chagos both of which are located within the IPWP (Fig. 8) (Pfeiffer et al. 2004; Pfeiffer and Dullo 2006; Cahyarini et al. 2016). It is unlikely that this difference is solely due to kinetic related offsets, since all three coral colonies show similar growth rates (Pfeiffer et al. 2004; Pfeiffer and Dullo 2006; Cahyarini et al. 2016). More negative δ13C values of the Porites colony from Bidadari Island are more likely due to the influence of river run off. Corals have the potential to record land–ocean carbon transfer because they draw on dissolved inorganic carbon (DIC) for calcification. The δ13C of DIC in coastal water is influenced by the δ13C signature from adjacent rivers (Moyer and Grottoli 2011). Bidadari Island is located in the Java Sea (Fig. 1) and in close proximity to three major rivers in the Greater Jakarta Area (Farhan and Lim 2012). Estuaries in SE Asia show low δ13CDIC values due to the contribution of 12C from riverside mangroves (Miyajima et al. 2009). The low δ13C of the coral skeleton in Kepulauan Seribu therefore likely reflects the enrichment in 12C of ambient seawater due to the input of riverine DIC to the Java Sea.
To a somewhat lesser extent, light intensity (photosynthesis) also has been shown to influence the value of skeletal δ13C in corals (Moyer and Grottoli 2011). Corals are usually depleted in 13C relative to ambient seawater as a result of kinetic and metabolic fractionation (Swart 1983; McConnaughey 1989; Allison et al. 1996; Grottoli and Wellington 1999). Environmental variables that influence coral metabolism should therefore also affect δ13C levels in coral skeletons (Grottoli 2002). Under low light conditions photosynthesis decreases and metabolic fractionation increases with the consequence of decreased skeletal δ13C (Grottoli 2002). Sediment input from rivers into the Java Sea potentially enhances turbidity in the semi-enclosed basin, leading to a reduction of light intensity and photosynthesis thus resulting in reduced skeletal δ13C values. This would be consistent with the relatively shallow depth of reef growth (Jordan 1998) and the presence of siliciclastic minerals in sediment samples below the wave base (Fig. 4) in the study area. Turbidity can also be produced by algae blooms caused by increased nutrient input either through river water or untreated sewage from populated islands (Baum et al. 2015). Additionally, higher nutrient availability could influence the δ13C of coral skeletons by an increase in the heterotrophic feeding rate or an increase in the photosynthesis rate related to higher zooxanthellate concentrations (Grottoli 2002). However, these processes have opposing effect on the δ13C of coral skeletons. The net effect therefore will depend on the relative magnitudes of these various influences but is expected to be small (Heikoop et al. 2000).
The largest single anthropogenic influence on the isotope signature of recent corals likely is caused by the input of anthropogenic CO2 to the global surface ocean. The accelerated admixture of fossil fuel derived CO2 leads to a shift towards lower δ13C values in atmospheric CO2, the so called Suess effect (Keeling 1979). This Suess effect is often recorded in declining coral skeleton δ13C values over the twentieth century (Swart et al. 2010). Linsley et al. (2019) showed that the Suess effect can shift skeletal δ13C in corals by more than 1‰ since the 1950s. The carbon isotope composition of corals in Fig. 8, which are largely derived from the twentieth century, is therefore likely depleted relative to bulk sediments or sedimentary grains of mixed and/or unknown age.
Florida Bay is the only location that shows a nearly complete overlap but broader total range in δ13C values compared to the sediments from Kepulauan Seribu (Fig. 8). In contrast to many other areas in the TWA, non-skeletal grains and inorganic mud are lacking in Florida Bay (Table 5). Average δ13CDIC of surface waters in Florida Bay is lowest in areas influenced by isotopically depleted river runoff from the Everglades, while the δ13C variability in other areas is controlled by remineralization of marine organic matter (Swart and Price 2002). Similar to Kepulauan Seribu, the lowest values in δ13C of carbonates deposited in Florida Bay seem to be a result of isotopically depleted river runoff and a lack of inorganic carbonate precipitation.
Since the pioneering work by Urey (1947), the δ18O values of carbonate minerals are known to be controlled by temperature of formation. Other controlling factors are the δ18O value (δ18Ow) of the precipitating fluid (Urey 1947; Epstein and Mayeda 1953), mineralogy (Tarutani et al. 1969), and “vital effects” in carbonate secreting organisms (McConnaughey 2003). The slope of the δ18O/temperature relationship in carbonates is commonly assumed to be in the range from − 0.2 to − 0.25 ‰/°C (Grossman and Ku 1986; Bemis et al. 1998), resulting in a decrease of 0.8–1 ‰ δ18O for a four degree increase in temperature.
Carbonates from Kepulauan Seribu are characterized by much lower δ18O values compared to other carbonate systems (Fig. 8). This can be explained by the combined effect of high sea surface temperatures and low δ18Ow in the Java Sea (Fig. 9). Modelling results confirm that for carbonates precipitated in equilibrium with seawater, SE Asia is globally the region with the most negative δ18O signature (Fig. 9c). Within the IPWP the differences in temperature are small (Fig. 9a), but δ18O water is significantly reduced in SE Asia compared to e.g. the central equatorial Indian Ocean or Pacific (Fig. 9b).
Differences in hydrology are reflected in the lower δ18O values of a Porites coral from Kepulauan Seribu compared to those from the Seychelles and Chagos (Fig. 8). The difference of more than 1 ‰ in δ18O is more than can be explained by intercolony offsets (Sayani et al. 2019) or the temperature difference of ~ 1 °C between the Java Sea and the other two sites. Most of the difference therefore can be attributed to the contrasting δ18O seawater values, which based on the difference in salinity should be around 0.9–1.1 ‰/δ18O assuming a slope of 0.22 ‰ for the salinity/ δ18O relationship (Fairbanks et al. 1997). The lower δ18O seawater values in the Java Sea result from heavy precipitation and increased runoff, and are also influenced by the δ18Ow of the precipitation itself via the so called “amount effect” (Rozanski et al. 1993; Lau and Yang 2003).
Facies also could influence the δ18O signature of bulk sediments due to differences in the “vital effects” of carbonate producing organisms (McConnaughey 2003; McConnaughey and Gillikin 2008) and mineralogy (Tarutani et al. 1969). However, the different δ18O values for individual grain types to a large extent seem to average out for bulk samples (Gischler et al. 2012). This results in relatively large overlaps between facies in most atolls, barrier reefs and platforms (Gischler et al. 2012), showing that the influence of facies in these settings is limited.
The carbonate mineralogy of shallow water (< 20 m) bulk sediments in the tropics and subtropics (IPWP, TWA) is predominantly composed of aragonite and HMC, with LMC usually present in low quantities < 10% (Fig. 10) (Ginsburg 1956; Purdy 1963; Weber and Schmalz 1968; Gischler 2006; Reijmer et al. 2009; Gischler et al. 2013; Schmitt and Gischler 2017; Husinec et al. 2019). The only exception is the Kuwait ramp system that shows LMC contents of up to ~ 40% (Gischler and Lomando 2005). In reef atoll and platform systems higher LMC content > 10% are usually spatially restricted and related to local dominance of rotaliid foraminifers, such as Amphistegina (Weber and Woodhead 1972), contribution from calcite precipitated in meteoric waters (Swart and Kramer 1997) or from microbial activity (Pusey 1975; Purdy and Gischler 2003). Otherwise, higher LMC (7–14%) contents generally occur only in deeper water between 20 and 70 m, mainly contributed from planktonic foraminifers, coccoliths and molluscs (Weber and Schmalz 1968; Gischler 2006). Aragonite also dominates the mineralogy of the study area in Kepulauan Seribu but HMC and LMC occur in nearly equal proportions. The main mineralogical difference to other reefal systems in the IPWP is the occurrence of clastic mineral phases (Table 1) and a higher LMC content (~ 10%) in the shallow water sediment (Fig. 10).
Opposed to open ocean settings, the semi-enclosed Java Sea is influenced by river runoff. Terrestrial sediments reaching Kepulauan Seribu (Table 1) most likely are derived from Sumatran rivers, since oceanic currents isolate Kepulauan Seribu from sediment input from Java and Kalimantan (Tomascik et al. 1997; Jordan 1998). The suspended particulate matter promotes oligophotic conditions in relatively shallow water. Wilson and Vecsei (2005) argued that oligophotic conditions in water depth below 20 m are typical for humid equatorial carbonate systems in SE Asia and would favor the deposition of e.g. relatively calcite-rich foramol facies. Foramol facies in Kepulauan Seribu dominates in water depth below 27 m (Jordan 1998). In water depth < 20 m, rotaliid foraminifers contribute to LMC (Table 4) but their abundance is not high enough to explain the generally increased LMC content. Another important source of LMC in Kepulauan Seribu seems to be mixed aragonite-calcite bivalve shells. More mineralogical analysis from humid equatorial settings are needed to test if the increased LMC content observed in Kepulauan Seribu is a common feature of shallow water (< 20 m) reef systems in humid, equatorial settings.
Fine fraction sediment from the Panggang reef platform is showing a higher HMC compared to the bulk fraction, likely related to micritization and bioerosion. Higher HMC contents in the fine fraction are also found in Belize, the potential source is micritization of skeletal grains to Mg calcite (Weber and Schmalz 1968). Micritization of skeletal grains is also an important diagenetic process in other shallow marine carbonate platforms, e.g. The Bahamas where the grains are altered to aragonite (Purdy 1963; Bathurst 1966; Kendall and Skipwith 1969; Trumbull 1988). Preferential alteration of the carbonate grains into high Mg calcite is potentially controlled by lower aragonite supersaturation in water off Kepulauan Seribu. Fresh water influx from rivers promotes low salinities in the semi-enclosed seas of SE Asia, leading to relatively low aragonite supersaturation compared to more arid regions as the tropical, western Atlantic (Park et al. 2010). Seawater Mg:Ca ratios are another important control on carbonate mineralogy, with lower Mg:Ca favouring calcite precipitation at a given temperature (Morse et al. 1997). A recent compilation of Mg:Ca ratios in seawater shows that values are not spatially constant but influenced by river waters with low Mg:Ca values (Lebrato et al. 2020). No seawater data is available for the semi-enclosed seas of SE Asia, but the high riverine input could potentially also reduce the Mg:Ca ratios in the Java Sea. Depending on the magnitude of this effect, the reduced Mg:Ca ratios could facilitate the precipitation of calcite in the study area.
The mineralogical composition in the shallow water (< 20 m), photozoan dominated carbonates from Kepulauan Seribu overlaps with the most aragonite rich part of the cool water carbonate field (Fig. 10). However, based on the dominance of reefal material, the sediments can be clearly differentiated from heterozoan dominated cool water carbonates. This might be more difficult for the foramol facies that dominates at the toe of slope and below (Jordan 1998). Wilson (2012) pointed out that oligophotic, humid equatorial carbonates share many similarities with cool water carbonates. However, a marked difference is that HMC components in Panggang have much higher Mg-contents (11.7–14 mol% MgCO3), compared to 4–12 mol% MgCO3 in carbonate sediment of e.g. South Australia (James et al. 2005; O’Connell and James 2015). This difference can be best explained by the different water temperature between both regions. The MgCO3 content in HMC of calcareous marine organism is strongly influenced by the water temperature, with an increase in Mg-content related to higher temperatures (Chave 1954). However, this difference in Mg-content will not be helpful in the distinction of ancient carbonates since HMC is losing Mg from the crystal lattice relatively early during diagenesis (Brachert and Dullo 2000).