Marine deoxygenation is one consequence of ongoing global change and is likely to be felt most acutely in the expansion of already low oxygen environments in the coastal and open ocean (Keeling et al., 2009;Levin, 2017;Schmidtko et al., 2017;Breitburg et al., 2018).
Expansion of marine hypoxia and anoxia alter the global cycling of key nutrients and redox sensitive metals (Gruber, 2008;DeVries et al., 2012), and drive ecological shifts in both benthic and pelagic communities (Levin, 2003;Stramma et al., 2010;Stramma et al., 2012;Horak et al., 2016). Despite growing recognition of the importance of low oxygen environments and the ramifications of their expansion, the short duration (decades at most) of ocean oxygen timeseries fundamentally limits the timescales at which oxygen and associated environmental variability can be studied.
Oxygen minimum zones (OMZs) are subsurface (~100-2000 m deep), generally open ocean features, where dissolved oxygen is low enough to impact biological and chemical cycles.
OMZs have changed in step with global climate in the past, especially during periods of rapid warming such as deglaciations (i.e., van Geen et al., 2003;Nameroff et al., 2004;Martinez and Robinson, 2010;Moffitt et al., 2015). However, models still fall short of reconstructing recent deoxygenation, and a greater understanding of long-term drivers of deoxygenation is required to improve future projections (Oschlies et al., 2018). Moreover, while marine deoxygenation is ongoing in the modern ocean (Schmidtko et al., 2017), some oxygen variability may be attributable to decadal scale variability rather than long-term climate change (Deutsch et al., 2014). With a better pre-modern baseline, such cyclical versus secular trends would be easier to tease apart. Thus, there is a particular need to develop and apply additional environmental and oxygenation proxies within the OMZ.
Great strides have been made toward developing new paleo-oxygenation proxies to constrain past OMZ dynamics. One notable example is the refinement and application of the I/Ca proxy for dissolved oxygen using shells of planktic foraminifera, a group of calcifying protists (Zhou et al., 2014;Lu et al., 2016;Hoogakker et al., 2018;Lu et al., 2020). Other trace element to calcium (TE/Ca) ratios in the shells of benthic and planktic foraminifera also have tremendous potential for recording the physical and chemical environment of the OMZ. For example, Mn/Ca ratios in benthic foraminifera are a promising proxy for bottom water hypoxia (Groeneveld and Filipsson, 2013;McKay et al., 2015;Brinkmann et al., 2021).
We explore the proxy potential of several conventional trace elements within the shells of the planktic foraminifer Globorotaloides hexagonus. This species has been widely associated with low oxygen waters (Fairbanks et al., 1982;Ortiz et al., 1995;Birch et al., 2013;Rippert et al., 2016), and was recently recovered live from discrete depth plankton tows through the Eastern Tropical North Pacific (ETNP) OMZ (Davis et al., 2021). Furthermore, G. hexagonus has a rich fossil history, with shells occurring in sediments dating back 14 Mya (Kennett and Srinivasan, 1983), making it an ideal candidate for reconstructing paleo-OMZs. We explore potential controls on Mg/Ca, Mn/Ca, Zn/Ca, and Sr/Ca ratios, which are routinely measured in calcite and are resolvable in foraminifera shells with the use of standard Inductively Couple Plasma Mass Spectrometry (ICP-MS), and in some cases even Inductively Coupled Plasma Optical Emission Spectrometry, analyses after the use of defined cleaning protocols (Barker et al., 2003;Marr et al., 2013a;Fritz-Endres and Fehrenbacher, 2021). As a result, these analyses are accessible to a wide range of researchers and could prove especially valuable for reconstructing past OMZ environments. Moreover, as TE/Ca in calcite may reflect diverse drivers ranging from temperature to oxygenation to source water (see Katz et al., 2010 for a review) and can be analyzed simultaneously, investigating multiple TE/Ca ratios could make multiproxy and multivariate records more accessible. Here we use Laser Ablation ICP-MS for two reasons. The first is that G. hexagonus shells are both rare and light in many sediments (Davis et al., 2021(Davis et al., , 2023)), meaning a sufficient sample for solution measurements may not always be possible. The second is in anticipation of the potential need to use high-resolution intra-shell analyses to analytically distinguish between primary and altered calcite.
All foraminiferal shells used in this study were collected by an opening-closing zooplankton net system, the MOCNESS (Multiple Opening/Closing Net and Environmental Sensing System; Wiebe et al., 1985), taken onboard the R/V Sikuliaq in January-February of 2017 in the ETNP (21° N, 117° W). Net tows consisted of both depth-stratified vertical profiles from the surface to 1000 m depth in 25-100 m intervals, as well as horizontal sequences of tows through low oxygen features (8 or 9 nets per tow, 222µm mesh) (Wishner et al., 2018(Wishner et al., , 2020(Wishner et al., , 2021)). A suite of environmental data was collected by MOCNESS sensors during each tow, including depth, temperature, salinity, dissolved oxygen, in situ fluorescence, and volume filtered through each net (Figure 1). Each net encompassed a range of environmental data, since an individual net was open for about 10 -20 minutes and sampled a depth stratum (in the case of vertical tows) from 25 to 100 m thick. Since it is unknown exactly where within the sampled stratum a foraminifer was collected, the environmental data range for the specific net from which an individual was analyzed is shown by the horizontal lines in the following figures. Samples from each net were preserved shipboard in 4% sodium-borate buffered formalin and seawater and then stored in the lab until foraminifera were removed in 2017-2018. All foraminifera were picked from tow material as described by Davis et al. (2021). The shells of live collected (with cytoplasm present) G. hexagonus were isolated from tow material. Because planktic foraminifera are not believed to be highly mobile, it was assumed that much or all of the adult whorl, calcified within the depth and environmental range sampled by the net in which it was captured. We stress that the use of the final or "F" chamber here is not completely analogous to its use in fossil material. In living foraminifera, the final chamber refers to the most recently calcified chamber at the time of capture, and not necessarily the final chamber to calcify prior to reproduction and death. As the relatively large mesh size (222 µm) used excludes juveniles (measured foraminifera ranged from 297 to 631 μm in length), all individuals are inferred to be adults.
Preparation for trace elemental analyses involved an oxidative cleaning step to remove residual organic matter from shells, following that described by (Barker et al., 2003) with some modification. Briefly, shells were bathed individually in a 1:1 mixture of NaOH and H2O2 for 10 minutes at 60 °C and then triple rinsed in deionized water to remove reagent. While foraminiferal trace element cleaning frequently includes sonication, this step was excluded due to the fragility of G. hexagonus shells and removal of clays and other infilling from tow collected specimens is unnecessary. Shells were kept whole to facilitate laser ablation ICP-MS and because recent findings demonstrate that fragmentation may artificially decrease some trace element ratios in shells (Fritz-Endres and Fehrenbacher, 2021). Shells were mounted on carbon conductive tape on a glass slide.
Shells were analyzed by a laser ablation system (Photon Machines 193 nm ArF laser with an ANU HelEx dual-volume laser ablation cell) coupled to an iCAP quadrupole ICP-MS in the College of Earth, Ocean, and Atmospheric Sciences at Oregon State University following previously established protocols (Fehrenbacher et al., 2015). Shells were ablated using a 65 μm spot size, a 4 Hz rep rate, and a fluence of 0.85 J cm -2 . Analytes presented here include 25 Mg, 44 Ca, 55 Mn, 66 Zn, and 88 Sr, with ablations of NIST 610 and NIST 612 run between every ~10 samples. Profiles include the data from 1 s after the start of ablation to when the laser broke through the chamber wall. Data from laser ablation analyses were processed using the LATools Software (Branson et al., 2019), and are presented as mean TE/Ca ratios either for a chamber (averaged if more than one ablation was possible within a chamber), or for a shell (presented as an average of every chamber ablated) (Supplementary Data). Ablation times ranged from 3 to 79 seconds, and ablation profiles less than 5 seconds (8 profiles) were excluded from these analyses.
Reproducibility was assessed by the mean difference in duplicate ablations through the same chambers. The difference between ablations was 0.27 mmol/mol for Mg/Ca, 0.03 mmol/mol for Sr/Ca, 5.15 μmol/mol for Mn/Ca, and 90.86 μmol/mol for Zn/Ca.
All statistics were carried out in R. In the case of regression analyses, non-linear least squares regressions were used and are reported here along with standard error. All correlations were carried out using a Pearson method, with a Holm's correction for multiple hypothesis testing and both correlation coefficient and p-value reported here.
Laser ablation ICP-MS profiles were resolved from 184 individual foraminifera shells spanning 20 different nets over a range of depth, oxygen, and temperature conditions within the ETNP OMZ (Figure 1).
The average Mg/Ca in individual G. hexagonus shells ranged from 1.02 to 5.85 mmol/mol. Mg/Ca ratios are positively correlated with both the average values of dissolved oxygen (corr = 0.31, p < 0.001) and temperature (corr = 0.31, p < 0.001) recorded in the corresponding nets (Figure S1). The latter ranged from 6.1 to 22.0 °C. Due to the bias in vertical distribution of G. hexagonus, with most individuals living well below the thermocline (Davis et al., 2021), only three shells were analyzed from the warmest temperature. Moreover, because this net integrated across the thermocline, it is possible that individuals collected were living preferentially at the low end of that temperature range (deeper), rather than the mean. Despite this potential bias, the mean temperature, is < 2 °C greater than the minimum in this sample, and thus is unlikely to be a major source of uncertainty in developing a quantitative relationship. A Pearson correlation with Holm's method, used to account for multiple hypotheses, indicates that individual shell Mg/Ca is well correlated with temperature, despite a wide range of interindividual variability.
Regressing individual G. hexagonus shell data against the mean temperature associated with each net and adopting an exponential fit, as has been found most suitable in other species (Lea et al., 1999;Mashiotta et al., 1999;Elderfield and Ganssen, 2000;Dekens et al., 2002;Anand et al., 2003;McConnell and Thunell, 2005;Cléroux et al., 2008;Sadekov et al., 2009;Livsey et al., 2020a), results in a relationship between temperature and Mg/Ca that can be described as:
with T as temperature in degrees Celsius, reported with standard error.
The same analysis can be run using Mg/Ca averages from each net, producing a regression within error of Eq. 1 (Mg/Ca = 1.40 (+-0.1) * e^(0.04 +-0.01)(T)). Given the paucity of datapoints at the warmest temperature, we also regressed individual shell Mg/Ca against temperature with results from the warmest net removed, resulting in an equation again within error of Eq 1 ((Mg/Ca = 1.78 (+-0.3) * e^(0.02 +-0.02)(T))) and virtually indistinguishable at lower temperatures (Fig. 2). We note that over a small temperature range the relationship could be equally well described by a linear relationship. Given the similarity between the three approaches, we selected the first, which is significant despite very high inter-individual variability, considers all available data, and retains an exponential relationship as has been found
for other species. these trends are robust to potential bias in the shallowest sample (Figure S2).
Despite, high interindividual variability, the relationship between Mn/Ca and O2 can be described by either a linear regression (r 2 = 0.10, p-value < 0.001) or an exponential curve (r 2 = 0.12, p-value < 0.001). Similar regressions for Zn/Ca are significant (p-values 0f 0.004 and 0.002 respectively), but with very low r 2 values (0.04 and 0.05). In contrast the relationship between Sr/Ca and O2 can be described by either a linear regression (r 2 = 0.24, p-value < 0.001) or a logarithmic curve (r 2 = 0.46, p-value < 0.001) (Figure 3).
The use of laser ablation to analyze individual chambers allows for a comparison of chamber-to-chamber differences in TE/Ca, or trends in trace element incorporation through ontogeny. A Kruskal-Wallis test demonstrates that each of the targeted elemental ratios change through ontogeny (p < 0.001 in all cases). Significantly lower Mn/Ca, Zn/Ca, and Mg/Ca ratios are observed in younger relative to older chambers. By contrast, higher Sr/Ca and ratios are observed in younger chambers (Figure 4).
Calibration of the Mg/Ca paleothermometer for G. hexagonus
Most variables of interest (temperature, oxygenation, depth) are highly colinear in this dataset, with all TE/Ca displaying robust correlations with multiple environmental drivers (Figure S1). Therefore, the following discussion will center around connections between environmental parameters and elemental ratios as supported by previous work rather than attempting to statistically deconvolve potential drivers. Given the rich literature on speciesspecific sensitivity of foraminiferal shells to calcification temperature (e.g., Nürnberg et al., 1996;Lea et al., 1999;Mashiotta et al., 1999;Elderfield and Ganssen, 2000;Lea et al., 2002;Anand et al., 2003), we present Mg/Ca results in the context of calcification temperature.
However, in at least some species, salinity (Lea et al., 1999;Kısakürek et al., 2008;Dueñas-Bohórquez et al., 2009;Mathien-Blard and Bassinot, 2009;Hönisch et al., 2013;Gray et al., 2018), and carbonate chemistry (Lea et al., 1999;Russell et al., 2004;Gray et al., 2018;Gray and Evans, 2019) have secondary influences on shell Mg/Ca. Here, salinity can be disregarded as a driver due to the narrow range of salinity across these samples (34.0-34.6), but the same assumption cannot be made about carbonate chemistry. While direct measurements of carbonate chemistry are not available, we would expect pH to be highly correlated with dissolved oxygen (Paulmier et al., 2011; Figure S2) and thus with temperature (Figure 1). Based on work in other planktic species, Mg/Ca would be expected to increase as pH or carbonate ion ([CO3 2-]) availability decreases (Russell et al., 2004;Allen et al., 2016;Evans et al., 2016;Gray and Evans, 2019). A potential pH/[CO3 2-] effect would then act counter to temperature, increasing Mg incorporation at depth. Thus, it is possible that the sensitivity of Mg/Ca to temperature alone in G. hexagonus is higher than predicted by the empirical relationship derived here. This would need to be tested through future work in additional locations or laboratory culture. One further caveat is the degree of interindividual variability observed. Interindividual variability has been observed in culture (Davis et al., 2017)
The robust relationships found between Mn/Ca and Zn/Ca in the shells of G. hexagonus and in situ dissolved oxygen point to a potential environmental control. There is an existing theoretical framework for such a relationship in Mn/Ca. As Mn is readily oxidized to MnO2, Mn 2+ , the cation assumed to substitute for Ca 2+ in the calcite lattice, should be more available in low oxygen water masses. Thus, Mn 2+ substitutions for Ca 2+ should occur more frequently and Mn/Ca should increase in shells formed at lower oxygen conditions (Barras et al., 2018;van Dijk et al., 2020). While there are other controls on Mn and Mn 2+ availability in natural systems, in the ETNP dissolved Mn shows a distinct peak specifically associated with low dissolved oxygen (Bolster et al., 2022; Figure 5). There is also evidence for an oxygenation control on the incorporation of Mn/Ca into benthic foraminiferal calcite (Munsel et al., 2010;Groeneveld and Filipsson, 2013;Koho et al., 2015;McKay et al., 2015;Ní Fhlaithearta et al., 2018;van Dijk et al., 2020). However, increased shell Mn/Ca has also been related to oxygenation or advection of water across the oxycline in live-caught planktic foraminfiera (Steinhardt et al., 2014;Davis et al., 2020). The directionality of the relationship found here is consistent with a dissolved oxygen control on Mn/Ca incorporation into G. hexagonus calcite (Figure 3A).
As multiple parameters covary with depth in this dataset, it is necessary to consider alternate drivers responsible for the higher Mn/Ca ratios, chief among these being the carbonate system (Figure 1; Figure S1&S2). Laboratory culture of the planktic species Orbulina universa demonstrated a negative correlation between shell Mn/Ca and [CO3 2-] ( Allen et al., 2016) and a positive correlation with DIC (Holland et al., 2017). Similarly, cultures of the benthic hyaline species Amphistigina gibbosa as well as the phylogenetically distant porcelaneous Sorites marginalis show increasing shell Mn/Ca values with increasing DIC (van Dijk et al., 2020).
These results are consistent with the expectation that increasing DIC (decreasing [CO3 2-]) would co-occur with decreasing oxygenation in the OMZ (Paulmier et al., 2011). Thus, the carbonate system and dissolved oxygen could act independently or in concert to drive higher Mn/Ca values in nets collected closer to the core of the OMZ. While more work in modern samples will be required to tease apart these drivers, results indicate that shell Mn/Ca may record the intensity of the OMZ and the co-occurring carbon maximum within the habitat of G. hexagonus.
We find that lower Zn/Ca ratios are found in shells from nets with higher oxygen (Figure 3B) and higher temperature. No previous work has directly compared Zn/Ca in foraminiferal shells to oxygenation, while temperature has a negligible effect on the Zn/Ca of benthic foraminifera (Marchitto et al., 2000;Titelboim et al., 2021). The carbonate system, however, has been widely implicated as a driver of Zn/Ca ratios (Marchitto et al., 2000;van Dijk et al., 2016;van Dijk et al., 2017), and Zn/Ca increases with increasing DIC (decreasing [CO3 2-]) in multiple species of cultured benthic foraminifera (van Dijk et al., 2016;van Dijk et al., 2017). This has been attributed to pH dependent speciation of Zn, and incorporation into foraminiferal calcite of Zn 2+ (van Dijk et al., 2016). Zn also displays a nutrient-like profile, increasing with depth through the OMZ (Conway and John, 2015;Janssen and Cullen, 2015) (Figure 5). Thus, both increased availability of [Zn] and [Zn 2+ ] may act in conjunction to increase Zn/Ca ratios with increasing depth and OMZ intensity.
Results suggest that Zn/Ca and Mn/Ca in G. hexagonus can be related to macroenvironmental drivers, via concentrations of dissolved ions in ambient seawater. This is further supported by observations of an increase in dissolved Mn and Zn at OMZ depths in the ETNP, in agreement with our foraminiferal record (Bolster et al., 2022; Figure 5). The variability in Zn/Ca and Mn/Ca ratios of G. hexagonus shells and dissolved Mn and Zn are also all higher at OMZ depths (Figure 5; Supplemental Table 2).
Other non-spinose foraminifera such as Neogloboquadrina dutertrei and Globorotalia truncatulanoides, may calcify within an organic aggregate microhabitat, as identified in part by high Ba/Ca values (Fehrenbacher et al., 2018;Richey et al., 2022). Shell Ba/Ca is also quite high in G. hexagonus (mean 39 µmol/mol across all ablation profiles done in these tow samples) and therefore, G. hexagonus may be additionally influenced by the redox conditions inside particle microenvironments. Particle microenvironments can support anaerobic respiration in a low oxygen water column when oxygen is further drawn down by respiration inside the particle (Alldredge and Cohen, 1987;Alldredge and Silver, 1988;Shanks and Reeder, 1993;Bianchi et al., 2018). This further oxygen depletion should result in higher Mn 2+ . These same particulate microenvironments might have a complex and non-linear influence on Zn. When reduced sulfate is available in particulate microenvironments, Zn 2+ precipitates into ZnS which can have implications for Zn cycling in the OMZ (Janssen and Cullen, 2015). While respiration within a particulate microenvironment could further decrease pH, increasing Zn 2+ , sulfate reduction at sufficiently low oxygen levels could decrease the availability of [Zn 2+ ]. Such sulfate-reducing metabolisms may occur within particles in the ETNP OMZ (Carolan et al., 2015), and could be partly responsible for the large range and variance of Zn/Ca ratios found at the lowest oxygen levels (Figure 5; Table S2).
The strongest correlation of any analyte with oxygen was found for Sr/Ca (Figure 3C).
As foraminiferal Sr/Ca ratios have not yet been evaluated with respect to oxygenation, we will first consider alternative controls. From relatively early in the history of foraminiferal trace element analyses, the Sr/Ca of planktic foraminifera shells has been linked to salinity and temperature (Lea et al., 1999;Elderfield et al., 2002;Cléroux et al., 2008;Kısakürek et al., 2008). Salinity differences are minimal in this dataset and therefore an unlikely driver of shell Sr/Ca. Previous studies demonstrate only a weak (Lea et al., 1999) or insignificant (Russell et al., 2004;Kısakürek et al., 2008) relationship between Sr/Ca and temperature in planktic foraminifera. However, a strong correlation between temperature and dissolved oxygen (corr = 0.92; Figure S1) make the two parameters nearly impossible to disentangle, and temperature cannot be discounted as an important influence on G. hexagonus Sr/Ca ratios. Given this uncertainly, other drivers need be considered.
The Sr/Ca ratios of calcareous foraminifera have also been linked to carbonate chemistry, similar to Mn/Ca and Zn/Ca ratios. However, findings are inconsistent. Multiple species of planktic foraminifera show an increase in shell Sr/Ca ratios with an increase in pH and/or [CO3 2-] (decreasing DIC) (Lea et al., 1999;Russell et al., 2004;Dueñas-Bohórquez et al., 2009;Holland et al., 2017). By contrast, some benthic foraminifera show a decrease in Sr/Ca with decreasing DIC (increasing pH and/or [CO3 2-]) (Keul et al., 2017;van Dijk et al., 2017). Still other benthic (Dissard et al., 2010;Raitzsch et al., 2010) and planktic species (Lea et al., 1999;Russell et al., 2004;Kısakürek et al., 2008) are apparently insensitive to the carbonate system as a Sr/Ca driver.
One related consideration for the incorporation of Sr into foraminiferal calcite is growth rate. Higher growth rate has been widely associated with increased foraminiferal Sr/Ca ratios especially in planktic species (Elderfield et al., 2002;Kısakürek et al., 2008;Holland et al., 2017;Geerken et al., 2022), and explicitly linked with an apparent influence of temperature and carbonate chemistry (Lea et al., 1999;Russell et al., 2004). The shell mass of cultured planktic foraminifera tends to increase with temperature (Lombard et al., 2009), [CO3 2-]/pH (Spero et al., 1997;Lea et al., 1999;Bijma et al., 2002;Russell et al., 2004;Lombard et al., 2010;Manno et al., 2012), and O2 (Kuroyanagi et al., 2013), which suggests all three parameters could increase growth. Further support for growth rate as a driver of Sr/Ca incorporation in G. hexagonus comes from individual shells. Assuming initial calcification of each chamber takes a similar amount of time, a larger chamber would calcify faster than a smaller chamber and therefore might have higher Sr/Ca. This is supported by chamber-specific ablations which demonstrate that younger, larger chambers have higher Sr/Ca compared to older, smaller chambers (Figure 4).
We note that significant decreases in Sr/Ca in younger chambers are not confined to the final chamber alone, indicating that the presence of this trend is unlikely to be driven entirely by incomplete calcification of the youngest chamber.
While we cannot disentangle the potential impacts of temperature, pH, and oxygen on Sr/Ca from this dataset, all three variables would be expected to result in slower shell growth rates within the low-oxygen, low-pH, and low-temperature core of the OMZ. As a result, Sr/Ca presents itself as a potentially useful proxy; a decrease in growth rate, and therefore lower Sr/Ca, may be associated with more intensive OMZ conditions rather than an isolated environmental driver. The use of Sr/Ca as a proxy for OMZ intensity would allow for an assessment of habitat within the OMZ, and therefore the depth habitat of an individual; a shell with higher Sr/Ca likely calcified farther from the core of the OMZ than one with lower Sr/Ca. At the population level, changes in Sr/Ca could point to changing strength of the OMZ, with lower Sr/Ca values indicative of more intense OMZ habitat being available to G. hexagonus.
One undeniable challenge in the reconstruction of OMZ environments from foraminiferal shell TE/Ca is preservation. It is well known that trace elemental ratios such as Mg/Ca can be altered by either calcite overgrowths and recrystallization (Boyle, 1983;Pena et al., 2005;Pena et al., 2008) or dissolution (Lorens et al., 1977;Dekens et al., 2002;Fehrenbacher et al., 2006;Johnstone et al., 2011;Regenberg et al., 2014;Branson et al., 2015). The use of elements such as Mn and Zn as proxies is especially complicated by postmortem modification of shell chemistry, with Mn alteration frequently noted (Boyle, 1983;Pena et al., 2005;Pena et al., 2008). A combination of chemical and mechanical cleaning may remove postmortem contaminants but can also remove the primary signal of elements including Mn and Zn (Fritz-Endres and Fehrenbacher, 2021). Distinguishing between primary and diagenetic signals via high resolution analytical technics, such as Secondary Ion Mass Spectrometry (SIMS) or Laser Ablation ICP-MS is one possibility (Bice et al., 2005;Marr et al., 2013b). In fact, our rationale for employing Laser Ablation ICP-MS here was in part in recognition that future analyses may need to employ this high-resolution technique to both exclude altered shell. However, the delicacy of thin and porous G. hexagonus shells present distinct challenges for differentiation of primary and diagenetic signals by microanalytical techniques. For example, altered calcite might be more commonly found on exposed surfaces including the inner and outer surface and pores and mixing of altered and relatively pristine regions would occur in any ablation pit that included a pore. To some extent, this would always be an issue in using LA-ICP-MS to identify altered zones, but one that would be amplified in a species with relatively large pores and thin walls. These challenges may be partially overcome by future characterization of the natural TE/Ca heterogeneity within G. hexagonus shells in addition to continued evaluation of cleaning protocols and careful site selection.
Although diagenesis may complicate some applications of Mn/Ca and Zn/Ca as proxies in G. hexagonus shells, the immediate outlook for other proxies is brighter. While Mg/Ca diagenesis requires careful consideration, a substantial body of research has already been amassed on cleaning and sample selection with an eye specifically to Mg/Ca preservation in foraminifera (e.g., Barker et al., 2003;Marr et al., 2013;Fritz-Endres and Fehrenbacher, 2021).
The upside is that temperature interpretations from Mg/Ca should be feasible at all but the most altered sites. Moreover, unlike Mn/Ca and Zn/Ca, Sr/Ca is not highly susceptible to postdepositional alteration (Lorens et al., 1977), although ratios may decrease if shells experience dissolution (McCorkle et al., 1995;Edgar et al., 2015). Thus, Mg/Ca and Sr/Ca ratios of G.
hexagonus shells from sediments show particular promise for reconstructing past OMZs. It is our hope that these conventional trace elements in planktic foraminifera, along with other archives of pelagic calcite such as fish otoliths (Limburg et al., 2015;Limburg and Casini, 2018;Altenritter and Walther, 2019;Cavole et al., 2023) may be used to improve the spatial and temporal resolution of records from low oxygen pelagic environments.
Trace metal analysis of individual G. hexagonus shells demonstrates several ways in which the species could be useful for reconstructing OMZ environments beyond the observational record.
As in other species of planktic foraminifera, Mg/Ca ratios of G. hexagonus shells can serve as a proxy for temperature once a species-specific equation is applied. Ratios of Mn/Ca and Zn/Ca both increase with decreasing dissolved oxygen, as would be expected if carbonate chemistry and/or oxygen were dominant controls. An especially strong correlation is observed between low Sr/Ca ratios and more intense OMZ environments. We hypothesize that this effect could be linked to growth rate and that Sr/Ca is a potential proxy for OMZ intensity, especially given relative robustness of Sr/Ca ratios to diagenesis. Thus Sr/Ca and more tentatively Mn/Ca and Zn/Ca could be useful in reconstructing OMZ intensity.