Abstract
Centimetre-scale laminae in tusk and molar dentine of late Pleistocene mastodonts and mammoths have been interpreted as annual growth bands produced, in part, by seasonal variation in growth rate. To test this interpretation, we measured the oxygen isotope composition (δ18O) of the CO3 fraction of dentinal hydroxyapatite from samples covering consecutive inferred years of growth in tusks. In mammals, changes in the δ18O value of dental tissues within individuals predominantly reflect variation in the δ18O value of body fluids, which is controlled mainly by the isotopic composition of ingested water. In Northern Hemisphere continental regions, winter precipitation has substantially lower δ18O values than does precipitation in other seasons. If ingested water tracks focal precipitation, then seasonal variations in dentinal isotope composition should result, the lowest δ18O values representing winter growth. We demonstrate that there are substantial variations in the oxygen isotope composition of proboscidean dentinal apatite, and that isotopic identifications of winter (i.e., low δ18O values) coincide with those based on growth rate (i.e., slow-growth zones). Finally, the potential of oxygen isotope analyses of terrestrial mammals for assessing the seasonality of paleoclimates is considered.