Calcium isotopic variability of cervid bioapatite and implications for mammalian physiology and diet

Highlights

• A lactation signal is preserved in bone Ca isotope composition of female red deer.

• Antlers display a different Ca isotope composition than the rest of the skeleton.

• Enamel Ca isotope composition changes at the time when cervids are weaned.

• Enamel serial micro-sampling allow to untangle dietary signals from other factors.

Abstract

There are clues that calcium (Ca) isotope composition of vertebrate bioapatite is influenced by diet and trophic level. These clues however conflict with several cases of mammal species exhibiting Ca isotope compositions which are inconsistent with their trophic levels. These observations support that diet may not be the only factor driving the Ca isotope composition in mammalian enamel and bone. To investigate this question, we selected a modern Cervus elaphus population (Bauges Natural Regional Park, Alps, France) to serve as a model for Ca isotope physiology of cervids and other mammals. Subsequently, we reinvestigated the case of the fossil Rangifer tarandus population from Jaurens (Late Pleistocene locality, 32.6 to 29.7 kyr BP, France), a population for which abnormal ⁴⁴Ca-depleted isotope compositions (regarding the low trophic level of reindeers) have been previously documented. By combining bone samplings and serial enamel micro-samplings, we discuss the main potential sources of Ca isotopic variability in bioapatite of young and adult individuals. This includes the effects of gestation, lactation, antlerogenesis, browser-grazer ecologies, osteophagia and natural mineral licks. Our results suggest an important effect of lactation on bone Ca isotope composition (δ^44/42Ca = +0.19 ± 0.09‰; 95% confidence interval), whereas other factors such as gestation or antlerogenesis seem more secondary. Our enamel micro sampling method allowed to detect when enamel Ca isotope composition could be affected by milk consumption or mineral supplementation. Thanks to these advances, we collected new data from the Pleistocene reindeers of Jaurens, which are now consistent with their known low trophic level and plant-based diet. This demonstrates that disentangling ecological and physiological signals within enamel Ca isotope compositions is possible by using serial micro-sampling. This approach allows retrieving accurate trophic information from this proxy and truly pushes forward the limits of Ca isotope applications regarding paleoecology, physiology and ethology.

Introduction

An increasing number of studies suggest that the stable Ca isotope composition of the hydroxylapatite of vertebrate bone and teeth, represents a record of diet within modern and fossil skeletal remains (Clementz et al., 2003; Chu et al., 2006; Reynard et al., 2010; Heuser et al., 2011; Clementz, 2012; Martin et al., 2015, Martin et al., 2017a, Martin et al., 2017b, Martin et al., 2018; Hassler et al., 2018). Body Ca originates mainly from food and water for land animals, but the Ca isotope composition of animals diverges from their food. At the opposite of carbon and nitrogen patterns, heavy Ca isotopes are discriminated against light Ca isotopes during their routing from food to tissues, notably in bone and enamel (Skulan and DePaolo, 1999; Chu et al., 2006; Hirata et al., 2008; Tacail et al., 2014; Heuser, 2016; Heuser et al., 2016). This results in a trophic level effect (hereafter TLE) with bones of herbivorous animals exhibiting more ⁴⁴Ca-depleted compositions compared to the plants they consume, and carnivorous predators following the same trend compared to their prey. This ultimately leads to an isotopic clustering of animal taxa as a function of their trophic level, with primary consumers exhibiting a heavier Ca isotope composition than tertiary consumers, and secondary consumers exhibiting an intermediate isotopic composition. This has been observed in land ecosystems (Chu et al., 2006; Reynard et al., 2010; Martin et al., 2017a, Martin et al., 2018; Hassler et al., 2018; Dodat et al., 2021) and marine ecosystems (Clementz et al., 2003; Clementz, 2012; Martin et al., 2015, Martin et al., 2017b), even though for marine animals the Ca originating from seawater likely buffers dietary Ca intakes. These studies are based on ⁴⁴Ca/⁴²Ca or ⁴⁴Ca/⁴⁰Ca analyses of the hydroxylapatite of bone and teeth, commonly expressed as δ^44/42Ca and δ^44/40Ca, respectively (equivalent to the variation in ‰ compared to the Ca isotope ratios of a reference material, further detailed in Section 2.6). Moreover, this technique allows to study the ecology of both modern and fossil specimens thanks to the good preservation potential of such mineralized tissues and their Ca (Heuser et al., 2011; Martin et al., 2017a). This body of studies supports that Ca isotopes are a promising tool for diet and trophic inferences in modern and paleontological contexts with a large temporal range of action, providing the fact that mineralized tissues are preserved. However, inferring a TLE within mammalian communities relies on first order observations of Ca isotopic variability among predators and their prey. As highlighted in previous work, many uncertainties remain concerning the mechanisms behind fractionation processes as related to physiology versus environmental sources. As such, not all taxa are strictly following the theoretical isotopic/trophic clustering in the faunas studied so far. For some specific faunas and trophic niches, diet seems hard to constrain with Ca isotopes only (Reynard et al., 2010; Melin et al., 2014). Moreover, bone or enamel Ca isotope compositions sometimes overlap between herbivores and predators, with herbivores occasionally exhibiting isotope compositions more ⁴⁴Ca-depleted than predators from the same fauna (e.g. hippopotamidae, mammoths and cervidae; see Martin et al., 2017a, Martin et al., 2018; Dodat et al., 2021). This raises important questions about what can generate such issues, and highlights how critical it is for the accurate use of this proxy to reconcile the evidences of TLE with the occasional decoupling recorded between trophic level and Ca isotope compositions.

The isotopic offset between diet and bone which generates the TLE (Δ^44/42Ca_diet-bone) is relatively constant among mammal species with a value of −0.54 ± 0.08‰ (2 standard error, 20 individuals from 6 mammal species; reviewed in Tacail et al. (2017)), despite resulting from the combination of numerous body Ca fluxes associated with Ca isotope fractionation (Skulan and DePaolo, 1999; Chu et al., 2006; Hirata et al., 2008; Tacail et al., 2014; Heuser et al., 2016; Tacail, 2017). The most impactful fluxes identified so far are the kidney Ca reabsorption from primary urines (Skulan et al., 2007; Heuser and Eisenhauer, 2010; Morgan et al., 2012; Tacail et al., 2014; Channon et al., 2015; Heuser et al., 2016, Heuser et al., 2019; Eisenhauer et al., 2019), the milk production and excretion (Chu et al., 2006; Hassler et al., 2021, Reynard et al., 2010) and the bone mineralization (Skulan and DePaolo, 1999; Skulan et al., 2007; Heuser and Eisenhauer, 2010; Reynard et al., 2010; Morgan et al., 2012; Channon et al., 2015), although the significance of this last parameter has been recently questioned (Tacail, 2017; Hassler et al., 2021, Tacail et al., 2020). Changing these fluxes, like during gestation, lactation (Ramberg Jr et al., 1970; Cross et al., 1995; Giesemann et al., 1998; Karlsson et al., 2001; Wysolmerski, 2002; Vanhouten and Wysolmerski, 2003; Gallego et al., 2006; Kovacs and Fuleihan, 2006; Tacail, 2017) or antlerogenesis (Mitchell et al., 1976; Muir et al., 1987a, Muir et al., 1987b), likely modifies the Ca isotopic equilibrium of the organism, the resulting Δ^44/42Ca_diet-bone offset, and could generate TLE discrepancies. Alternatively, the consumption of milk during nursing can also generate TLE discrepancies by changing the diet Ca isotope composition of non-weaned individuals compared to weaned individuals (Chu et al., 2006; Li et al., 2016, Li et al., 2020; Tacail et al., 2017, Tacail et al., 2019). Finally, Ca enriched mineral supplementation such as with mineral licks and osteophagia, as well as the Ca isotopic variability inherent to plants (Holmden and Bélanger, 2010; Gussone and Heuser, 2016; Schmitt, 2016; Moynier and Fujii, 2017; Martin et al., 2018; Griffith et al., 2020), are also able to blur the trophic clustering of Ca isotope compositions. The aim of this study is thus to investigate how these different factors can affect bone and enamel Ca isotope compositions in modern animals, then to use this background to unravel a case study of TLE discrepancy previously documented.

We carried out the first part of this project by monitoring the bone and enamel Ca isotope compositions of a modern cervid population (red deer, Cervus elaphus) inhabiting the Bauges Natural Regional Park (NRP), Alps, (Savoie, France). This modern population then served as a model to discuss TLE discrepancies, and more precisely to discuss the case of the reindeers (Rangifer tarandus) from the Pleistocene locality of Jaurens (Corrèze, France). Martin et al. (2017a) showed that reindeers from this locality (dated between 32.6 and 29.7 kyr BP, Guérin et al., 1979) exhibit ⁴⁴Ca-depleted compositions in tooth enamel down to a δ^44/42Ca value of −1.75 ± 0.09‰ (2 standard deviation; 3 individuals). This is different from the Ca isotope compositions of the other herbivores of this locality, and closer to the one of lions and wolfs (Martin et al., 2017a). In accordance with the documented effects of nursing on the body Ca isotope composition of the young (Chu et al., 2006; Li et al., 2016, Li et al., 2020; Tacail et al., 2017, Tacail et al., 2019), Martin et al. (2017a) proposed that these negative δ^44/42Ca values could result from the consumption of maternal milk by reindeers at the time when their analyzed molars were mineralizing (second molars, M2). Although this hypothesis is tempting, no direct evidence was available to prove that this phenomenon was responsible for the ⁴⁴Ca-depleted isotope composition reported among Jaurens fossil cervids, leading to this new investigation.

To ensure the completion of our study, it was necessary to assess at which degree the Ca isotope composition of bone and enamel are comparable. Previous studies found differences in δ^44/42Ca values between bone and enamel or bone and enameloid (Heuser et al., 2011; Tacail et al., 2014; Martin et al., 2015, Martin et al., 2017a), but it was unclear whether this offset (Δ^44/42Ca_bone-enamel) was due to different Ca isotope fractionation coefficient during mineralization (α _blood-bone and α _blood-enamel), to different mineralization timing, or to diagenesis. Our modern red deer population is free of diagenetic influence, and therefore, mineralization timings and differences between α _blood-bone and α _blood-enamel are the only factors affecting bone and enamel Ca isotopic differences. In other words, characterizing α _blood-bone and α _blood-enamel in red deer allow to accurately compare the different ontogenetic time periods recorded by these two tissues.