Isotopic analysis of tooth enamel carbonate from modern North American feral horses: implications for paleoenvironmental reconstructions

摘要

The accuracy of paleoenvironmental reconstructions based on isotope analyses of equid teeth is currently uncertain because the exact relationship between the isotope composition of modern feral equids and their environment has not been thoroughly studied. We analyzed the carbon and oxygen isotope values (δ~(13)C and δ~(18)O) of tooth enamel carbonate and the δ~(13)C values of fecal samples from modern feral horses. We compared those values with the δ~(13)C values of local vegetation and the δ~(18)O values of local waters. Herds were studied in two contrasting localities: eastern Oregon, where grasslands consisted of 100% C_3 species, and New Mexico, where >95% of the grasses were C_4 species. Carbon isotope analyses of fecal material and tooth enamel suggest that horses consumed primarily grass, but some New Mexico horses also consumed significant amounts of shrubs and/or forbs. Microhistological analyses of fecal samples show that Oregon horses consumed 95% grass, and Oregon enamel δ~(13)C values are consistent with a diet containing 100% C_3 plants. Microhistological analyses of fecal samples from New Mexico indicate a diet averaging 75% grass, while enamel δ~(13)C values suggest that diets averaged 85% C_4 plants (range=72–97%). Thus, reconstructions of the C_3/C_4 ratio of grasses in ancient grasslands that are based on the δ~(13)C values of fossil equid teeth may underestimate the abundance of C_4 grasses. The mean δ~(18)O values of tooth enamel paralleled the trends observed in the mean δ~(18)O values of precipitation. However, the mean δ~(18)O values of enamel carbonate from Oregon and New Mexico differed by only 3.3‰, which is less than the difference in the mean δ~(18)O values of precipitation (6.5‰). In addition, the range of δ~(18)O values within New Mexico enamel samples (6.5‰) was greater than the difference between mean enamel 18O values at each site. Calculated values for the δ~(18)O of water ingested by horses are 2–3‰ more positive than mean δ~(18)O values for corresponding precipitation, suggesting that horses consumed waters that were enriched in δ~(18)O due to evaporation. While our results confirm that local climatic and hydrological conditions can influence the δ~(18)O values of equid enamel, they also show that the δ~(18)O values of equid teeth are not always a direct proxy for the isotope ratios of precipitation.