SKELETAL COMPLETENESS

We study the processes of laminae formation in the Oligo-Miocene volcano-sedimentary palaeolakes of the Gürçü-Dere Valley (Anatolia, Turkey). These deposits contain a rich macro- and micro-flora and vertebrate and insect fauna. The deposition and fossilisation of these organisms' remains are related to the development of the lakes' surface and bottom microbial mats. These strata are made of two types of rhythmite or varve, imitating dark and clear laminae sequences. The first type is (detritic + organic material) / diatoms at the Bes-Konak palaeolake, probably not of seasonal origin (rhythmite), but providing evidence of short ecological 'mini-catastrophes', a few weeks long, in confined environments. The second type is (diatoms + organic matter) / dolomite at the Iybeler palaeolake, probably of seasonal origin (varve). For the first time, we show that fossil insects, i.e. the Notonecta spp., are highly relevant to determining the relative duration and mechanisms of the depositions of these laminae.

Numerous experiments on modern bone assemblages have demonstrated that carnivores preferentially ravage bone portions that have high fat content such as ribs, vertebrae, pelves, and the spongy ends of long bones. If marrow is present in long bones, carnivores will break them to access it. If humans remove the marrow first, carnivores typically ignore midshaft fragments and focus on spongy portions that contain bone grease. These portions are swallowed and the grease is extracted within the gut. Many modern human groups regularly implement bone boiling technology to extract grease, and this technology is also known to have existed during the Late Pleistocene and Holocene - particularly among high-latitude hunter-gatherer groups. Because most actualistic studies of bone surface modification and carnivore bone portion choice are based on unboiled bones, the results may not be applicable to archaeological assemblages in which bone boiling technology was utilized. An experimental evaluation of the effects that bone boiling has on carnivore bone portion choice and relative proportions of human and carnivore bone surface modification indicates that; 1) The extraction of bone grease from bone ends during boiling, particularly prolonged boiling, causes carnivores to become less selective in bone portion choice; and 2) As a likely result of this decreased selectivity, the relative proportions of tooth- and percussion-marked midshaft fragments in heavily boiled assemblages (those boiled longer than 10 hours) is not likely to remain a reliable indicator of the timing of carnivore and human interaction with an assemblage.

Differences amongst ungulate body-size classes in archaeofaunal assemblages are frequently used to infer economic structure, human transport and processing decisions, or pre-burial taphonomic processes. It was found that bones of larger ungulates (fallow deer, Dama mesopotamica) are generally more fragmented than analogous elements of smaller ungulates (mountain gazelle, Gazella gazella). This pattern is consistent across several Levantine Paleolithic and Epipaleolithic bone assemblages and does not stem from recovery or identification biases, or from animal-induced taphonomic causes such as carnivore ravaging. We suggest that the greater fragmentation of larger animals is an artifact of either differential human processing of large and small animals, or of post-depositional attritional processes. Our analysis points to the greater likelihood of the latter scenario. We conclude that inter-taxonomic comparisons should consider the possibility that key zooarchaeological measures may be biased due to differential size-related fragmentation, affecting inferences on human behavior, taphonomic processes, and paleoenvironments.