Contents lists available at ScienceDirect

Chemical Geology

journal homepage: www.elsevier.com/locate/chemgeo

LA-MC-ICP-MS study of boron isotopes in individual planktonicforaminifera: A novel approach to obtain seasonal variability patterns

Dennis Mayka,b,c,*, Jan Fietzkec, Eleni Anagnostouc, Adina Paytand

a Department of Earth Sciences, University of Cambridge, Cambridge, Cambridgeshire, UKb British Antarctic Survey, Cambridge, Cambridgeshire, UKcGEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Schleswig-Holstein, GermanydUniversity of California Santa Cruz, Santa Cruz, CA, USA

A R T I C L E I N F O

Editor: Michael E. Boettcher

Keywords:Boron isotopeOrbulina universaLA-MC-ICP-MSSeasonal variabilityforaminiferaPaleo-reconstructionpHGeochemistry

A B S T R A C T

Boron isotope (δ11B) analysis using bulk foraminifera samples is a widely used method to reconstruct paleo seawater pH conditions. Although, these analyses exhibit high analytical precision, short term information is lostdue to the pooling of tests with distinct and diverse boron isotope signatures resulting in average values for thetime interval encompassed in the sample. Here we present and assess the analysis of δ11B of individual for-aminifera by means of Laser Ablation Multi-Collector Inductively Coupled Plasma Mass Spectrometry (LA-MC-ICP-MS) to obtain seasonal variability patterns and to test the limits of precision of LA-MC-ICP-MS on theplanktonic foraminifera Orbulina universa. The results show that relative seasonal differences (of ∼11‰) can becaptured from either uncleaned or cleaned individual O. universa tests with an average precision of± 2.9‰ (2SE). The δ11B variability among foraminifera representing the same season is on average 7.4‰ (2 SD) irre-spective of cleaning state. With our approach, analyses on oxidatively cleaned O. universa do not require the useof a matrix matched standard to obtain B isotope values in the range of those expected for solution multi-specimen analyses. Our results are useful for considering the potential spread caused by foraminifera vital effectsand for obtaining information of seasonal ranges of pH and possible bias related to seasonality hidden withinconventional solution based δ11B analyses.

1. Introduction

The isotopic composition of boron in marine carbonate is a usefultool for reconstructing past ocean pH, an important variable needed forcalculating paleo-pCO2 (e.g., Hemming and Hanson, 1992; Hönischet al., 2009; Foster and Rae, 2016). Boron has two stable isotopes 10B(19.9% abundance) and 11B (80.1% abundance). Due to its approx.14–20 Ma residence time in seawater (Spivack and Edmond, 1987;Lemarchand et al., 2000, 2002), boron is considered to be homo-geneously distributed throughout the oceans with an isotopic ratio of39.61 ± 0.04‰ (2 SE). Boron in seawater is predominantly present intwo forms; tetrahedrally coordinated borate [B(OH)4−] and trigonallycoordinated boric acid [B(OH)3]. The relative abundance of these spe-cies varies with water pH (Zeebe and Wolf-Gladrow, 2001). The ratio of10B and 11B in borate ion [B(OH)4−], reported in δ-notation relative toNIST-SRM951, is given by the relationship:

= − −

−

− − −

pH sw pK Bδ B sw δ B borate

δ B sw α Bδ B borate α B* log[

11 11

11 ( 11 1000( 1))]

where pK*B describes the dissociation equilibrium between boric acidand borate ion in seawater (Dickson, 1990), δ11Bsw is the seawaterboron isotopic composition, αB is the equilibrium constant for boronisotope fractionation between boric acid and borate ion in seawater(1.0272, Klochko et al., 2006; Nir et al., 2015). Assuming borate ion ispreferentially incorporated into carbonate (e.g., Hemming and Hanson,1992), the δ11B composition of marine carbonates can be used to cal-culate the pH of the solution from which the mineral precipitated.However, δ11B ratios in marine biogenic carbonates, specifically for-aminifera, are often different from that of inorganic carbonates formingunder the same pH conditions. This is because of processes related tobiological calcification - often referred to as “vital effect” (e.g., Hönischet al., 2003; Zeebe et al., 2003; Noireaux et al., 2015; Martínez-Botíet al., 2015; Henehan et al., 2016; Farmer et al., 2019). Therefore,species-specific relationships are developed between seawater borate

https://doi.org/10.1016/j.chemgeo.2019.119351Received 14 March 2019; Received in revised form 22 October 2019; Accepted 27 October 2019

⁎ Corresponding author at: Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge, Cambridgeshire, CB2 3EQ, United Kingdom.E-mail address: dm807@cam.ac.uk (D. Mayk).

Chemical Geology 531 (2020) 119351

Available online 28 October 20190009-2541/ © 2019 Elsevier B.V. All rights reserved.

T

and foraminiferal δ11B. For O. universa this relationship is described byHenehan et al., 2016 as:

δ11Bborate = (δ11Bforaminifera + (2.49±9.01)) / (1.02±0.50)

Boron isotope analyses on MC-ICP-MS of foraminifera are typicallycarried out on multi-specimen samples, comprised of several individualforaminifera shells (tests) that are cracked and cleaned together (e.g.,Sanyal et al., 2001; Foster, 2008; Rae et al., 2011). Although this pro-cedure produces precise results, it requires pooling of 100–200 in-dividuals (for planktonic foraminifera) and boron purification throughcolumn chemistry. When multiple individuals are combined to obtainone value, temporal resolution is lost in the process. Recent advances inthe methodology of isotope analysis in foraminifera make it possible toanalyse the isotopic composition of individual foraminifera (e.g., Fordet al., 2015; Pracht et al., 2018; Sadekov et al., 2019; Standish et al.,2019) opening a new field of research questions that can be addressed,such as investigation of the variability within and among individualforaminifera. This may provide new insight into short timescale varia-bility (representing the life span of an individual specimen) and help tounderstand the process of isotope incorporation during calcification.Analyses of single foraminifera have been primarily focused on isotopesand elemental ratios of major constituents of the carbonate tests (e.g.,oxygen; Metcalfe et al., 2019, and Mg/Ca; Eggins et al., 2003, 2004,Ford et al., 2015). Marine carbonates often contain 1–100 ppm boron(Vengosh et al., 1991), therefore, δ11B analysis in single foraminifera ischallenging; only a few studies reporting δ11B in single foraminiferahave been published (e.g., Kaczmarek et al., 2015; Sadekov et al., 2019;Standish et al., 2019). In this study, we present a novel approach toanalyse δ11B in individual foraminifera by LA-MC-ICP-MS, with the goalto decipher intra-individual and seasonal variability of δ11B recorded inforaminifera, and test the method’s applicability for palaeoceano-graphic research.

2. Material and procedures

2.1. Environmental setting

Orbulina universa (size fraction: 500–600 μm, est. shell weight:40–60 μg (Allen et al., 2011)), were picked from sinking particulatematter collected using McLean Mark VII-W sediment traps equippedwith a 0.5 m2 funnel opening deployed in the Santa Barbara Basin, CA(Thunell, 1998). Sinking particulate matter including the foraminiferawas collected every ∼15 days. Sample bottles were buffered and poi-soned with sodium azide solution prior to deployment. To investigateseasonal variability samples were assigned one of two groups. The fallgroup (32 individuals) consisted of samples collected from October toDecember 2013, the spring group (19 individuals) included samplescollected from March to May 2014. The mean surface seawater pH(total scale) at the collection site was 8.02, with a range of up to 0.07during the growing period of the foraminifera, representing 15 daysprior to collection start date and through the 15 days of collection. Weassumed that each O. universa either was fully grown or started to growat the time of the trap deployment (see Supplement). The short-term pHvariability within the region of the sediment traps deployment can beup to 0.7 pH (Hofmann and Washburn, 2018), with a total high fre-quency variability of 4.7‰ among sediment traps (Supplemental Data).Observed seasonal pH values at the site translate to a mean of 16.6‰δ11Bborate.

2.2. Cleaning methods

To test the effect of cleaning on the spread of the δ11B data amonggroups, two commonly used cleaning methods were applied sequen-tially on the same samples that were previously analysed without anycleaning; labelled as cleaning-I and cleaning-II, respectively.

2.2.1. Cleaning-IA set of specimens were retrieved from the ablation cell (spring

n = 19, fall n = 12). Each test was individually cleaned using acleaning procedure modified after Glock et al., 2016; 2019. Thiscleaning treatment aimed at removing impurities from the foraminifera,including trapped particles and labile organics.

Each foraminifera specimen was bathed in ∼500 μL of ethanol(ROTIPURAN ≥ 99,8%, p.a.) and ultrasonicated for 20 s. This was re-peated three times, after which the procedure was repeated for anotherthree times using Milli-Q water to remove any ethanol residue from theforaminifera. Following, 350 μL of oxidative solution (0.3% H2O2 in0.1 M NaOH) was added to each foraminifer, and bathed at 80 °C for40 min, ensuring that air bubbles were released frequently. Each for-aminifer was then removed from its solution using a 63 μm sieve andMilli-Q water. Finally, each specimen was rinsed with ethanol beforebeing transferred to a new adhesive pad for re-analysis.

2.2.2. Cleaning-IIAs many of the foraminifera cleaned using the above procedure

(Cleaning-I) as possible were removed from the adhesive pad (springn = 10, fall n = 9) following the second LA-MC-ICP-MS analysis. Thetests were then individually cleaned following a different oxidativecleaning approach (Barker et al., 2003; Rae et al., 2011; Henehan et al.,2013).

The foraminifera were carefully cracked open under a microscopeusing a fine needle. 250 μL of oxidative solution (1% H2O2 in 0.1 MNH4OH) was added to each foraminifer. Note: This is the same amountconventionally used to clean ∼100–200 O. universa individuals fromsediment trap and tow samples (Henehan et al., 2016), therefore itrepresents at least 100x more oxidative solution per individual for-aminifer. Here too a water bath at 80 °C was employed, but for a totaltime of 15 min, with each sample ultrasonicated for a few seconds every5 min to release trapped bubbles. Bubbling activity (related to CO2

production from organic matter oxidation) dissipated after the first5 min, which indicated the near completion of the oxidative reaction.Following, the foraminifera tests were rinsed three times with Milli-Qwater. Once dried, the foraminifera were individually transferred ontoanother adhesive pad, with the internal part of the test facing up for thethird analysis.

2.3. LA-MC-ICP-MS analyses

Measurements were carried out at GEOMAR, Helmholtz Centre ofOcean Research in Kiel, Germany using an ESI New Wave ResearchUP193FX excimer laser, operating at 193 nm, connected to a ThermoFisher (former VG) Axiom MC-ICP-MS. Both 10B and 11B were collectedby electron multipliers and 12C by a Faraday cup, using low mass re-solution (∼400 Res) for maximum sensitivity. Before analyses, sampleswere rinsed during sieving (125 μm sieve) with unbuffered deionizedwater, dried in an oven at 40 °C and mounted on a self-adhesive padwithout any further treatment. The order in which the foraminiferawere analysed was randomized to avoid bias, and samples were as-signed to one of two groups of equal size (A and B). Prior to analysesdifferent spot sizes and laser fluences were tested to maximize signalintensity and to limit the ablation area, so that multiple measurementson the foraminifera tests could be performed. The optimal setting wasdrilling spots of 75 μm with a fluence of ∼1.8 J cm−2 and an ablationrate of 10 Hz (laser shots per second). Most foraminifer tests were stableenough to withstand a constant ablation of> 300 shots on a singlepoint using this configuration. Therefore, an ablation-time of 30 s foreach ablation spot was chosen. Uncleaned and cleaned-I samples wereablated from the outside, while, the cleaned-II samples were crackedopen and ablated from the inside out. The carrier gas was a 1:1 argon/helium mixture at a total flow rate of 1.64 l min-1. The ICP-MS wastuned for best boron signal intensity while maintaining plasma ro-bustness, identified using the “normalized Ar index” (NAI) approach

D. Mayk, et al. Chemical Geology 531 (2020) 119351

2

(Fietzke and Frische, 2016) throughout all measurement sessions(Table 1).

To quantify the internal variability within each group and the re-lative difference between groups, NIST610 soda-lime glass standardwas used, applying the empirically determined offset of -0.55‰ be-tween NIST610 and boric acid standard NBS951 (Fietzke et al., 2010).To realize the full potential of our method, the analytical strategy forcleaned-II samples was modified. Carbonate standards JCp-1 and JCt-1were included into the measurement procedure of the cleaned-II sam-ples to confirm accuracy of our δ11B LA-MC-ICP-MS data. JCt-1, a giantclam calcite standard was measured alongside the foraminifer samplesas an unknown, resulting in a mean δ11B value of 15.9‰±1.8 (2 SD,n = 37). This value is in good accord with published bulk JCt-1 data(16.3‰±0.6 (2SD), Gutjahr et al., 2014). This, in addition to theabsence of interferences (Fig. S4), we considered as evidence that thereis no need for a matrix matched standard to obtain accurate δ11B usingour method, further supported by our longterm (>3 yrs) mean δ11B forJCp-1 of 24.0 ± 1.0‰ (2 SD) which is consistent with reported values(24.3 ± 0.4 (2 SD) Gutjahr et al., 2014).

Because the low NAI sessions did not include carbonate standards,we avoid comparing the δ11B composition of samples analysed usingdifferent NAI (0.5 or 3.5). Instead we focus on the variability withineach analyses group. Nevertheless, it should be noted that our mea-surements at NAI = 0.5 yield JCp-1 values of 22.1 ± 1.5‰ (2 SD),indicating only a minor offset from solution analyses (Gutjahr et al.,2014) (see Supplement for more details).

An optimized standard-sample bracketing measurement sequencewas used to correct for possible instrumental mass bias. It consisted of500 μm long line scans on NIST610, JCp-1 and JCt-1 (JCp-1 and JCt-1were only used with cleaning-II samples), each followed by ∼50 s ofbackground signal collection (no laser firing), and one selected point oneach test of the foraminifera from group A (typically ∼25 individuals),followed by another set of standard line ablation and sampling of thespecimen in group B, concluded by another set of standard measure-ments. After every spot ablation the background signal was collectedfor 30 s. While 11B background signal intensity was typically< 1000

cps, the target 11B signal intensity for ablation data collection has been∼100,000 cps. To achieve the later the laser spot size for JCp-1 was setto 35 μm and for JCt-1 to 50 μm.

The sequence was repeated four times, resulting in four spots foreach foraminifer from each of the groups. Thus, standards wheremeasured after every ∼25 samples to ensure consistent and stablemeasurement conditions. No additional drift corrections were deemednecessary as no systematic drifts during the measurements were ob-served.

2.4. Data processing

The 10B, 11B and 12C signal intensities, as well as signal variability,decreased with the incremental cleaning progress (Fig. 1 A to C). Ab-lation intervals, shown here as increased signal plateaus, lasted for 30 s.Prior and after ablation gas blank data (no ablation) were collected. Theinitial spike in the signal of the uncleaned sample (Fig. 1 A) representssurficial contamination that was encrusted on most of the tests but wasrelatively easy to ablate. In other studies (e.g., Sinclair et al., 1998) thedata collected during the first few seconds of ablation was typicallydiscarded, and ablation was used to clean the foraminifera before theactual analyses. However, since the surface contamination signals inthis study are so distinct and easily identified, it was possible to removethese data points during the data processing and thus no additional pre-ablation was done. After the second cleaning procedure, some for-aminifera exhibited an additional 12C secondary peak at the end of theablation plateau, which we attribute to the laser ablating through a verythin individual into the adhesive pad. This secondary peak was treatedlike the initial peaks and was removed from the results by the dataprocessing routine.

The δ11B values were computed from the raw data following a dataprocessing sequence applied to each individual ablation spot of everyindividual specimen (see also Supplement S1):

First, boron isotope count rates were dead time corrected and sub-sequently background corrected. This was done by subtracting theaverage background intensity that was calculated from ∼30 integratedseconds of the background collected after each spot ablation from therespective ablation trend. Thereafter, all signals below 5000 cps for 10B,22,000 cps for 11B and 5000,000 cps for 12C were removed to extractthe actual ablation plateau from the baseline. The 10B/11B ratios werethen calculated with a slope regression analysis (Fietzke et al., 2008)(see Supplement for details). Since no statistically significant differ-ences between ablation spots on the same foraminifer were observed(e.g., Fig. 4 B), the results of all four ablation spots per individual werepooled so that the sum of the extracted plateaus per individual for-aminifer consisted of ∼66 data points (intensities [cps]) per for-aminifer. Based on the counting statistics, the δ11B twofold standarderror was calculated on the processed data for each individual. For easeof the comparative analysis, results of all analysed foraminifera col-lected during the fall (October to December 2013, three sediment trapcollection intervals, n = 32) and those from four sediment-trap inter-vals in the spring (March to May 2014, n = 19) were grouped togetheras representatives for their respective seasons (Fig. 2 bottom).

2.5. B/C analyses

As reported above 10B, 11B and 12C were collected simultaneously.Reported B/C ratios are based on the sum of the boron intensities di-vided by the carbon intensity. In order to convert intensity ratios toabsolute ratios [μmol/mol] a calibration was done according to themethod reported in Wall et al., 2019. To infer B/Ca ratios from con-verted B/C ratios we used a stochiometric ratio of C/Ca ∼ 1 as anapproximation for foraminiferal calcite.

Table 1Instrument and laser settings that were used for the boron isotope measurementon individual foraminifera. Uncleaned and cleaned-I samples measured under(1) intermediately hot plasma conditions (NAI∼0.5, ThO/Th<0.02), cleaned-II samples measured under (2) hot plasma conditions (NAI ∼3.5, ThO/Th<0.001).

AXIOM MC-ICP-MS

Cool gas 17.5 l min−1

Auxilary gas 0.9 l min−1

Sample gas 0.82 l min-1 He / 0.82 l min-1 Ar (1)1.15 l min-1 Ar (2)

RF power 1000 W (1)1120 W (2)

Reflected power 1-2 WAccelerating voltage 4980 VCones RAC 19 & RAC 705Resolution ∼ 400Laser ablation settings for samplesSpot size 75 μmFluence ∼ 1.8 J cm−2

Repetition rate 10 HzAblation time 30 sScan mode Spot analysis (300 shots)Laser ablation settings for NIST610Spot size 25 μmFluence ∼ 1.8 J cm−2

Repetition rate 5 HzScan speed 10 μm s−1

Wash out 2 sScan mode Line analysis (500 μm in length with 15 μm space between

lines) for constant intensity

D. Mayk, et al. Chemical Geology 531 (2020) 119351

3

2.6. Statistics

Normality of the data, where applicable, was confirmed using theShapiro-Wilk (Shapiro and Wilk, 1965) test as well as a visual inspec-tion by QQ-plots. Before statistical tests were performed, clear outliersencompassing unusually light δ11B values were removed from the da-taset (two data points from the cleaned-I spring samples, one data pointfrom the cleaned-II spring and fall samples [see Fig. 2 bottom]). Theseoutliers were identified by (1) their deviation from the normal dis-tribution in QQ-plots and (2) their distance from their respective dis-tribution’s median being larger than 1.5 times inter-quartile range. Theeffect of the different cleaning methods on intra group (grouped sam-ples from one season; e.g., spring or fall) variance was tested using F-tests. Two-sample T-tests were used to test for significant offsets be-tween seasons within each cleaning treatment. Before the T-test, unityof variance of the seasons within each cleaning treatment was con-firmed by F-tests. To test if cleaning had an effect on the offset betweenseasons, data were normalized by subtracting the means of each fallgroup from the associated spring group. This way a one-way ANOVAcould be run including all fall groups to test for differences betweentheir means. Thereafter, a Tukey HSD posthoc test was carried out to

test which spring groups are significantly different and thus have adifferent offset to their corresponding fall group. All statistical analyseswere carried out in R v.3.5.2 (R Core Team, 2018).

3. Results

In Fig. 2 we show the seasonal and intra seasonal variability of theanalysed foraminifera samples for each cleaning treatment. The averageprecision with which δ11B was analysed within a single O. universawas± 2.9‰ (2 SE). Although two different NAI settings were used toanalyse uncleaned, cleaned-I and cleaned-II foraminifera, respectively,within each NAI setting and within each treatment, seasonal data wererun at random, and therefore their relative δ11B variability is a robustfeature (i.e. not dependent on day and sequence of running). Ad-ditionally, within the low NAI the relation between boron concentra-tion and δ11B of the foraminifera follows the opposite trend than wouldbe expected from a mass bias effect (see Supplement for details) asreported in Sadekov et al., 2019 and Standish et al., 2019. This suggeststhat the δ11B variations within each session are not driven by matrixeffects within the low NAI settings and hence represent a real signature(i.e. likely not necessitating matrix effect corrections).

Fig. 1. Ablation intensity (kcps: 1000 and Mcps: 1000,000 counts per second) trends of 11 individual O. universa. [A] Ablation trends for 10B, 11B, 12C of uncleanedO. universa. [B] Ablation trends of cleaned O. universa after first cleaning. [C] Ablation trends of cleaned O. universa after second cleaning - additional spikes in the12C trends are due to shooting through the foraminifera and into the adhesive pad. Note: Measurement conditions for A and B differ from C (see text).

D. Mayk, et al. Chemical Geology 531 (2020) 119351

4

Generally, cleaning had the effect of reducing B/Ca ratios of theindividual specimens (3–4 fold [Fig. 3]), converging to similar ratiosbetween seasons for the cleaning-II samples towards values reported forO. universa based on bulk pooled foraminifera analyses (e.g., Henehanet al., 2016; Allen et al., 2011) but had a different effect on the springand fall samples’ δ11B variances. Cleaning-I significantly reduced theδ11B variance of the fall samples [F-test, F12, 27 = 5.265, p = 0.004]but had no effect on the δ11B variance of the spring samples [F-test, F31,55 = 1.423, p = 0.292]. Secondary cleaning-II had no effect on thevariance of the fall samples [F-test, F8, 12 = 0.304, p = 0.063] or thespring samples [F-test, F7, 31 = 1.549, p = 0.57]. Regardless of thecleaning treatment used, the mean of the spring and fall samples re-mained significantly different within treatments; uncleaned [T-test, T82

= 11.662, p < 2.2 10−16], cleaning-I [T-test, T45 = 6.631, p = 3.610-8], cleaning-II [T-test, T17 = 5.658, p = 2.83 10-5]. The effect ofcleaning on the offset between the seasonal means was tested afternormalizing the data. A one-way ANOVA showed that cleaning influ-enced the difference between seasonal means [ANOVA, F2, 94 = 12.78,p = 1.23 10-5]. The seasonal difference in the cleaned-I samples is

reduced but the seasonal difference in the uncleaned and the cleaned-IIsamples remained constant. To summarize: We observe a general de-crease in B/Ca ratios with increasing cleaning effort. While δ11B appearto become heavier, we think this observation needs further experi-mental evidence, since our first two analytical sessions did not includean appropriate carbonate consistency standard to prove accuracy of theresults.

After cleaning-II, the combined data for both seasons(14.8 ± 3.1‰ (2 SE, n = 19)) is in agreement with expected O. uni-versa δ11B data (14.4 ± 0.2‰ (2SD), based on local sea water datausing the calibration of Henehan et al., 2016). Interestingly, seasonmeans of fall (9.5‰) and spring (20.1‰) of the cleaned-II samplesexhibit an offset of +5.7‰ (spring) and ∼ -4.9‰ (fall) from whatwould be expected according to mean seawater δ11B (Fig. 2 and Sup-plemental Data). Intra-group variance was not impacted by cleaningtreatment.

Fig. 2. Results of the δ11B analyses of the individual foraminifera reported against cleaning treatment. Error bars represent the± 2 SE. A–C show results of theindividual foraminifera. A*-C* show box plots of the δ11B values of the seasonal groups as used for the statistical comparison. D shows the expected seasonal δ11Bvariability for O. universa calculated from pH measured around the sampling location in the Santa Barbara Basin. Thick black line represents the distribution median,box extends mark quartile one and three and whiskers are within 1.5 inter-quartile range.

D. Mayk, et al. Chemical Geology 531 (2020) 119351

5

4. Discussion

4.1. The effect of cleaning

Cleaning of O. universa in this study reduced the initial signal spike,the overall signal intensity and the signal variability of 10B and 11B(Fig. 1). We attribute this effect to the removal of boron associated with(1) external contamination, (2) contamination within the pores (e.g.,Glock et al., 2019) and (3) internal organic material (e.g., Eggins et al.,2004) which is rich in boron (Fig. 3). Cleaning-I resulted in a reducedoffset between seasons, which could be due to inconsistent efficiency ofremoving organics and other contaminants, or different composition ofcontaminants that are season dependent. The residual contaminationwas subsequently removed during cleaning-II. After cleaning-II, theoverall ablation intensities were greatly reduced which indicated re-moval of phases enriched in boron, such as internal organic bandings(e.g. Geerken et al., 2018, 2019), and other within-shell contaminants.The difference between seasonal means of uncleaned and cleaned-II

samples remained constant at ∼10.5‰ on average, suggesting thatboth methods are likely suitable to identify relative seasonal variations.Notably, despite the similarity in the seasonal difference in δ11B, theabsolute δ11B values for samples of both seasonal groups increasedfollowing cleaning-II. The cause for this difference requires furtherexperiments, since in this study an external quality control standard(JCt-1) testing the accuracy of the standardization was used only in theanalyses of specimens treated using cleaning-II. Specifically, sincecleaning-II resulted in mean B/Ca and δ11B comparable to publishedbulk data, this cleaning procedure should be applied in the future toobtain the δ11B composition of skeletal foraminifera calcite.

4.2. Detection of variability within and between samples

The highest temporal and spatial resolution that could be achievedby the method presented in this study is on the level of the individualforaminifera, representing the life span and depth occupied by eachindividual foraminifer. Intra-shell variability was too small to be

Fig. 3. δ11B vs B/Ca ratios of the group averages of the uncleaned, cleaned-I and cleaned-II samples. Black line represents linear regression through all points andgrey area represents the 95% confidence interval of the regression line.

Fig. 4. Variability in δ11B and the effect of cleaning encompassed on the different levels of examination. (A) typical internal error on a spot measurement throughoutall measurement sessions. (B) Average variability within individual O. universa. This also represents the highest possible resolution of δ11B values achieved in thisstudy. (C) Average variability between specimens from the same catch interval. (D) Offset between the two seasons fall and spring examined.

D. Mayk, et al. Chemical Geology 531 (2020) 119351

6

resolved because even though the ablation spot-size can be small en-ough to focus on different features, such as test chambers and theirinternal bands (Eggins et al., 2004; Sadekov et al., 2005; Allen et al.,2011; Fehrenbacher et al., 2015; Bonnin et al., 2019), analytical pre-cision for δ11B is insufficient to resolve differences on the data-pointlevel (Fig. 4 B). Since the instrumental settings that were used in thisstudy were set to be as close as possible to the limits of the countingstatistic (Fig. 4 A), allowing us to maximize the resolution withoutlosing too much precision, a different sampling approach and improvedinstrument architecture will be needed if we are to resolve the δ11Bcomposition of different internal components of each shell. Interest-ingly, average δ11B variability of the spot measurement is not stronglyaffected by the different cleaning methods employed (Fig. 4 B).

The highest temporal resolution resolved in this study is capturedamong individuals from the same collection intervals (i.e., the samesediment trap cap [Fig. 4 C]), representing changes on time scales of thelife span of the specific foraminifer (∼16 days [Faber and Be, 1987])and would thus also be immensely useful for comparison with core topanalyses (see Supplement S2). Our data, whether uncleaned, aftercleaning-I or cleaning-II, show that any single foraminifer responds tovery specific and unique conditions (i.e., seawater pH, temperature,food and light availability, respiration and photosynthesis rates) at thetime and location of its growth. The range of δ11B for specimens col-lected at one location and over a period of weeks to months is 7–9‰ [2SD]) (Fig. 2), and appears to be relatively constant within each seasonand irrespective of the cleaning employed (Fig. 4 C). Short-term en-vironmental pH variability at the sediment trap site can be as high as0.7 pH units (Hofmann and Washburn, 2018), which translates to∼4.7‰ Δδ11BO.universa, explaining 56–79 % of the foraminifera δ11Bvariability we see. The remainder, therefore, could represent vital ef-fects (e.g., physiology, genetics, fitness) of individual foraminiferawhich is also manifested as scatter around the regression line of theδ11B calibration (i.e. Henehan et al., 2016). It is thus not reasonable toexpect that single foraminifera analyses will become a substitute forconventional solution multi specimen analyses which represent averageconditions experienced by the pooled individual tests. To test howmany specimen would be required by LA-MC-ICP-MS analysis toachieve representative δ11B data at precision levels comparable to so-lution based multi-specimen analyses we estimate the expected 2 SEMbased on the average variability observed within the groups (collectionintervals) of this study and the student-t distribution (Fig. 5). The errorestimation shows that> 30 individuals would be required to achieve atwo-fold standard error of the mean comparable to the average preci-sion achieved on the individual foraminifer (2.9‰, 2 SE). 100 in-dividual foraminifera would result in an estimated 2 SEM of 1.5‰ andmore than 250 specimens would be required to theoretically decreasethe error below 1‰ (2 SEM) with LA-ICP-MS. This implies that bulkmeasurements would need to consider the effect of this variabilityduring data interpretation as well. Notably, although the seasonal anddaily pH fluctuations in the Santa Barbara Basin are likely higher thanin open ocean settings accentuating the variabilities observed, the un-derlying systematic should be alike for any setting.

Results for our cleaned-II data show that the mean of both seasons’δ11B agrees remarkably well with those expected from seawater values.Nevertheless, the mean δ11B differs among seasons by a total of 10.6‰,being 5.7‰ heavier in the spring and 4.9‰ lighter in the fall thanmean seawater borate, which is larger than the range expected from pHmeasured in the seawater at this site (Fig. 2). As mentioned before56–79 % of the seasonal variations can be explained by the range of pHfrom short-term environmental variability. When compared to re-presentative seawater δ11Bborate minima and maxima using the pH andtemperature time series of Hofmann and Washburn, 2018; Washburn,2019, and Kapsenberg and Hofmann, 2016 (Fig. 2 bottom), we see thatmaxima excursions towards heavier δ11Bborate in seawater may be thedriver behind the observed heavier δ11B values in the foraminiferacalcite from spring in comparison to the fall. This would suggest that

the δ11B in O. universa we capture is skewed towards the extrema ratherthan representing the average δ11Bborate of the seawater, which re-mained relatively stable throughout the time frame encompassed in thisstudy. Additionally, our observations may indicate season-specific off-sets associated with “vital effects” resulting in O. universa δ11B beingshifted towards heavier values during spring (e.g. through enhancedphotosynthesis) and towards lighter values during fall (e.g. throughenhanced respiration), potentially as a function of light/energy avail-ability and symbiont activity affecting the relative contributions of re-spiration and photosynthesis within a foraminiferal growing environ-ment. Without knowledge of the underlying drivers (environmentaland/or physiological) of these shifts precise pH reconstruction may bebiased by seasonality and uneven foraminifera annual abundances(Raitzsch et al., 2018).

The question therefore remains: Which processes affect the averageδ11B obtained by combining several individuals in a multi-specimenanalysis of a sediment sample, and therefore what does the averagevalue measured represent? For example, seasonal or interannualchanges to the foraminifera flux of any species and/or phenotypes intothe seafloor (Thunell et al., 1983; Thunell and Reynolds, 1984), or se-lective post burial dissolution of a sub set of tests (Berger and Parker,1970) may all bias the results of the conventional single species multi-specimen analysis to a particular time of year or time period which isover-represented in the accumulating archive (Raitzsch et al., 2018),resulting in a value that is very precise but possibly not a true averageof the environmental conditions of the time period the multi-specimenanalyses targeted. We propose that this issue could be resolved bylooking at multi-species reconstructions and refining possible inter-pretations by considering species-specific seasonal biases.

Furthermore, our results suggest that thorough cleaning of thesamples is needed to ensure that the δ11B results produced by LA-MC-ICP-MS are within the expected range to those obtained by multi-spe-cimen analyses. In our study, when combining the results of all thecleaned-II samples, a δ11B mean of 14.8 ± 3.1‰ (2 SE, n = 19) wasachieved for O. universa sediment trap samples. This result is wellwithin the range of expected δ11B (Fig. 2) using the seawater pH andtemperature variability at the site of collection and the multi-specimenO. universa calibration (Henehan et al., 2016).

5. Conclusion and recommendations

In this study, we present a novel approach to analyse δ11B in in-dividual foraminifera of the species O. universa with an average preci-sion of± 2.9‰ (2 SE). We show that it is possible to investigate therelative variability between individual foraminifera and between sea-sons. δ11B consistent with expected values based on local pH mea-surements can be obtained on oxidatively cleaned O. universa and re-sults approach expected δ11B from multi-specimen analyses whenaveraging data from multiple individuals (here, 19 specimens).Furthermore, we observed that environmental variations in pH andtemperature describe the majority of the variability in δ11B varianceamong individual foraminifera collected at different seasons, with20–40% of the variance requiring additional processes affecting for-aminifera δ11B, such as season dependant vital effects not resolved bythe multi-specimen calibration. This suggests that care should be takenwhen combining tests to obtain average values, as these averages maybe affected by processes that can skew the test representation. Wepropose that the method laid out in this study may be used to identifybiases in conventional multi-specimen analyses that could rise from un-even accumulation of tests during different seasons and periods of highfrequency short term environmental variability. It also offers a uniquetool to quantify individual O. universa vital effects, and if enough in-dividuals are measured may represent the environmental short-termvariability (range) of pH.

D. Mayk, et al. Chemical Geology 531 (2020) 119351

7

Declaration of Competing Interest

The authors declare that they have no known competing financialinterests or personal relationships that could have appeared to influ-ence the work reported in this paper.

Acknowledgements

The authors would like to thank Robert Thunell and Emily Osbornefrom the University of South Carolina for providing the foraminiferasamples that were used in this study and Barbara Balestra for pickingthe foraminifera from the sediment trap caps. Many thanks go toNicolaas Glock for his invaluable support on the cleaning procedures.We would also like to thank Oscar Branson and three anonymous re-viewers for their comprehensive reviews which helped to substantiallyimprove the manuscript.

Appendix A. Supplementary data

Supplementary material related to this article can be found, in theonline version, at doi:https://doi.org/10.1016/j.chemgeo.2019.119351.

References

Allen, K.A., Hönisch, B., Eggins, S.M., Yu, J., Spero, H.J., Elderfield, H., 2011. Controls onboron incorporation in cultured tests of the planktic foraminifer Orbulina universa.Earth Planet. Sci. Lett. 309, 291–301. https://doi.org/10.1016/j.epsl.2011.07.010.

Barker, S., Greaves, M., Elderfield, H., 2003. A study of cleaning procedures used forforaminiferal Mg/Ca paleothermometry. Geochem. Geophys. Geosyst. 4, 1–20.https://doi.org/10.1029/2003GC000559.

Berger, W.H., Parker, F.L., 1970. Diversity of planktonic foraminifera in deep-sea sedi-ments. Science (80-.) 168, 1345–1347. https://doi.org/10.1126/science.168.3937.1345.

Bonnin, E.A., Zhu, Z., Fehrenbacher, J.S., Russell, A.D., Hönisch, B., Spero, H.J., Gagnon,A.C., 2019. Submicron sodium banding in cultured planktic foraminifera shells.Geochim. Cosmochim. Acta 253, 127–141. https://doi.org/10.1016/j.gca.2019.03.024.

Dickson, A.G., 1990. Thermodynamics of the dissociation of boric acid in potassiumchloride solutions from 273.15 to 318.15 K. J. Chem. Eng. Data 35, 253–257. https://

doi.org/10.1021/je00061a009.Eggins, S., De Deckker, P., Marshall, J., 2003. Mg/Ca variation in planktonic foraminifera

tests: implications for reconstructing palaeo-seawater temperature and habitat mi-gration. Earth Planet Sci. Lett. 212, 291–306. https://doi.org/10.1016/S0012-821X(03)00283-8.

Eggins, S.M., Sadekov, A., De Deckker, P., 2004. Modulation and daily banding of Mg/Cain Orbulina universa tests by symbiont photosynthesis and respiration: a complica-tion for seawater thermometry? Earth Planet. Sci. Lett. 225, 411–419. https://doi.org/10.1016/j.epsl.2004.06.019.

Faber, W.W., Be, A.W., 1987. Growth of the spinose planktonic foraminifer orbulinauniversa in laboratory culture and the effect of temperature on life processes. J. Mar.Biolog. Assoc. U.K. 67, 343–358. https://doi.org/10.1017/S0025315400026655.

Farmer, J.R., Branson, O., Uchikawa, J., Penman, D.E., Hönisch, B., Zeebe, R.E., 2019.Boric acid and borate incorporation in inorganic calcite inferred from B/Ca, boronisotopes and surface kinetic modeling. Geochim. Cosmochim. Acta 244, 229–247.https://doi.org/10.1016/j.gca.2018.10.008.

Fehrenbacher, J.S., Spero, H.J., Russell, A.D., Vetter, L., Eggins, S., 2015. Optimizing LA-ICP-MS analytical procedures for elemental depth profiling of foraminifera shells.Chem. Geol. 407–408, 2–9. https://doi.org/10.1016/j.chemgeo.2015.04.007.

Fietzke, J., Frische, M., 2016. Experimental evaluation of elemental behavior during LA-ICP-MS: influences of plasma conditions and limits of plasma robustness. J. Anal. At.Spectrom. 31, 234–244. https://doi.org/10.1039/c5ja00253b.

Fietzke, J., Heinemann, A., Taubner, I., Böhm, F., Erez, J., Eisenhauer, A., 2010. Boronisotope ratio determination in carbonates via LA-MC-ICP-MS using soda-lime glassstandards as reference material. J. Anal. At. Spectrom. 25, 1953–1957. https://doi.org/10.1039/c0ja00036a.

Fietzke, J., Liebetrau, V., Günther, D., Gürs, K., Hametner, K., Zumholz, K., Hansteen,T.H., Eisenhauer, A., 2008. An alternative data acquisition and evaluation strategyfor improved isotope ratio precision using LA-MC-ICP-MS applied to stable andradiogenic strontium isotopes in carbonates. J. Anal. At. Spectrom. 23, 955–961.https://doi.org/10.1039/b717706b.

Ford, H.L., Ravelo, A.C., Polissar, P.J., 2015. Reduced El Niño-Southern Oscillation duringthe last glacial maximum. Science (80-.) 347, 255–258. https://doi.org/10.1126/science.1258437.

Foster, G.L., 2008. Seawater pH, pCO2and [CO2-3] variations in the Caribbean Sea overthe last 130 kyr: a boron isotope and B/Ca study of planktic foraminifera. EarthPlanet. Sci. Lett. 271, 254–266. https://doi.org/10.1016/j.epsl.2008.04.015.

Foster, G.L., Rae, J.W.B., 2016. Reconstructing ocean pH with boron isotopes in for-aminifera. Annu. Rev. Earth Planet. Sci. 44, 207–237. https://doi.org/10.1146/annurev-earth-060115-012226.

Geerken, E., de Nooijer, L.J., Roepert, A., Polerecky, L., King, H.E., Reichart, G.J., 2019.Element banding and organic linings within chamber walls of two benthic for-aminifera. Sci. Rep. 9, 3598. https://doi.org/10.1038/s41598-019-40298-y.

Geerken, E., Jan De Nooijer, L., Van DIjk, I., Reichart, G.J., 2018. Impact of salinity onelement incorporation in two benthic foraminiferal species with contrasting mag-nesium contents. Biogeosciences 15, 2205–2218. https://doi.org/10.5194/bg-15-2205-2018.

Glock, N., Liebetrau, V., Eisenhauer, A., Rocholl, A., 2016. High resolution I/Ca ratios of

Fig. 5. Standard error of the mean (SEM) estimation based on the group variability observed in this study. Errors were estimated using the student-t distribution.

D. Mayk, et al. Chemical Geology 531 (2020) 119351

8

benthic foraminifera from the Peruvian oxygen-minimum-zone: a SIMS derived as-sessment of a potential redox proxy. Chem. Geol. 447, 40–53. https://doi.org/10.1016/j.chemgeo.2016.10.025.

Glock, N., Liebetrau, V., Vogts, A., Eisenhauer, A., 2019. Organic heterogeneities in for-aminiferal calcite traced through the distribution of N, S, and I measured withNanoSIMS: a new challenge for element-ratio-Based paleoproxies? Front. Earth Sci. 7,1–14. https://doi.org/10.3389/feart.2019.00175.

Gutjahr, M., Bordier, L., Douville, E., Farmer, J., Foster, G.L., Hathorne, E., Honisch, B.,Lemarchand, D., Louvat, P., McCulloch, M., Noireaux, J., Pallavicini, N., Rae, J.,Rodushkin, I., Roux, P., Steward, J.A., Thil, F., You, C.F., 2014. Boron IsotopeIntercomparison Project (BIIP): development of a new carbonate standard for stableisotopic analyses. EGU Gen. Assem. Conf. Abstr. 16, 5028.

Hemming, N.G., Hanson, G.N., 1992. Boron isotopic composition and concentration inmodern marine carbonates. Geochim. Cosmochim. Acta 56, 537–543. https://doi.org/10.1016/0016-7037(92)90151-8.

Henehan, M.J., Foster, G.L., Bostock, H.C., Greenop, R., Marshall, B.J., Wilson, P.A.,2016. A new boron isotope-pH calibration for Orbulina universa, with implicationsfor understanding and accounting for ‘vital effects.’. Earth Planet. Sci. Lett. 454,282–292. https://doi.org/10.1016/j.epsl.2016.09.024.

Henehan, M.J., Rae, J.W.B., Foster, G.L., Erez, J., Prentice, K.C., Kucera, M., Bostock,H.C., Martínez-Botí, M.A., Milton, J.A., Wilson, P.A., Marshall, B.J., Elliott, T., 2013.Calibration of the boron isotope proxy in the planktonic foraminifera Globigerinoidesruber for use in palaeo-CO2 reconstruction. Earth Planet. Sci. Lett. 364, 111–122.https://doi.org/10.1016/j.epsl.2012.12.029.

Hofmann, G.E., Washburn, L., 2018. SBC LTER: ocean: time-series: mid-water SeaFET pHand CO2 system chemistry with surface and bottom dissolved oxygen at SantaBarbara Harbor/Stearns Wharf(SBH), ongoing since 2012-09-15. Environ. DataInitiative. https://doi.org/10.6073/pasta/d95118a18b7124501a85962e8a8147b4.Dataset accessed 7/15/2019.

Hönisch, B., Bijma, J., Russell, A.D., Spero, H.J., Palmer, M.R., Zeebe, R.E., Eisenhauer,A., 2003. The influence of symbiont photosynthesis on the boron isotopic composi-tion of foraminifera shells. Mar. Micropaleontol. 49, 87–96. https://doi.org/10.1016/S0377-8398(03)00030-6.

Hönisch, B., Hemming, N.G., Archer, D., Siddall, M., McManus, J.F., 2009. Atmosphericcarbon dioxide concentration across the mid-pleistocene transition. Science (80-.)324, 1551–1554. https://doi.org/10.1126/science.1171477.

Kaczmarek, K., Langer, G., Nehrke, G., Horn, I., Misra, S., Janse, M., Bijma, J., 2015.Boron incorporation in the foraminifer Amphistegina lessonii under a decoupledcarbonate chemistry. Biogeosciences 12, 1753–1763. https://doi.org/10.5194/bg-12-1753-2015.

Kapsenberg, L., Hofmann, G.E., 2016. Ocean pH time-series and drivers of variabilityalong the northern Channel Islands, California, USA. Limnol. Oceanogr. 61, 953–968.https://doi.org/10.1002/lno.10264.

Klochko, K., Kaufman, A.J., Yao, W., Byrne, R.H., Tossell, J.A., 2006. Experimentalmeasurement of boron isotope fractionation in seawater. Earth Planet. Sci. Lett. 248,276–285. https://doi.org/10.1016/j.epsl.2006.05.034.

Lemarchand, D., Gaillardet, J., Lewin, É., Allègre, C.J., 2000. The influence of rivers onmarine boron isotopes and implications for reconstructing past ocean pH. Nature408, 951–954.

Lemarchand, D., Gaillardet, J., Lewin, É., Allègre, C.J., 2002. Boron isotope systematics inlarge rivers: implications for the marine boron budget and paleo-pH reconstructionover the Cenozoic. Chem. Geol. 190, 123–140.

Martínez-Botí, M.A., Foster, G.L., Chalk, T.B., Lunt, D.J., Rohling, E.J., Badger, M.P.S.,Schmidt, D.N., Sexton, P.F., Pancost, R.D., 2015. Plio-Pleistocene climate sensitivityevaluated using high-resolution CO2 records. Nature 518, 49–54. https://doi.org/10.1038/nature14145.

Metcalfe, B., Feldmeijer, W., Ganssen, G.M., 2019. Oxygen isotope variability of plank-tonic foraminifera provide clues to past upper ocean seasonal variability.Paleoceanogr. Paleoclimatol. https://doi.org/10.1029/2018pa003475.

Nir, O., Vengosh, A., Harkness, J.S., Dwyer, G.S., Lahav, O., 2015. Direct measurement ofthe boron isotope fractionation factor: reducing the uncertainty in reconstructingocean paleo-pH. Earth Planet. Sci. Lett. 414, 1–5. https://doi.org/10.1016/j.epsl.2015.01.006.

Noireaux, J., Mavromatis, V., Gaillardet, J., Schott, J., Montouillout, V., Louvat, P.,

Rollion-Bard, C., Neuville, D.R., 2015. Crystallographic control on the boron isotopepaleo-pH proxy. Earth Planet. Sci. Lett. 430, 398–407. https://doi.org/10.1016/j.epsl.2015.07.063.

Pracht, H., Metcalfe, B., Peeters, F.J.C., 2018. Oxygen isotope composition of finalchamber of planktic foraminifera provides evidence for vertical migration and depthintegrated growth. Biogeosci. Discuss. 1–32. https://doi.org/10.5194/bg-2018-146.

R Core Team, 2018. R: A Language and Environment for Statistical Computing.Rae, J.W.B., Foster, G.L., Schmidt, D.N., Elliott, T., 2011. Boron isotopes and B/Ca in

benthic foraminifera: Proxies for the deep ocean carbonate system. Earth Planet. Sci.Lett. 302, 403–413. https://doi.org/10.1016/j.epsl.2010.12.034.

Raitzsch, M., Bijma, J., Benthien, A., Richter, K.U., Steinhoefel, G., Kučera, M., 2018.Boron isotope-based seasonal paleo-pH reconstruction for the Southeast Atlantic – amultispecies approach using habitat preference of planktonic foraminifera. EarthPlanet. Sci. Lett. 487, 138–150. https://doi.org/10.1016/j.epsl.2018.02.002.

Sadekov, A., Lloyd, N.S., Misra, S., Trotter, J., D’Olivo, J., Mcculloch, M., 2019. Accurateand precise microscale measurements of boron isotope ratios in calcium carbonatesusing laser ablation multicollector-ICPMS. J. Anal. At. Spectrom. 34, 550–560.https://doi.org/10.1039/c8ja00444g.

Sadekov, A.Y., Eggins, S.M., De Deckker, P., 2005. Characterization of Mg/Ca distribu-tions in planktonic foraminifera species by electron microprobe mapping. Geochem.Geophys. Geosyst. 6. https://doi.org/10.1029/2005GC000973.

Sanyal, A., Bijma, J., Spero, H., Lea, D.W., 2001. Empirical relationship between pH andthe boron isotopic composition of Globigerinoides sacculifer: implications for theboron isotope paleo-pH proxy. Paleoceanography 16, 515–519. https://doi.org/10.1029/2000PA000547.

Shapiro, A.S.S., Wilk, M.B., 1965. Biometrika trust an analysis of variance test for nor-mality (complete samples) published by : Oxford University Press on behalf ofBiometrika Trust Stable. Biometrika 52, 591–611.

Sinclair, D.J., Kinsley, L.P.J., Mcculloch, M.T., 1998. High resolution analysis of traceelements in corals by laser ablation ICP-MS. Geochim. Cosmochim. Acta 62,1889–1901. https://doi.org/10.1016/S0016-7037(98)00112-4.

Spivack, A.J., Edmond, J.M., 1987. Boron isotope exchange between seawater and theoceanic crust. Geochim. Cosmochim. Acta 51, 1033–1043.

Standish, C.D., Chalk, T.B., Babila, T.L., Milton, J.A., Palmer, M.R., Foster, G.L., 2019. Theeffect of matrix interferences on in situ boron isotope analysis by laser ablation multi-collector inductively coupled plasma mass spectrometry. Rapid Commun. MassSpectrom. 33, 959–968. https://doi.org/10.1002/rcm.8432.

Thunell, R.C., 1998. Particle fluxes in a coastal upwelling zone: sediment trap results fromSanta Barbara Basin, California. Deep Sea Res. Part II Top. Stud. Oceanogr. 45,1863–1884. https://doi.org/10.1016/s0967-0645(98)80020-9.

Thunell, R.C., Curry, W.B., Honjo, S., 1983. Seasonal variation in the flux of planktonicforaminifera: time series sediment trap results from the Panama Basin. Earth Planet.Sci. Lett. 64, 44–55. https://doi.org/10.1016/0012-821X(83)90051-1.

Thunell, R.C., Reynolds, L.A., 1984. Sedimentation of planktonic foraminifera: seasonalchanges in species flux in the Panama Basin. Micropaleontology 30, 243. https://doi.org/10.2307/1485688.

Vengosh, A., Kolodny, Y., Starinsky, A., Chivas, A.R., McCulloch, M.T., 1991.Coprecipitation and isotopic fractionation of boron in modern biogenic carbonates.Geochim. Cosmochim. Acta 55, 2901–2910. https://doi.org/10.1016/0016-7037(91)90455-E.

Wall, M., Fietzke, J., Crook, E.D., Paytan, A., 2019. Using B isotopes and B/Ca in coralsfrom low saturation springs to constrain calcification mechanisms. Nat. Commun. 10(1), 1–9.

Washburn, L., 2019. SBC LTER: reference: sea-aurface water temperature, Santa BarbaraHarbor, Santa Barbara, CA, USA, 1955 to present, ongoing. Environ. Data Initiative.https://doi.org/10.6073/pasta/64fe7f595fcdb84207f3a6b2aadafe5a. Datasetaccessed 7/15/2019.

Zeebe, R.E., Wolf-Gladrow, D.A., 2001. CO2 in Seawater-Equilibrium, Kinetics, Isotopes.Elsevier Oceanography. Gulf Professional Publishinghttps://doi.org/10.1016/S0924-7963(02)00179-3.

Zeebe, R.E., Wolf-Gladrow, D.A., Bijma, J., Hönisch, B., 2003. Vital effects in foraminiferado not compromise the use of δ 11 B as a paleo- p H indicator: Evidence frommodeling. Paleoceanography 18https://doi.org/10.1029/2003PA000881. n/a-n/a.

D. Mayk, et al. Chemical Geology 531 (2020) 119351

9