Environmental, trophic, and ecological factors influencing bone collagen δ2H

Abstract

Organic deuterium/hydrogen stable isotope ratios (i.e., 2H/1H, expressed as δ2H value in ‰) in animal tissues are related to the 2H/1H in diet and ingested water. Bone collagen preserves the biochemical 2H/1H isotopic signal in the δ2H value of collagen’s non-exchangeable hydrogen. Therefore, δ2H preserved in bone collagen has the potential to constrain environmental and trophic conditions, which is of interest to researchers studying of both living and fossil vertebrates. Our data examine the relationship of δ2H values of collagen with geographic variation in δ2H of meteoric waters, with local variations in the ecology and trophic level of species, and with the transition from mother’s milk to adult diet. Based on 97 individuals from 22 marine and terrestrial vertebrates (predominately mammals), we found the relationships of collagen δ2H to both geographic variation in meteoric water δ2H (R2=0.55) and to δ15N in bone collagen (R2=0.17) statistically significant but weaker than previously reported. The second strongest control on collagen δ2H in our data is dietary, with nearly 50 percent of the variance in δ2H explained by trophic level (R2=0.47). Trophic level effects potentially confound the local meteoric signal if not held constant: herbivores tend to have the lowest δ2H values, omnivores have intermediate ones, and carnivores have the highest values. Body size (most likely related to mass-specific metabolic rates) has a strong influence on collagen δ2H (R2=0.30), by causing greater sensitivity in smaller animals to seasonal climate variations and/or high evapotranspiration leading to 2H-enrichment in tissues. In marine mammals weaning produces a dramatic effect on collagen δ2H with adult values being universally higher than pup values (R2=0.79). Interestingly, the shift in δ15N at weaning is downward, even though normally hydrogen and nitrogen isotope ratios are positively correlated with one another in respect to trophic level. Our findings suggest that in carnivores, which have an especially high variance in δ2H, large samples are needed to separate signals from precipitation, trophic level, body size, and age. For δ2H of fossil collagen to be useful as a proxy of environmental or dietary information, these confounding effects need to be understood, which means careful selection of a study species. Further, δ2H from a single fossil bone collagen is likely to be uninterpretable.