3.1. Morphology of Gephyrocapsa during the MB acme
The records of Gephyrocapsa coccolith size and mass from all the studied sites show coherent trends and very similar values during the MB (Text S1 and Fig. S1-S3). From the beginning of the interval, the Gephyrocapsa complex shows low values of average size and mass of the medium Gephyrocapsa fraction, which are coeval with relatively increased, or stable, average values of the same morphometric parameters for the small Gephyrocapsa fraction (Fig. S1- S3). Such a feature is particularly well evidenced at the central part of the MB, between MIS 13 to 11 and MIS 9. From MIS 8 towards the end of the MB, the medium Gephyrocapsa fraction experiences a coeval increase in the values of mean coccolith size and mass, compared to stability of the values of the small Gephyrocapsa fraction (Fig. S1-S3). This morphometrical evolution in our records unequivocally evidences an enhanced presence of mid-sized specimens composing the Gephyrocapsa complex towards the central part of the MB. When morphometrical measurements are used for the calculation of the Morphological Divergence Index (MDI) by Beaufort et al. (2022), a similar pattern arises, with a notable degree of agreement along the different studied latitudes (Fig. 3). In line with the interpretation for this index during the Pleistocene (Beaufort et al., 2022), the lower values of MDI during eccentricity minima in our record (Fig. 3) are considered indicative of a reduction in morphological diversity, that characterizes the enhanced proliferation of Noelaerhabdaceae specimens during the acme episodes of the Pleistocene. The observation of this feature outside the equatorial and tropical latitudes, at Sites 977 and U1385, at mid latitudes, and Site U1314, at high latitudes, is a new critical insight in this study (Fig. 3). It can be stated that Gephyrocapsa coccolithophore assemblages are experiencing a response of enhanced proliferation, which is coeval at every latitude of the Atlantic Ocean and the western Mediterranean region during the MB.
Since the pronounced change in coccolith size and mass described above occurred during the Mid Brunhes Dissolution Interval (MBDI) (Barker et al., 2006), it could be argued that these changes could have been driven, completely or partially, by an intensification of dissolution processes in the water column and/or sediments. However, our data provides two lines of evidence that suggest that calcite dissolution played a negligible role on coccolith morphometries. Firstly, the absence of a correspondence between the morphometrical trends in Gephyrocapsa and the dissolution trend recorded by the planktic foraminiferal fragmentation index of Site U1314 (Fig. 3). Secondly, the coherence between the morphometrical trends and MDI at the different studied sites (Fig. 3 and S1-S3), which are located at significantly different regional depths, latitudes, and environments (Fig. 2).
Contrasting with the homogeneous evolution of size and mass, the evolution of the mean bridge angle of Gephyrocapsa during MB exhibits differences between sites (Fig. 3 and S6). A shift of reduction in values towards the center of the MB is more marked at Sites 925 and 977, in contrast with the more stable evolution of values at other locations (Fig. 3). This indicates, firstly, that the morphology of the Gephyrocapsa specimens proliferating during the acme episode is different through the different latitudes and, secondly, that the evolution of diversity within the Gephyrocapsa complex is also contrasted at the different latitudes and environments (Fig. 3 and S6; Text S1). These results do not appear to indicate the occurrence of a rapid spread of a single mid-sized Gephyrocapsa morphotype, generated in the tropics, a plausible explanation of the morphometrical evidence of the Gephyrocapsa acme outside the low latitudes (Fig. 3). But, more likely, supports a process of global environmental forcing and selection of a wide variety of mid-sized taxa at the different studied locations during the MB. Such observation could challenge the existing model, attached to the effect of reduced seasonality in the equator/tropics, as a single forcing mechanism on Noelaerhabdaceae evolution or adaptation during eccentricity minima (Beaufort et al., 2022). In this sense, earlier proposals pointing to global processes, as changes in nutrient delivery or its budget in the ocean (Flores et al., 2012; Rickaby et al., 2007; Zhang et al., 2021), deserve to be reconsidered.
Overall, there is agreement between our morphological data and previously available records of a different nature that Gephyrocapsa coccolithophores experienced an enhanced proliferation during the acme episode of the MB (Fig. 3) (Beaufort et al., 2022; Flores et al., 2012; Rickaby et al., 2007; Zhang et al., 2021). This observation could be interpreted as a result in the intensification in the function of photosynthesis and production or organic carbon by this group. Nonetheless, all available micropaleontological, geochemical and morphometrical records fail to assess if the higher growth rates were accompanied by a change in the intensity of calcification (i.e., the production of inorganic carbon per individual cell). Assessing the existence of this possible response within the variety of taxa composing the Gephyrocapsa complex at the different latitudes (Fig. 3) is a key point to discern the available ideas about the nature and extent of the processes operating during the acme episode, and to decipher the complete dimension of the role of Gephyrocapsa during the MB.
3.2. Enhanced calcification of the Gephyrocapsa complex
The trends of the three proxies of calcification, the SN Thickness, k value and PIC/POC, evidence higher values during the MB, namely between MIS 13 to 9 interval, and a pronounced decrease at the end of the interval, through the MIS 8 at all sites (Fig. 3). Interestingly, the trends in these calcification proxies are highly similar at each site and between them, when we consider separately each of the traditionally defined species or morphotypes within the Gephyrocapsa complex (Text S2 and Fig. 3). Also, the trends are similar for both the initially differentiated medium and small Gephyrocapsa size classes for morphometrical analysis (Fig. S5) and similar between all locations. These results unequivocally suggest a common physiological change in the cell size normalized calcite content of all the species and/or morphotypes within the Gephyrocapsa complex.
Despite the overall similar trends across sites, some differences in the SN Thickness can be recognized (Fig. 3). The SN Thickness profile at the tropical Site 925 displays a different trend of variability compared to that of other locations (Table S2). In contrast, the k value and the PIC/POC ratio display less variability between sites (i.e., highly similar range of values; Fig. 3) and good correlation between locations (Table S2), suggesting that the derived k value and PIC/POC are more consistent. Thus, we focus our discussion on the changes in calcification on these two proxies.
The coupling between the high concentration of Gephyrocapsa coccoliths in sediments with maximum values in sedimentary tracers to approximate the concentration of calcium carbonate in sediments (Fig. 4) supports the prevailing existent view that the enhanced production of Gephyrocapsa is responsible to increase the global production and accumulation of pelagic carbonate in sediments during the MB (Barker et al., 2006; Flores et al., 2012; Rickaby et al., 2007; Wang et al., 2003). Notably, our results show that, aside from the increased production of carbonate, by the higher coccolithophore proliferation and cellular divisions of Gephyrocapsa (Barker et al., 2006; Flores et al., 2012; Rickaby et al., 2007; Wang et al., 2003), there is an increase in the amount of calcite produced per cell, tracked from the changes in the intensity of calcification of individual coccoliths across multiple Gephyrocapsa species or morphotypes (Fig. 4). This feature has never been described, nor quantified before for the MB, and moreover, it has not been observed as an accompanying factor in the ~ 400 kyr Noelaerhabdaceae acme episodes during the Pleistocene.
Based on our results for the acme of the MB, we consider the existence of a relationship between the production of highly calcified coccoliths during natural bloom episodes, as the MB, and production of coccoliths in a higher quota per cell. If this were the case, it could be argued that either the production of a greater number of coccoliths and/or the higher intensity of calcification may be together the expression of a physiological management aiming to “accommodate” a higher PIC production per cell, in response to certain external/environmental forcing. As a complementary note, the increase in coccolith calcification accompanied by an overall reduction in the average size, towards the proliferation of mid-sized specimens (Figs. 3 and 4), challenges the idea that a decrease in coccolith size is necessarily coupled with a reduction of the degree of calcification (Suchéras-Marx and Henderiks, 2014).
The taxonomical convention has considered all individuals with a value of bridge inclination around 45°, average mid-sized values of coccoliths around 3 µm and a qualitative higher robustness, which globally dominates the sedimentary records during the MB, to be ascribed to the concept of "G. caribbeanica" (Bollmann et al., 1998; Flores et al., 2012; González-Lanchas et al., 2020; González-Lanchas et al., 2021a; González‐Lanchas et al., 2021b; Saavedra‐Pellitero et al., 2017). The universal increase in calcification observed in all the specimens belonging to the Gephyrocapsa complex, together with a variable evolution of diversity and assemblage structure across latitudes (Figs. 3 and 4), suggest that what has been traditionally clumped as unique "Gephyrocapsa caribbeanica" could be the product of the increase in coccolith calcification intensity over a wide variety of morphotypes, globally distributed during the MB. Consequently, we cannot rule out that, instead of just one species, multiple different mid-sized Gephyrocapsa morphotypes proliferated and calcified more intensely during that episode. This would constitute the entire Gephyrocapsa complex which, therefore, could be a synonym of what has been traditionally considered as “G. caribbeanica”.
3.3. Ocean alkalinity as the driver of the Gephyrocapsa acme?
The synchronous morphometrical evidence of the Gephyrocapsa acme and increased intensity of calcification of Gephyrocapsa across the complex in all the latitudes analyzed in this study (Figs. 3 and 4), suggest the existence of an enhancement in the functions of photosynthesis and calcification of this group during the MB and that the origin of both changes was triggered by a common driver. This appears to represent a process of environmental/oceanic forcing, on a global scale, to explain the acme of Gephyrocapsa at the MB.
The main factors controlling the intensity of coccolithophore calcification in natural environments are still under debate (McClelland et al., 2016; Raven and Crawfurd, 2012). It has been put forward that changes in temperature, light, nutrients and seawater carbonate chemistry all have an important role (Gafar et al., 2019; Gafar and Schulz, 2018). Changes in temperature, light and nutrients are more regional and highly dependent on the G/I cyclicity (Rehfeld et al., 2018), which means that they could be apparently ruled out as a possible factor controlling the changes in coccolith calcification intensity at the observed secular scale (Fig. 4). Since the re-equilibrium time of the ocean carbon cycle (i.e., the residence time of carbon in the ocean) is of the magnitude of ~ 100 kyr (Dickens et al., 1995), the changes in seawater carbonate chemistry, together with changes in pH and total alkalinity, are the most suitable candidates for ocean change at the observed scale (Bach et al., 2015; Müller et al., 2021; Müller et al., 2015). There is emerging consensus that HCO3− is the primary inorganic carbon source for calcification, implying that calcification rates increase with increasing amount of HCO3− transferred to the coccolithophore intracellular carbon pool (i.e., the intracellular reservoir and source for calcification) (Bach et al., 2015; Brownlee and Taylor, 2004). Results from culture observations evidence, furthermore, that both photosynthesis and calcification rates increase with enhanced availability of inorganic carbon substrate (Rickaby et al., 2016) and, in particular, as a response of increased concentration of HCO3− (Bach et al., 2013). The existence of a global process of addition of carbon and HCO3− to the ocean carbon pool during the MB, could easily explain the joint stimulation of both functions and the enhanced proliferation of highly calcified Gephyrocapsa specimens in our record (Figs. 3 and 4). Therefore, it is worthwhile to explore the possible changes in global seawater alkalinity during the MB as potential driver.
On the geological perspective, carbon limitation throughout the Pleistocene, due to the long-term Cenozoic CO2 decline, points HCO3− as likely to have played a key role on long-term changes in coccolithophore calcification. This notion has already been proposed in reference studies about the geological changes in coccolithophore calcification, in order to conciliate the apparent decoupling between pCO2 and calcification during the Pleistocene in comparison with its coupling during the late Neogene (Bolton et al., 2016). At orbital timescales, ocean alkalinity is the result of the balance between river input of Ca2+ and carbonate ions (i.e., CO32− and HCO3−) due to terrestrial weathering, and their consumption by carbonate primary producers (i.e., coccolithophores, foraminifers and corals), followed by carbonate burial in the sea floor (Berner and Berner, 2004) (Fig. 1). The MB interval is characterized by the record of exceptional fluvial discharge and intensified chemical weathering, starting from 600 ka (i.e., MIS 15), as discussed from sedimentological and geochemical records from different settings (see Chen et al. (2020) and referenced therehein). The intensified weathering from MIS 15 is conspicupus during the Pleistocene and common to multiple records (Yang et al., 2006; Yao et al., 2010), linked to the anomalous climate conditions and duration of this interglacial (Chen et al., 2020). This fact could have critically triggered an excess of alkalinity in the surface ocean at the beginning of the MB (Fig. 1). The intensification of African and Asian monsoon systems and chemical weathering during the strong interglacials MIS 13, 11 and 9 (Chen et al., 2020; Yao et al., 2010; Yin and Guo, 2008), together with intense denudation and drag during sea level lowstands at glacial MIS 12 and 10 (Chen et al., 2020), would explain a mainted high supply of continental alkalinity during the MB. We consider these facts could have critically contributed to modify the chemical conditions of the surface ocean carbon pool (Fig. 1), thereby facilitating coccolithophorid calcification (Bach et al., 2015), and stimulating the high proliferation of the different highly calcified Gephyrocapsa species or morphotypes in our records (Fig. 4).
While the lack of studies directly documenting the influence of changes in alkalinity on coccolith morphometries limit our interpretation, there are several lines of indirect evidence that support this control. As the global distribution of total alkalinity largely matches that of salinity (Millero et al., 1998), this parameter could be considered roughly equivalent to alkalinity. Several studies have documented a strong and positive correlation between coccolith thickness and calcification with salinity in Noelaerhabdaceae, although the cause for this physiological response is unclear (Bollmann and Herrle, 2007; Bollmann et al., 2009; Green et al., 1998; Linge Johnsen et al., 2019). Building on these studies and our results, we propose that alkalinity, rather than salinity, could represent a major control on coccolith calcification and, plausibly, coccolith morphometries. Future studies will be needed to validate this hypothesis.
3.4. Hypothesis about the double-edged role of Gephyrocapsa during the MB
The increased chemical weathering and transfer of alkalinity from the beginning of the MB would have caused a deepening of the carbonate compensation depth, enhanced preservation of CaCO3 and a major reduction of pCO2 at that orbital scale. This is not observed in the pCO2 (Fig. 3) and global records of reference for the interval (Barker et al., 2006), indicating that, if this addition of alkalinity to the ocean carbon pool had any influence on the acme of Gephyrocapsa, the occurrence of coupled changes in the expected mode of compensation of the ocean carbonate system may have occurred.
The enhanced concentration of Gephyrocapsa coccoliths in sediments (Fig. 4) and Gephyrocapsa nannofossil accumulation rates in our record (NAR; Fig. S7) are consistent between them and with other studies (Beaufort et al., 2022; Flores et al., 2012; Rickaby et al., 2007; Zhang et al., 2021), supporting the notion that this taxon was an important vector of CaCO3 to the seafloor, at a global scale, during the MB. Building on this, the magnitude of the increase in calcification in our results, over 50% in the morphometrical PIC/POC calculation at all the sites (Fig. 4), fits well with the model output of a ~ 50% increase in PIC/POC during the MB, able to promote an intensification in the biological pump sufficient to maintain the constant pCO2 levels at the 400 kyr scale (Barker et al., 2006). The coupled increase in calcification intensity, or enhanced production of calcite per cell (Figs. 3 and 4), provides the last missing piece to support an enhancement of rain ratio and ballast effect promoted by this group during the MB (Barker et al., 2006; Saavedra-Pellitero et al., 2017). Taking all these pieces together, we propose a hypothesis in which the Gephyrocapsa complex may have acted as a sink and source of alkalinity (double-edged role), keeping the alkalinity cycling in the ocean, as detailed below:
“The enhanced chemical weathering from the beginning of the MB would have promoted fertilization, by the enhanced availability of DIC and alkalinity in the surface ocean, possibly complemented by the nutrient addition coupled to the intensified river input (Rickaby et al., 2007; Zhang et al., 2021). This situation would have triggered a process of environmental forcing on morphological modification and/or selection on Gephyrocapsa, promoting the proliferation of a variety of mid-sized and highly calcified taxa within (Figs. 3 and 4). Such response of enhanced production of organic and inorganic carbon represents an overall intensification of both functions of photosynthesis and calcification, acting as a sink for the increased DIC and alkalinity in the ocean at that stage (Fig. 1). The resulting increase in the rain ratio of PIC/POC to the deep ocean (Fig. 1) would have contributed to improve the efficiency of the organic carbon pump during the MB (Barker et al., 2006), now better supported by the role of the thicker and more numerous carbonate particles produced by Gephyrocapsa (Figs. 3 and 4). The high production, ballast and accumulation of organic carbon resulting from these processes, may have enhanced the remineralization of organic carbon and the rates of respiration, contributing to promote dissolution of the CaCO3 above the ocean sea floor (Fig. 1). Such reinforcement in dissolution is in line with the intensity of the MBDI recorded in sedimentary proxies (Barker et al., 2006) and extracted from isotope records (Hoogakker et al., 2006). The higher rates of carbonate dissolution would have returned Ca 2+ and carbonate ions to the ocean carbon pool, reducing the CaCO3 burial and maintaining relatively high levels of seawater alkalinity (Fig. 1). In this sense, the role of the Gephyrocapsa complex during the MB can be considered, as well, to contribute as a source of recicled alkalinity for the system.
This sequence of processes explains a self-sufficient mechanism of recycled and maintenance of alkalinity within the ocean carbon pool and stimulation of the Gephyrocapsa acme during the MB. We consider, nonetheless, that the activation and duration of the double-edged role of Gephyrocapsa with seawater alkalinity necessitates the external control, by continental weathering, to be explained. In tone with this, the overall decreased chemical weathering after interglacial MIS 9 (Chen et al., 2020; Yao et al., 2010), towards the end of the MB, may have triggered a progressive reduction of the ocean surface DIC and alkalinity, by the decreased continental input. These new environmental conditions may have exerted a different forcing on Gephyrocapsa, as observed by the shift towards increased MDI values and reduced calcification of specimens, common to all latitudes from MIS 8 onwards (Figs. 3 and 4). The reduction in the carbonate consumption and ballast effect by the end of acme conditions would have even amplified the preservation of inorganic carbon in the sea floor, balancing the existent inputs of alkalinity from continental weathering through carbonate deposition in the deep sea, after a few kyr (Fig. 1). In other words, this represents the recovery of the typical conditions of carbonate compensation towards the end of the interval ”
Finally, the increased amplitude of the 100 kyr cycle in atmospheric CO2 after the MIS 11/12 transition, (~ 430 ka, the so called MB event, sensu stricto) may relate to a diminished reservoir of ocean DIC and alkalinity, leading to a greater G/I amplitude response to the same forcing (Rickaby, 2022). If correct, an eventual loss of alkalinity from the ocean, by an “exacerbated” enhanced proliferated and production of highly calcified Gephyrocapsa at the beginning of the MB, could have contributed to an eventual reduction in the ocean DIC under a certain critical threshold, before the MIS 11/12 transition. These thoughts may give rise to further hypothesis and studies approaching a possible influence of the seawater chemical conditions triggered by the Gephyrocapsa acme within the causes of the increased amplitude of the G/I cyclicity after 400 ka.