Andrew Sillen

87Sr/86Sr ratios in modern and fossil food-webs of the Sterkfontein Valley: implications for early hominid habitat preference

This research addresses the potential contribution of strontium isotopes to the reconstruction of early hominid behavior at the Swartkrans site in the Sterkfontein Valley of Gauteng Provence (formally known as the Transvaal), South Africa. We report that, while there is considerable variability in the 87Sr/86Sr of whole soils within a 15 km radius of this site, available soil and grassland plant 87Sr/86Sr is much less variable and generally above 0.730. This value is higher (more radiogenic) than the 87Sr/86Sr of plants growing within the greenbelt surrounding the Blaaubank stream adjacent to Swartkrans and streamwater itself (0.721). The difference between grassland and riparian strontium isotope composition suggests a method for determining habitat utilization by early hominids. In this study, a geological explanation for a natural difference between Blaaubank stream and grassland Sr is suggested, based on relatively less radiogenic Sr (having lower 87Sr/86Sr values) in the carbonate component of the local dolomite when compared to other nearby geological formations. The explanation was tested initially using a top-down approach in which the 87Sr/86Sr ratios of water, soil, and plants from the entire Blaaubank catchment were measured. Next, a bottom-up approach was used to examine Swartkrans Member I faunal species known to have obtained their Sr from well-defined habitats. The results are that (1) pollution is not the explanation for the relatively low 87Sr/86Sr ratios of the Blaaubank stream, (2) Swartkrans Member I carbonate has a similar 87Sr/86Sr to that of modern Blaaubank water, indicating that relationships seen today existed in the Pleistocene, and (3) Pleistocene riparian fauna have relatively low 87Sr/86Sr ratios when compared to fauna adapted to drier habitats. Together these results make it possible to interpret the strontium isotope composition of Pleistocene early hominids from Swartkrans in terms of habitat utilization.

87Sr86Sr as a dietary indicator in modern and archaeological bone

Strontium isotope ratios (87Sr86Sr) in human and animal bone are inherited, via foodstuffs, from rocks in the areas in which an individual lived. If such an area includes more than one rock type with differing 87Sr86Sr, or is coastal region with terrestrial 87Sr86Sr different from that of the sea, then bone strontium isotope ratio measurements provide a measure of the relative importance of foods from each isotopic zone. In the southwestern Cape of South Africa, marine and coastal terrestrial areas (covered by marine sands) should yield foods with 87Sr86Sr close to the marine average, while the Cambrian and pre-Cambrian shales and sandstones of the hinterland ought to be enriched in 87Sr (from radioactive decay of 87Rb during the considerable time elapsed since they were laid down). Analyses of modern animal bones show that these patterns do prevail, furnishing us with a new means of reconstructing prehistoric diets. Preliminary measurements on archaeological human material recovered from coastal shell-midden sites apparently reflect heavy reliance on coastal foods. However, these bones have been shown to be contaminated with diagenetic marine strontium from the shells in the middens. Experiments on faunal specimens show that diagenetic strontium can be removed by repeated washing in acetic acid/sodium acetate buffer. Mineral with 87Sr86Sr similar to the burial environment dissolves first, while subsequent washes remove strontium with different 87Sr86Sr. For two coastal human skeletons, however, the marine/coastal terrestrial nature of the isotopic signature remains. Further experimental work is required in order to establish the limits of the technique; however, a combination of the solubility profile method and strontium isotope analysis promises to provide a powerful tool for reconstructing palaeodiets.

Strontium-calcium ratios (Sr/Ca) of Australopithecus robustus and associated fauna from Swartkrans

Abstract Strontium-calcium ratios (Sr/Ca) are normally reduced at higher trophic levels in foodwebs, due to discrimination against strontium in favour of calcium by animals. This phenomenon has not generally been applied to the study of fossil foodwebs and the diets of early hominids because of diagenetic changes which obscure or obliterate biological Sr/Ca. The examination of compartments of fossil apatite having differing solubility, however, is a promising method for independently measuring biological and diagenetic Sr/Ca. In this study, Sr/Ca in Member I fossils from the site of Swartkrans were examined using a solubility profile procedure. Sr/Ca relationships observed among Swartkrans fauna match those seen in modern African foodwebs, suggesting that biological Sr/Ca accounts for the observed variation. When specimens of the fossil hominid Australopithecus robustus were examined, Sr/Ca values were inconsistent with that of a root, rhizome or seed-eating herbivore, suggesting that the diet of this species was more diverse than previously believed, and almost certainly included the consumption of animal foods.

Chemistry and Paleodietry Research: No More Easy Answers

While isotopic and elemental analyses of prehistoric skeletons have made an important contribution to paleodietary research over the last 10 years, certain problems in the application of these techniques only now are emerging. These problems, affecting both isotopic and trace-element studies, mainly are due to the peculiar interdisciplinary nature of the field, rather than to any technological barrier. With minor exceptions, techniques developed largely in other sciences have been grafted on to archaeological problems. This no longer suffices because gaps remain in the scientific grounding of these techniques that need to be addressed before more complicated archaeological questions can be resolved. While the necessary studies may seem to be of little immediate anthropological interest, they are vital if continued progress in paleodietary research is to characterize the years ahead.

Solubility profiles of synthetic apatites and of modern and fossil bones

Abstract Strontium has been shown to vary within trophic levels in natural foodwebs so that previous criteria used to identify biogenic strontium in fossils, namely consistent differences between herbivores and carnivores, may be inappropriate. In this article, two alternative approaches are explored. These include first, experimental studies of the behaviour of synthetic apatites in the solubility profile system, and second, solubility profile studies of fossils from the Pleistocene site of Swartkrans. The synthetic apatite studies show that simple mixtures of apatites dissolve discretely in the system. Study of Swartkrans specimens reveals that, while diagenetic Sr is clearly present, some recoverable Sr varies in a manner similar to that seen in modern African foodwebs, and is therefore unlikely to be of diagenetic origin.

Sr and Sr/Ca in marine and terrestrial foodwebs in the Southwestern Cape, South Africa

Abstract Strontium and calcium have been measured in a range of plants and animals (both marine and terrestrial) from the southwestern Cape of South Africa as part of an investigation of modern and prehistoric foodwebs in the region. First, the meat of marine molluscs and crustaceans are shown to have Sr and Sr/Ca values comparable to those of terrestrial plants. Thus, the consumption of these marine foods in this region cannot produce the markedly elevated Sr levels seen in archaeological human skeletons from coastal sites; such levels are shown to be a diagenetic phenomenon. Second, reduction in Sr/Ca in higher trophic levels is seen only when predators are compared with their prey. However, individual herbivore or carnivore species cannot be taken to represent other animals in their respective trophic level. These data imply that Sr/Ca is inappropriate for determining meat intake in complex prehistoric human diets in this region. The technique may be more useful in examining specific prey-predator relationships, including those in the early hominid fossil record.

Trace element and isotopic aspects of predator-prey relationships in terrestrial foodwebs

Abstract Strontium-calcium ratio (Sr/Ca) and stable carbon isotope ratio (13C/12C) analyses of bone and enamel apatites may be used for reconstructing predator-prey relationships in vertebrate fossil assemblages. The approach is based on the rationale that 13C/12C in predator apatite closely resemble those of their prey, while Sr/Ca is predictably reduced in specific predator-prey pairs. As an example, we compare the relationship of the extant leopard, Panthera pardus, and its preferred prey in modern foodwebs, with the relationship of fossil leopards (from the Member 1 fossil faunal assemblage at Swartkrans Cave), and their likely prey. The results suggest that Swartkrans Member I leopards depended heavily on Papio baboons, with some lesser contribution from Hyrax (Procavia sp.). The approach has considerable potential for elucidating the preferred prey of extinct predators such as sabertooth cats.