Calibration of the pH-δ¹¹B and temperature-Mg/Li proxies in the long-lived high-latitude crustose coralline red alga Clathromorphum compactum via controlled laboratory experiments

Abstract

A solid understanding of global oceanic change throughout Holocene time is needed to contextualize and interpret recent observations of rapid warming (Moore, 2016), ocean acidification (Popova et al., 2014; Qi et al., 2017), increasing meltwater input (Halfar et al., 2013; Notz and Stroeve, 2016) and circulation changes (Liu et al., 2017; Rahmstorf et al., 2015; Yang et al., 2016) in the Arctic and subarctic Oceans. Precisely reconstructing acidification and temperature variations throughout the Holocene will provide a vital context for interpreting current environmental changes and future climate projections in the region. However, existing paleoenvironmental reconstructions are sparse and uncertain, largely owing to limited availability of high fidelity paleoceanographic archives, such as marine carbonates, in high latitude waters. Coralline algae of the genus Clathromorphum have emerged as key candidates for reconstructing high-latitude environmental variability at annual to sub-annual resolution. Here, we present the first empirical calibrations of boron isotope-pH and Mg/Li-temperature relationships within the long-lived, crustose coralline red alga Clathromorphum compactum. Calibration experiments were performed in triplicate, growing wild-collected specimens for four months at three controlled temperatures (6.4–12.4 °C) and four pCO₂ conditions (352–3230 ppm), to test the effects of these environmental parameters on the isotopic and elemental composition of the algal skeleton. We find that boron isotopes within the skeleton of C. compactum (δ¹¹B_cc) are well correlated with δ¹¹B of seawater borate (δ¹¹Β_borate), defining the following equation: δ¹¹Β_cc (2σ) = 1.46 (0.06) δ¹¹Β_borate + 6.91 (0.72). This equation can be used to reconstruct δ¹¹Β_borate of the coralline alga’s ambient seawater, from which past seawater pH can be calculated. We also identified a strong correlation between skeletal Mg/Li ratio and seawater temperature, defined by the equation: Mg/Li (2σ) = 0.17 (0.02) temperature (^oC) + 1.02 (0.16). Therefore, despite the strong biological control that this species appears to exert on calcification site pH (elevated 1.0–1.6 pH units above seawater pH, inferred from δ¹¹B_cc > δ¹¹Β_borate), and the apparent relationship between skeletal extension rate and skeletal Li/Ca and Mg/Ca, the δ¹¹B_cc and Mg/Li ratios of the coralline alga’s skeleton strongly and significantly respond to ambient seawater pH and temperature, respectively. These results support the use of δ¹¹B and Mg/Li within C. compactum for pH and temperature reconstructions of northern high-latitude oceans.

Introduction

Recent observations of dramatic warming (Moore, 2016), acidification (Popova et al., 2014, Qi et al., 2017), increasing meltwater input (Halfar et al., 2013, Notz and Stroeve, 2016), and circulation changes (Rahmstorf et al., 2015, Yang et al., 2016, Liu et al., 2017) in the Arctic and subarctic oceans suggest that these regions are particularly sensitive to impacts of anthropogenic global change. However, a better understanding of pre-anthropogenic environmental variations throughout Holocene time is needed to contextualize and interpret these changes, and to inform predictions of future change. Existing paleoceanographic reconstructions for these regions are relatively sparse and highly uncertain, largely because of the limited availability of reliable paleoceanographic archives in high-latitude waters, such as annually layered marine carbonates.

Crustose coralline red algae have been used for paleoecological reconstructions of geological intervals dating back to the early Tertiary (Adey, 1979). However, they have only recently been used to reconstruct environmental change of the Common Era (Hetzinger et al., 2009, Kamenos and Law, 2010, Burdett et al., 2011, Chan et al., 2011, Halfar et al., 2011, Williams et al., 2011, Kamenos, 2012, Fietzke et al., 2015, Hetzinger et al., 2018). This was largely because of incomplete understanding of the coralline algae’s complex anatomy and the lack of suitable methodologies for analysing geochemical signals within their skeletons with the required spatial resolution and analytical accuracy. Coralline algae of the genus Clathromorphum have emerged as a promising climate archive of high-latitude marine environments, because of their resolvable annual growth bands and multi-century lifespan (Frantz et al., 2005, Halfar et al., 2007, Halfar et al., 2013) and their wide distribution in mid- and high-latitude environments throughout the northwest Atlantic, North Pacific, and Arctic Oceans (Adey et al., 2008) (Fig. 1). This has led to increased assessment of the genus’s climate archiving potential.

Coralline algae are the most diverse and abundant calcareous organisms within intertidal and shallow subtidal zones around the world (Steneck, 1986). They are keystone species within such ecosystems because they provide substrate for larval settlement of invertebrates onto rocky and sediment-dominated substrates. They also provide food for grazers, function as nursery grounds for a number of species (Steneck and Martone, 2007, Chenelot et al., 2011), and increase the structural integrity of reefs and sediments within the photic zone (Heyward and Negri, 1999, Roberts, 2001). Depending on their degree of control over calcification, dissolution, and grazing of their relatively soluble high-Mg calcite skeleton, ocean acidification and warming may have highly deleterious effects on their growth and survival (Borowitzka and Larkum, 1987, Hoegh-Guldberg et al., 2007, Ries et al., 2009, Fabricius et al., 2015). Specifically, coralline algae in colder, higher-latitude waters, which are characterized by lower calcium carbonate saturation states owing to increased solubility of atmospheric CO₂ and increasing meltwater input, could reach their limits of resilience even sooner than species of coralline algae in warmer, lower latitude locations.

Coralline algae of the genus Clathromorphum grow at a rate of 300–400 μm yr⁻¹ in the warmer fringes of the subarctic ocean and Aleutian Islands, and 100 μm yr⁻¹ or less in the colder reaches of the subarctic and Arctic oceans (Adey et al., 2013). Clathromorphum produce annually resolved layers of high Mg-calcite crystals over their lifetime, delineated by seasonal changes in skeletal density (Chan et al., 2017). Living tissue and photosynthetic epithelial cells cover the meristem, protecting the calcified layers beneath the meristem, the perithallium, from diagenetic alteration (Alexandersson, 1974) and from herbivory (Steneck, 1986). A continuous record of skeletal accretion is thereby maintained in the perithallium (Steneck, 1982, Adey et al., 2013), which is well-suited for archiving high-resolution paleoenvironmental variability (Fig. 2). This mode of calcification contrasts that of other species of coralline algae that lack the meristem, and instead calcify across several cell layers below the coralline alga’s surface, creating a diffusive, time-integrated band of calcification (Adey et al., 2013), limiting the temporal resolution of the archive.

Calcification within coralline algae is likely regulated by a number of metabolic processes that influence the carbonate system, including photosynthesis and respiration (Smith and Roth, 1979, Gao et al., 1993, Beer and Larkum, 2001, Hurd et al., 2011, Martin et al., 2013). For example, the calcification rate of coralline algae has been directly linked to photosynthetic rate (Pentecost, 1978) and the availability of dissolved inorganic carbon (Digby, 1977, Gao et al., 1993). Seawater dissolved inorganic carbon is converted to CO₂ for photosynthesis within the algae through the action of the enzyme carbonic anhydrase, ion transporters, and proton pumps (McConnaughey and Falk, 1991, McConnaughey and Whelan, 1997, Comeau et al., 2013, Hofmann et al., 2016). Photosynthesis, in turn, removes CO₂, thereby increasing local calcite saturation state and promoting calcification (Gao et al., 1993). Calcification in coralline algae likely occurs in semi-enclosed regions within and between cell walls (Adey et al., 2013). However, calcein staining and boron isotope studies indicate that, for some species of coralline algae, the sites of calcification are partially open to seawater exchange (Pauly et al., 2015, Donald et al., 2017). The role of photosynthesis in coralline algal calcification is unclear in the genus Clathromorphum, as they calcify even under extended periods of darkness (Adey, 1998, Adey et al., 2013). This may arise from the algae’s ability to store energy during periods of ample light and photosynthesis (Adey et al., 2013), and supports prior assertions that organic templates and ion pumps are also important in coralline algal calcification (Borowitzka and Larkum, 1987, Adey, 1998, Rahman and Halfar, 2014).

Boron isotopes in several marine carbonates have been used as a proxy of paleo-seawater pH (Hemming and Hanson, 1992, Zeebe and Wolf-Gladrow, 2001, Foster and Rae, 2016). The basis for the boron isotope proxy of seawater pH stems from the observation that both the borate abundance and the boron isotopic composition of borate in seawater (δ¹¹B_borate) increases systematically with increasing seawater pH (Zeebe and Wolf-Gladrow, 2001). The relationship between δ¹¹B_borate and seawater pH (pH_sw) is described as follows:

$$p{H}_{\mathit{sw}}=pK\begin{array}{c}\ast \\ B\end{array}-\log \left[-\frac{{\delta }^{11}{B}_{\mathit{sw}}-{\delta }^{11}{B}_{\mathit{borate}}}{{\delta }^{11}{B}_{\mathit{sw}}-\left({\alpha }_B{\delta }^{11}{B}_{\mathit{borate}}-1000\left({\alpha }_B-1\right)\right)}\right]$$

where pK^*_B is the constant describing the dissociation equilibrium between boric acid and borate ion in seawater (Dickson, 1990a), δ¹¹B_sw is the seawater boron isotopic composition (in delta notation relative to NIST SRM 951 boric acid), and α_B is the equilibrium constant for boron isotope fractionation between boric acid and borate ion in seawater (1.0272; (Klochko et al., 2006) within uncertainty of other estimates (Nir et al., 2015)).

Many species of calcium carbonate precipitating organisms elevate their calcifying fluid pH (pH_cf) to promote calcification, as revealed by pH-microelectrode, pH-sensitive dyes, and/or boron isotope studies (Al-Horani et al., 2003, Krief et al., 2010, Ries, 2011a, Venn et al., 2011, Anagnostou et al., 2012, McCulloch et al., 2012, Venn et al., 2013, Holcomb et al., 2014, Sutton et al., 2018). Therefore, to estimate pH_sw from skeletal δ¹¹B (δ¹¹B_cc) of coralline algal calcite, which prior studies suggest occurs primarily as borate ion, (Cornwall et al., 2017, Donald et al., 2017), species-specific relationships must be empirically defined between seawater δ¹¹B_borate and δ¹¹B_cc. These relationships are used to convert δ¹¹B_cc into δ¹¹B_borate, which can then be substituted into Eq. (1) to solve for pH_sw.

There is no established calibration of the δ¹¹Β_cc vs. pH_sw proxy in any species of coralline red alga. However, coralline algal δ¹¹Β has been used to estimate pH at the alga’s site of calcification. For example, δ¹¹Β_cc was found to be substantially elevated relative to δ¹¹Β_borate of the ambient seawater in two species of low-latitude branching, non-articulate coralline algae (Neogoniolithon sp. and Sporolithon durun) and within one species of low-latitude articulate coralline red alga (Amphiroa anceps) cultured over a range of controlled pCO₂ conditions. The increase in δ¹¹Β_cc relative to δ¹¹Β_borate translates to a 0.5–1.5 unit increase in pH_cf relative to ambient pH_sw (Cornwall et al., 2017, Donald et al., 2017). Similarly, a wild specimen of C. nereostratum exhibited substantially elevated δ¹¹Β_cc relative to δ¹¹Β_borate of ambient seawater, which was also attributed to the alga’s pH_cf being at least 0.6 units greater than its ambient pH_sw (Fietzke et al., 2015).

The Mg/Ca ratio of calcite has been proposed as a temperature proxy in both inorganically (Berner, 1975) and organically precipitated calcite (Chave, 1954). In coralline algae, the majority of skeletal Mg²⁺ is incorporated into their high-Mg calcite lattice, rather than being associated with organic matter or other Mg-bearing mineral phases (Ries, 2006, Kamenos et al., 2009). The Mg/Ca ratio of coralline algal calcite varies as a function of seawater temperature (e.g. Chave, 1954, Kamenos et al., 2008, Williams et al., 2014, Hetzinger et al., 2018), seawater Mg/Ca (Stanley et al., 2002, Ries, 2006), seawater pH (Ries, 2011b), and growth rate (Moberly, 1968, Kolesar, 1978, Sletten et al., 2017) in a manner similar to that of inorganically precipitated calcite (Gabitov et al., 2014b). The Mg/Ca of coralline algal calcite has also been shown to vary amongst species (through so-called ‘vital effects’), and with seasonal cycles in insolation and sea ice cover affecting light levels (Moberly, 1968, Adey et al., 2013, Sletten et al., 2017, Williams et al., 2018). However, these effects are difficult to isolate in wild specimens due to their seasonal covariation. Furthermore, most Mg/Ca-temperature relationships in coralline algae are calibrated using wild specimens grown in poorly constrained seawater temperatures, such as those derived from satellites (Williams et al., 2014), often by averaging temperatures over 2° × 2° latitudinal-longitudinal areas (Hetzinger et al., 2018), and/or by aligning maxima and minima to temporally coordinate the elemental and temperature timeseries (Kamenos et al., 2008). All of these approaches could lead to inaccuracies in the Mg/Ca vs. temperature calibrations and in the final reconstructed seawater temperatures.

Lithium/calcium ratios (Li/Ca) in inorganically precipitated calcite has been shown to vary inversely with temperature (Marriott et al., 2004a, Marriott et al., 2004b). However, the Li/Ca compositions of foraminifera, corals, coralline algae and other marine calcifiers are also influenced by variations in seawater Li/Ca ratio, carbonate ion concentration, calcium carbonate polymorph mineralogy, and calcification rate (Delaney et al., 1985, Hall and Chan, 2004, Case et al., 2010, Caragnano et al., 2014, Montagna et al., 2014, Fowell et al., 2016, Dellinger et al., 2018).

The Mg/Li composition of aragonitic corals (Case et al., 2010, Montagna et al., 2014, Fowell et al., 2016) and calcitic and aragonitic foraminifera (Marchitto et al., 2007, Bryan and Marchitto, 2008, Marchitto et al., 2018) exhibits a strong positive dependence on temperature. According to a Rayleigh model of calcification (e.g. Elderfield et al., 1996), growth rate could have similar effects on the partitioning of Li⁺ and Mg²⁺ between the carbonate skeleton and the organism’s calcifying fluid. Therefore, dividing Li/Ca by Mg/Ca reduces the secondary impacts of calcification rate on the elemental relationship with temperature, thereby rendering Li/Mg a more reliable recorder of seawater temperature than Li/Ca or Mg/Ca alone (Bryan and Marchitto, 2008, Marchitto et al., 2018). However, a prior field-based Mg/Li calibration of the coralline alga L. kotschyanum did not show any improvement over the Mg/Ca or Li/Ca temperature proxies (Caragnano et al., 2014).

Here we present the first rigorous calibration of the pH-δ¹¹Β and temperature-Mg/Li relationships for specimens of the arctic/subarctic coralline alga C. compactum, which were cultured for 120 days in a controlled laboratory experiment comprising three seawater temperature conditions crossed with four CO₂-manipulated seawater pH conditions.