The benthic response to the climatic perturbations
As it was expected the general distribution of the analysed components is highly dependent on the facies type. However, they also show a clear temporal pattern related to both the time and probably the amplitude of the carbon isotope excursions.
Bryozoans
The distribution of bryozoans is very heterogeneous (Fig. 7). On the one hand, they are more common in reefs and reef debris than in other facies (Fig. 4), on the other hand, they are far more abundant during the isotope excursions, especially during the Lau and Ireviken excursions (Fig. 7). Also, within the reefs, the distribution of bryozoans is very uneven. Reefs that are mainly formed by corals and stromatoporoids possessed only low abundances of bryozoans and vice versa. In contrast, when reefs comprise a high proportion of microbial carbonate bryozoans tend to be more abundant. This is especially the case in bryostromatolites (Claussen et al. 2022) (Fig. 5h, 15e, f), where bryozoans and microbes were reef-building and therefore rock-forming (Fig. 7, 8).
A possible explanation could be, that the combination of microbial growth in combination with thriving bryozoans is associated with extreme conditions (Piller and Harzhauser 2023). Similar modern associations are known from Joulters Cays (Cuffey at al. 1977, 1979; Cuffey and Fonda 1979), Coorong lagoon (Palinska et al. 1999), and the Netherlands (Harrison et al. 2022). They occur in for corals hostile environments, that experience, e.g., seasonal emergence, hypersalinity, seasonal changing temperatures with hot conditions (40°C in the water body), fast changing salinities due to tidal influences, temporary euxinic and/or toxic conditions (e.g. heavy metals), and high sulfate contents (cf. with Cuffey at al. 1977, 1979; Cuffey and Fonda 1979, Palinska et al. 1999, Harrison et al. 2022). Also, for fossil counterparts this is assumed (Piller and Harzhauser 2023). In reefs where stromatoporoids and corals dominate, bryozoans usually occur in low abundances as they have a low potential to compete with common reef dwellers (Jackson, 1983; McKinney and Jackson, 1989). This is why they usually prefer to live in cryptic habitats (Cuffey 1977; Kobluk 1981, 1988), which is also a common phenomenon in the Silurian of Gotland.
The high abundance of bryozoans in some reefs on Gotland, especially in the bryostromatolites, is probably due to their high tolerance to a wide range of environmental conditions and, additionally, due to the fact that some species were highly opportunistic (Palinska et al. 1999, Piller and Harzhauser 2023). The frequently observed alternation of microbial cements with encrusting bryozoans indicates rapidly changing conditions (Fig. 5h, 15e, f). Since the bryostromatolites grew in very shallow water, even seasonal emergence would be a possible explanation, which is supported by the frequent occurrence of Palaeomicrocodium as it was observed before by Claussen et al. (2022, Fig. 11). in several specimens Recent bryozoans have been observed to die off seasonally, be overgrown by microbes, and then immediately reoccupy the same substrate and area, as it was observed in Coorong Lagoon (Palinska et al. 1999) and in The Netherlands (Harrison et al. 2022). This is supported by the work of Ernst and Königshof (2008) as they interpreted some Palaeozoic bryozoans as pioneer species. However, the perennial corals and stromatoporoids would not be able to withstand such harsh conditions.
An important question to be solved is, why were Bryozoans at isotope excursions more abundant? This can possibly be explained by seasonality. Although Gotland was located in the tropics, evenly banding in corals (Fig. 15b, d), stromatoporoids (Fig. 15a, c) and bryostromatolites (Fig. 15e, f) were discovered. It is unknown what let to seasonal changes, possible are seasonal monsoons, or changes in aridity. The post-sedimentary, early diagenetic gypsum crystals indicate increased salinities during phases of extremely shallow, hypersaline water on the platforms, probably related to subaerial emergence. Very salt-rich, heavy water percolates downwards through the rock and precipitates gypsum crystals in cavities, that later have been transformed to pseudomorphs. Increased nutrient levels would also be plausible, which would promote the growth of bryozoans and microbes, while oligotrophic conditions would tend to favour the growth of corals and stromatoporoids. An increased phosphate content, which is an essential nutrient, might be indicated by the frequency of phosphate components (see below).
Corals and stromatoporoids
Similar to bryozoans, corals and stromatoporoids are also unevenly distributed on Gotland (Fig. 7). They are common in reefs and reef debris, occur in almost all facies settings, but are only very rarely found in the oncolitic shelf limestones of the Eke Beds or in the distal shelf (Fig. 4). When stromatoporoids are present, they are in most cases more abundant than corals and are often associated with low diversity (Fig. 7). Presumably both needed oligotrophic conditions due to their assumed photosynthetic symbionts (Zapalski 2014, Zapalski and Berkowski 2019; Król et al. 2024). Therefore, they seem to exclude themselves from bryozoans (Fig,7; as the PCA in Fig. 16 also shows). Corals and stromatoporoids often occur at the end of the isotope excursions and during the relatively weak Linde excursion (Fig. 7). Usually, microbes are rare when stromatoporoids are abundant (Fig. 7, 8). Stromatoporoids seem to predominate in reefs. If they are present in the samples, then they occur often in larger quantities than corals. The dominance of stromatoporoids may be related to the environmental conditions of the reefs in which they occur. Possibly the driving factors could be water energy or enhanced salinity, as stromatoporoids dominated in the extremely shallow biostromes of Kuppen (middle Hemse Group, Linde excursion) (Fig. 6d) (e.g., Kershaw 1987a, b, 1981, 2023; Keeling and Kershaw 1994; Kershaw and Keeling 1994; Sandström and Kershaw 2008) and in Galgberget (Tofta Formation, Ireviken excursion). Among the corals, it is remarkable that the highest tabulate values are mainly caused by Coenites sp.. This is probably because they were easier to be fragmented than most other tabulates species and thus produced more sediment.
Microbial carbonates
Microbial carbonates show both, a facies as well as a temporal dependence (Fig. 4, 8). They occur mainly in shallow facies areas (reefs, oncolitic back reefs, oncolitic shelf) (Fig. 4). In addition, they are only found in significant quantities during periods of high isotope values (with the exception of Furilden (upper Slite Beds)) (Fig. 8). In the periods with low isotope values, they are only very subordinate even in very shallow facies areas. Although they are common during excursions, they appear to be very insignificant in almost all samples of the comparably weak Linde excursion (Fig. 8).
During the extremely strong Lau excursion, microbial carbonates appear as oncoids (Fig. 5c, 6c) even in the shelf areas in front of the reefs, sometimes rock-forming (Eke Formation), whereas they usually only produced thin crusts around bioclasts.
Microbes need nutrients and are less demanding in terms of oxygen availability and temperature. Due to the great species diversity of recent microbes, there is hardly any habitat that is not colonised by microbes - although most of them do not calcify. Similar to bryozoans, they are opportunists (Kershaw et al. 2006). If the nutrient content in the water increases, they can grow rapidly and dominate the entire ecosystem (e.g., harmful algal blooms in the Great Barrier Reef (Bell 1992), and in South Florida (Paul et al. 2005)).
Microbes calcify extracellularly, i.e., in contact with seawater. While porostromate microbial carbonates represent calcified bacterial colonies, probably cyanobacteria, spongiostromate microbial carbonates are poorly understood. It can be assumed that they are calcified consortia of various non-colony-forming bacteria, such as biofilms (Riding and Sharma 1998). A change in seawater chemistry, especially carbonate saturation, should therefore be directly reflected in the calcification rate of the microbes. Other organisms (metazoans) calcify intracellularly, which means that they are less dependent on the carbonate saturation of the seawater.
Possible scenarios for altering the calcium carbonate saturation of the sea water would be:
-
(a) Carbonate weathering during sea-level fall. This would result in the erosion of freshly deposited limestones on the exposed shelves. At the beginning of the isotope excursion there are indications of short-term sea-level falls, but the rocks of the isotope excursion formed mostly during transgressive phases or highstands.
-
(b) Carbonate input by aeolian dust input (Kozłowski 2015), which requires an arid climate.
-
(c) Increase in water temperature on the platform. This would lead to higher CO2 release and lower CO2 uptake into the water body. There is, however, no indication for increased temperatures during times of the isotope excursion. In contrast, the increased d18O values would rather argue for lowered temperatures (see discussion in Munnecke et al. 2010).
-
(d) Change in ocean circulation from an anti-estuarine to estuarine circulation. A change from an upwelling situation, with upwelling of colder, CO2-rich water to a downwelling situation, as proposed by Bickert et al. (1997), would result in an enhanced carbonate saturation during times of downwelling (i.e. times of isotope excursions).
Theoretically, the high abundance of microbial limestones can be explained by the reduction of grazers. This, however, seems unlikely because (a) there is no evidence if the extinction event at the beginning of the Lau Event has affected gastropod species (Samtleben et al. 2000; Calner 2005), and (b) remains of grazing organisms are present in thin sections from Gotland (e.g., Fig. 6b). The presence of grazers was reported for the isotope excursions during the Silurian period by Jarochowska et al. (2014)d tzel et al. (2023).
Phosphatic components
Phosphatic components were unexpectedly common in many thin sections and therefore have been quantified (Fig. 10, 12). Bryozoan species with phosphatic layers (Fig. 11c – f) or pearls (Fig. 9e – h, 11a – b) are known since a long time (Oakley, 1934; Conti and Serpagli, 1988; Ma et al., 2014). It is suggested that both were formed during the life of the bryozoan colonies, possibly during digestion (Oakley 1934, 1966; Conti and Serpagli 1988; Ma et al. 2014), indicated by clearly visible grains as nuclei (Fig. 9e – h, 11 a – b). As their occurrence in the Silurian is time-specific (Oakley 1934, 1966), their formation may have been triggered by environmental influences (Oakley 1934, 1966; Conti and Serpagli 1988; Ma et al. 2014). However, several species were observed in which this had not yet been observed so-far.
Since many of the phosphatic components are produced by bryozoans, it is not surprising that samples with many bryozoans also contain many phosphatic pearls. Consequently, periods with isotope excursions are characterised by a high abundance of bryozoans as well as phosphatic pearls and linings (Fig. 10, 12). Therefore, the bryostromatolites in particular contain many of these bryozoan pearls. In addition to this, the bryozoan genus Ptilodictya, which occurs throughout the entire sequence and forms phosphatic linings, also shows significantly thicker linings during the isotope excursions (Fig. 11d) than at other times (Fig. 11c).
Interestingly, phosphatic brachiopods also show a clear "preference" for times of isotope excursions, suggesting that these times facilitated the mineralisation of phosphatic shells or other biogenic structures. This hypothesis is supported by the fact that there are also further occurrences of high abundances in phosphatic brachiopods during the Lau excursion in the Barrandian, which was in mid-latitudes at that time (Mergl et al. 2018). This argues for another reason for the increase in phosphatic components than just elevated abundance of bryozoans.
Phosphate is an important and limiting nutrient (Redfield 1934,1958), which means that it may indicate increased nutrient levels in the seawater. A plausible explanation would be an increase in dissolved phosphate in seawater, although the reason for this is unclear. Processes that influence the phosphate content in seawater are:
-
(a) Deposition of organic-rich sediments in oxic ocean water (Noffke 2014, Sinha et al. 2021).
-
(b) Iron fixation as pyrite in the sediment. Iron fixation of phosphate is one of the major sinks of phosphorous (Föllmi 1996) in oceans and soils, directly followed by deposition of organic-bound phosphates (Dijkstra et al. 2014). Under anoxic conditions phosphate will be released from sediments (Noffke 2014, Sinha et al. 2021). When H2S is produced in the anoxic waters by sulphur-reducing-bacteria, it reacts with the iron to form pyrite (FeS), which removes iron from this cycle (Noffke 2014, Sinha et al. 2021).
-
(c) Terrestrial input of phosphate or phosphate-containing minerals by rivers or by dust can increase the contents (Dijkstra et al. 2014).
-
(d) and finally upwelling can bring P-enriched deep water from the oxygen minimum zones to the shelf (Föllmi 1996).
A model for the Silurian isotope excursions is discussed in the chapter 'Implications for the Silurian climate'. In this chapter, a potential phosphate cycle for the Silurian of Gotland is proposed with regard to possible sources and sinks.
Habitat preferences of reef dwellers
Based on the distribution of the components microbes, bryozoans, stromatoporoids and corals (Fig. 4) in the samples from Gotland, their hypothetical preferred habitat is determined (Fig. 17). Therefore, assumptions were made on the environmental parameters of ‘nutrient’ supply, ‘water energy’, ‘salinity’, and ‘seasonality’. For the parameter light, conclusions can potentially be drawn from the facies setting. However, this only would indicate whether light could have been available but gives no information on whether the organisms actually needed light for photosynthesis. This is under debate and therefore unclear for tabulate corals and stromatoporoids (Wood 1993).
Nutrient availability
In reefs with high numbers of corals and stromatopores, bryozoans and microbes are rare. Corals and stromatopores are compared to modern reef builders in terms of nutrient availability because they favoured oligotrophic to mesotrophic conditions (Nestor 1984; Wood 1993; Kershaw 1998; Whalen et al. 2002; Zatoń et al. 2015). When stromatoporoids occur in reefs on Gotland, they often dominate the fauna. Other reef inhabitants, especially heterotrophic organisms and microbes, occur with low abundance, which supports the assumption of oligotrophic conditions. In the example of the biostrome at the locality Kuppen (Fig. 6d), corals occur only subordinately. This leads to the assumption that stromatoporoids could live well under oligotrophic conditions, which is also supported by Kershaw (1998), whereas corals possibly favoured mesotrophic conditions. On Gotland, bryozoans and microbes occur often in combination. Bryozoans as heterotrophic organisms rely on food supply, but are also found in normal Silurian reefs, which are assumed to be mesotrophic. However, there they occur in low abundances, presumably because they were unable to assert themselves against the competition, especially from the tabulate corals.
On the other hand, increased formation of microbial mats is often associated with eutrophication (Wood 1993; Whalen et al. 2002). This means that both high abundance of microbes and bryozoans can be regarded as indicators for enhanced nutrient availability. Therefore, bryozoans and microbes were categorised from mesotrophic to eutrophic (Fig. 17), while the microbes were considered to be somewhat more resistant. Presumably, nutrient availability is the most influential factor for the distribution of organisms in the Silurian of Gotland. This correlation is also evident in the PCA (Fig. 16) carried out and seems to confirm this assumption.
Water energy
Due to the distribution of organisms in different facies areas, it is possible to conclude which habitats are favoured with respect to water energy. In facies areas with high energy, microbes generally form oncoids around bioclasts (e.g. Hummelbosholm) (Fig. 6c) (cf. Cherns 1983). To withstand in these environments, the microbes which formed oncoids (mainly porostromates) formed crusts during their lifetime. In contrast, spongiostromate microbes can also be found in reefs with high amounts of fine allochthonous micrites, such as in the Halysites biostromes (Lower Visby Formation, ?Ireviken Excursion), and the Heliolites biostromes of Blåhäll (Halla Formation, Mulde Excursion) (Fig. 6g). In addition to this, cauliflower-like oncoids in limestones with micritic matrix from back reef areas in Bodudd (Hemse/Eke transition, Lau Excursion), Ronehamn (Eke Formation, Lau Excursion) (Fig. 5c), and Bankvät can be attributed to calm water settings (cf. Cherns 1983). As fine allochthonous micrite and mud can only be deposited with low water energy this can be taken as evidence that microbes also lived in very calm environments (cf. Cherns 1983). Due to this, it can be assumed that they were insensitive regarding water energy. However, it is likely that they would not be preserved at high water energies.
The bryozoans, however, seemed to be more adapted to environments with lower water energies. For feeding they are not necessarily dependent on high water energy, as they can actively generate water currents with their cilia. In general, bryozoans have a low tolerance to high water energies, as most of the colonies, for example fenestrates, or cyclostomates would break easily. However, there are some species in the Silurian that were also adapted to moving water. These had flexible joints (Matsutrypa lindstroemi, as well as several species from the genus Ptilodictya) and were therefore flexible.
Stromatopores are rarely found in deposits formed with low water energy, however, they seemed to favour shallow marine and more turbulent water conditions (Kershaw 1990, 1998). They occurred rarely in distal shelf reefs. Due to their smooth surface, it is likely that they were more adapted to higher energy facies areas. For instance, they dominate the fauna in the widespread biostromes in Kuppen (Fig. 6d), which are assumed to be formed in highly energetic environments.
Some coral species break easily and are therefore only adapted to low-energy habitats. One example of this is the important Silurian sediment former Coenites sp.; while others tend to form shallow colonies, like Heliolites sp., Favosites sp., and therefore seem to be adapted to higher water energies which is confirmed by the fact that they are common in patch reefs. However, two exceptions of low-energy tabulate coral reefs can be found on Gotland. In Ireviken, a biostromal reef almost completely formed by halysitid corals formed in deeper-water settings in the Lower Visby Formation (Zapalski and Berkowski 2019), and in Blåhäll (Halla Formation, Mulde Excursion) (Fig. 6g) another deep-water biostrome formed, here dominated by heliolitid corals (Calner et al. 2000). Possibly these reefs formed at a pycnocline, as in modern oceans nutrients accumulate at the pycnocline (Whitney et al. 2013, Fripiat et al. 2021).
An adaptation to different water energies can also be recognised in the large patch reefs of the Högklint Formation which formed in a shallowing-upward sequence. The reef community changed from tabulates as main reef builder in the lower (i.e. deeper) part of the reef to stromatoporoids in the upper part. The topmost part of the reefs is made up of algal limestones (Kershaw 1987b; Watts and Riding 2000).
Salinity
Most sediments from Gotland contain fragments of organisms which indicate a fully marine environment, with common marine salinity. Aberrant salinities can only be inferred from the increased abundance of salinity-tolerant organisms (e.g. microbes, ostracods, gastropods). It is also suggested that stromatoporoids also had the ability to tolerate slightly elevated salinities (Kershaw 1987a, 2023). Therefore, the finding of pseudomorphs after gypsum especially in reefs might appear surprising because the precipitation of gypsum crystals from sea-water requires hypersaline conditions that - except for microbes - no organisms would tolerate. The pseudomorphs, however, are clearly post-depositional as they formed exclusively in cavities (Kershaw 1987a, Färber et al. 2014, Claussen et al. 2022). They probably formed directly after reef formation in a time of exposure or at least non-deposition by hypersaline brines percolating downward through the rocks (Färber et al. 2014). Such unconformities are common on Gotland, some of them show even signs of subaerial exposure/karstification, for example at the Högklint-Tofta boundary (Fig. 5a is showing gravitational cements; typical for vadose diagenetic environments, i.e. subaerial exposure), the Slite-Halla boundary (Laufeld and Martinsson 1981, Calner 2002, Calner et al. 2004a), in Kuppen (Hemse Group; Keeling and Kershaw 1994; Sandström and Kershaw 2002, 2008; Calner et al. 2004b) and especially at the Hemse-Eke boundary (Cherns 1982, 1983). Nevertheless, an evaporative gypsum formation would require an arid climate. Therefore, species-poor communities (composed of often monospecific communities of rhynchonellid brachiopods, bryozoans, ostracods, bivalves), that are found in very shallow facies areas, especially in the east of Gotland, are more likely caused by slightly increased salinities in an overall arid climate rather than by reduced salinities.
Another indicator of exposure might be the problematic crust-like Palaeomicrocodium. Antoshkina (2014) assumed that it was formed possibly under subaerial conditions, indicated by measurements of trace elements (Mn, Co, Ni, Zn, Cu, and V) in these crusts. On Gotland Palaeomicrocodium can often be found in bryostromatolites. It can be assumed that the reefs formed in very shallow water, possibly directly at the sea surface. Claussen et al. (2022) showed that Paleomicrocodium formed repeatedly during reef growth, as it frequently alternated with bryozoan crusts (cf. Claussen et al. 2022, Fig. 6).
Seasonality
Evidence of seasonality can be recognised in stromatoporoids and corals, which are supposed to live for several years to decades, as both of them often show regular growth bands. This banding was also observed by Manten (1971, p.434) and attributed by him to possibly formed due to “(seasonal?) fluctuations”. Even in relatively deep-water milieus there is evidence of seasonality visible in growth banding of halysitid corals (Nohl and Munnecke 2019). However, too extreme seasonality can be considered as unlikely for both corals and stromatoporoids, as the tolerance of both groups would probably be exceeded (too warm, cf. modern coral bleaching, too strongly fluctuating nutrient supply, strong clay input due to possible monsoonal climate, etc.). It is likely that stromatoporoids and corals required comparably stable conditions.
Regular banding can also be recognised in bryostromatolites (Claussen et al. 2022). There in many areas microbes and bryozoans were intergrowing and produced multiple alternating thin layers (Fig. 5h, 15e, f), which might be interpreted as seasonal. In contrast to the perennial corals and stromatoporoids, microbes and bryozoans can still form reef-like structures even when they die seasonally as shown, e.g. in modern bryostromatolites (Harrison et al. 2022).
Implications for the Silurian climate
In the past thirty years various models were developed that aimed to explain the strong isotope excursions during the Silurian as well as the related environmental changes (e.g. Jeppsson 1990; Wenzel and Joachimski 1997; Bickert et al. 1997; Cramer and Saltzman 2005; Kozłowski and Sobień 2012; Bowman et al. 2019; Frýda et al. 2021; Stolfus et al. 2023). Many of these models deal with the spreading of anoxic and oxic zones in the oceans during the Silurian and their potential causes (e.g. McLaughlin et al. 2012; Vandenbroucke et al. 2015; Emsbo et al. 2017; Bowman et al. 2019, Hartke et al. 2021, Frýda et al. 2021; Stolfus et al. 2023). However, they are in many aspects contradictory and vary, e.g., regarding depth, intensity and timing of the anoxic water bodies. The main reason for the discrepancies is the lack of direct witnesses from the deep ocean, which is completely subducted, which means that all climatic or oceanographic models rely on information from shelf deposits, and this also applies to the present manuscript. In the following a model (Fig. 18) is proposed, which is based on the models of Jeppsson (1990), Bickert et al. (1997) and Cramer and Saltzman (2005), but tries to explain also the results of the present study. We suggest two distinct climatic stages, the first formed during times of isotope excursions, while the second stage corresponds to the time interval between isotope excursions.
First stage: Times of isotope excursions
The occurrence of gypsum is mostly restricted to times of isotope excursions (Fig. 13; Kershaw 1987a) indicating an arid climate. During times of arid climate, the water on the shallow shelf becomes slightly enriched in salinity which in turn results in a downwelling of this slightly heavier water (anti-estuarine circulation). This scenario was first proposed by Bickert et al. (1997).
Increased amounts of phosphatic components indicate that the phosphate concentration in the ocean, at least on shelves, was elevated. Continental input, which could be a strong source of phosphate (Föllmi 1996), is unlikely as the carbonates were very pure in these time intervals, with only very little amount of clastic input such as clay. Besides, elevated terrestrial input would have inhibited the growth of benthic fauna as many of the organisms were filter-feeders. Another potential source would be the dissolution of iron phosphates (vivianite) and re-mineralization of organic matter from sea floor deposits, while preventing the phosphate from being redeposited with iron. However, this kind of phosphate release in elevated quantities is only possible under anoxic conditions (Noffke 2014, Sinha et al. 2021) as the iron phosphates dissolves within the sediment (Föllmi 1996; Noffke 2014, Sinha et al. 2021). Under oxic conditions, the sedimentary surface will be oxic (Noffke 2014). This will prevent the release of phosphate, as phosphate will be fixed in the aerobic layers by iron. So, widespread anoxia might offer an explanation for enhanced P content in the water which facilitated the growth of organisms with phosphatic shells, pearls or linings.
In numerous studies, it was suggested that global ocean anoxic events (OAEs) happened during the major Silurian isotope excursions (e.g. McLaughlin et al. 2012; Vandenbroucke et al. 2015; Emsbo et al. 2017; Smolarek et al. 2017; Bowman et al. 2019, Hartke et al. 2021, Frýda et al. 2021). Almost all of them assume increased anoxia or extended areas of anoxia during the excursions. In contrast to this, del Rey et al. (2023) assumed on the basis of U-isotopy that an extended anoxic phase existed immediately prior to the excursion and lasted only until the beginning of the excursion. A direct proof is difficult as many black shales might have been deposited in the deep ocean and thus are not preserved. The theory of OAEs during the excursions can thus only be indirectly proven, e.g., by geochemical measurements of redox sensitive trace elements (Emsbo et al. 2017, Frýda et al. 2021) as well as by other not redox-sensitive indicators of ocean anoxia such as d34S, 205Tl, Mn (Bowman et al. 2019) and 87Sr86Sr (Emsbo et al. 2017).
The relatively wide distribution of graptolite shales in the Silurian (Lüning et al. 2000a, b, Loydell et al. 2013) proves that the outer shelf areas and probably also the continental slope were often or even permanently anoxic and thus the boundary between oxic and anoxic water was probably relatively shallow (Wilde et al. 1991). Interestingly, the distribution of black shales does not seem to be strictly related to the d13C excursions. Although some black shales are reported from times of isotope excursions (e.g., Smolarek et al. 2017), the very widespread and often very organic-rich black shales in the Llandovery such as the “hot shales“ in N-Africa (Lüning et al. 2000a, b) are not mirrored in the d13C curve, indicating that the isotope excursions are not significantly controlled by the formation of organic-rich sediments, at least not by the ones deposited on the shelves or slopes. It is, however, very likely that there was a strong stratification of oxic and anoxic water masses, presumably induced by density differences due to salinity (Jeppsson 1990; Frýda et al. 2021).
In times of global anoxia already at shallow depth a strong fractionation in d13C between surface and deep water is expected because most of the 12C-rich organic matter produced by phytoplanktic organisms sinks to the sea floor and is not recycled by bacteria. Such a situation can be observed in the modern Black Sea (Fry et al. 1991). An anti-estuarine circulation would bring surface water with strongly elevated d13C values to the shallow platforms (Bickert et al. 1997), which could explain the heavy d13C values measured in shallow-water carbonates.
The interface between oxia and anoxia was presumably shifted away from the shelves towards the open ocean due to the downwelling of the dense water. This leads to a peripheral mixing effect that transports phosphate-enriched deep water back to the ocean surface. With the anti-estuarine circulation, the phosphate-enriched water is transported back into the shallow platform areas. In the initial phase, the phosphorus content in the sea water was low. Due to the presumably initially oligotrophic to slightly mesotrophic conditions, extensive reefs developed, which consisted mainly of corals and stromatoporoids. However, the longer this situation lasted, the stronger the anoxia became. As a result, more phosphorus was transported into the shallow seas, which greatly increased benthic productivity. This can be seen in an increase in spongiostromate microbial mats (interpreted as biofilms), bryozoan crusts, phosphatic components and subsequently in the formation of bryostromatolites. After some time, the global anoxia in the deep-water area declined and phosphate became fixed again with iron in the sediment. As a result, nutrients were withdrawn from the ocean waters and corals and stromatoporoids were able to dominate again.
The sedimentary records from this period let suggest that benthic life flourished during periods of isotope excursions, while plankton was suppressed (Jeppsson 1990; Jeppsson et al. 1995; Urbanek and Teller 1997; Jeppsson and Aldridge 2000; Jeppsson and Calner 2003; Gelsthorpe 2004; Stricanne et al. 2006; Cooper et al. 2014). It is unclear what exactly caused the absence of a plankton blooms during isotope excursions. It is possible that other essential nutrients, such as nitrate, were limited and suppressed planktonic productivity (cf. Saltzman 2005).
This could explain the observed intervals of reef formation during isotope excursions as well as the occurrence of extensive reef growth with decreasing isotope ratios at the end of isotope excursions. Possibly the same process took place also in other regions during the Silurian. This could potentially be an explanation for a contemporaneous enrichment of phosphatic brachiopods in the Prague Basin (Mergl et al. 2018).
Stage 2: Times with no isotope excursion
In the time intervals between the isotope excursions, the deposition of marls is elevated. The cay minerals, presumably derived by riverine input, reduced the purity of the limestones and indicate that the climate became more humid. It inhibited the growth of benthic organisms such as microbes, bryozoans and other Silurian reef builders, which explains the decline in carbonate deposition during these times (Jeppsson 1990, Bickert et al.1997).
There is no direct evidence as to whether the deep ocean was oxic or anoxic at this time. Investigations in several studies (Kiipli 2004; Gambacorta et al. 2019) of various markers, such as TOC, U/Th and trance elements concentrations indicate that the deep ocean was oxic. In such a case, it must be assumed that there was an oxygen Minimum Zone (OMZ) at relatively shallow water depths - analogous to today's conditions and as also suggested by Jeppsson (1990), and Cramer and Saltzman (2005). The OMZ is caused by the aerobic decomposition of organic material in the water column by microbes and zooplankton. During this process, O2 and biomass (e.g. glucose) are consumed and CO2 and H2O produced, which results in a strong decline of O2 in the OMZ which ranges in depth in modern oceans from 150 to 1000m (Paulmier and Ruiz-Pino 2009). Due to the lower atmospheric O2 content in the atmosphere in the Silurian (Haxen et al. 2023) the OMZ was probably shallower than today, and probably reached at least the outer parts of the shelf (Fig. 18).
The dissolved phosphate, derived from organic matter, is fixed by iron in the oxic ocean water and deposited at the sea floor. As a result of the increased freshwater input from land due to the humid climate, an estuarine circulation was developed, as suggested by Bickert et al. (1997). This circulation is driven by density changes (freshwater is less dense than ocean water) and it results in an upwelling of deeper waters.
From the riverine input nutrients were transported into shelf areas and caused enhanced plankton productivity (Jeppsson 1990, Hoshiba and Yamanaka 2013). Due to the high amount of planktic biomass that was produced and sank down, the oxygen content in the AOM decreased and eventually became anoxic. This resulted in the deposition of organic-rich deposits (black shales) on the outer shelves and probably also on the upper slopes (Jeppsson 1990, Bickert et al. 1997, Cramer and Saltzman 2005). Through the continuous recycling of biomass in the open ocean (similar to today’s situation), the 13C values remain close to normal-marine values. While slightly positive values probably existed in the surface water, it may have been slightly negative in the OMZ.
These described processes result in the accumulation of increasing amounts of phosphorus in deep-sea sediments. They dissolve when a strong anoxia zone is formed again in the deep water.