Alba Regia. Annales Musei Stephani Regis. – Alba Regia. Az István Király Múzeum Évkönyve. 21. 1981 – Szent István Király Múzeum közleményei: C sorozat (1984)

Közlemények – Mitteilungen - Bartosiewicz László: Csabdi–Télizöldes: Taphonomy in the Western Section of the Neolithic Site. p. 235–240.

ELEMENT max - SPECIMEN max Figure J : Extremes, between which the sampling of a mammalian skeleton with no replacement may occur. The continuous line symbolizes an ideal case on the basis of initial probabilities (Table 1), while the dashed line shows a theoretical situation where the inventory of new elements is exhausted step by step in the first phase of sampling. Empirical values can only fall between the two curves. Material and Method For the purposes of this paper material from the western section of the settlement was chosen, and animal bones from ten squares and associated pits were evaluated. Among the species identified wild and domestic cattle, red deer and wild swine were grouped together as large mammals. Small mammals included domestic swine, sheep and goat, roe deer and sporadic bones of other species such as dog and beaver. In order to reduce the number of zeroes potentially increasing the number of useless combinations and thus biasing the matrix used for calculations, further abstraction was necessary in the presentation of skeletal elements. All kinds of bones were categorized in three groups according to the quality of meat they may have carried. This classification largely corresponds to that of Uerpmann (1972). Modifications in his system were not made in order to correct its few disadvantages in interpretation (Bartosiewicz 1982), which are only of minor significance in this study. For purely technical considerations however, all long bones of the skeleton were placed in the same category (B). This was intended to solve the problem raised by the great number of unidentified long bone splinters which potentially fall into two categories (upper limb bones: A, lower arm and tibia : B) in Uerpmann's classification. All computations were executed using the B1MED procedure revised in 1981 (BMDP 1981). Results Table 2 contains the univariate statistics of bone remains found in all proveniences and the sub-division of the sample into excavation units and features respectively. The overall sample is heavily dominated by large mammal bones of the В category which are most numerous in the skeleton (ribs, long bones) and represent medium value meat. This group of bones also dominates among the remains of small mammals, although averages only half of the previous class of skeletal elements. Bone remains of small mammals are even more under-represented in the A and С categories. In both cases the remains of small mammals reach only one quarter of the mean value characteristic of large mammals. The occurrence of A class elements (vertebrae, flat bones) is least consistent: the standard deviation of this variable exceeds its mean value in the group of small mammals. The number of skeletal elements represented in all proveniences averages 25 with a 0.27 coefficient of variation. When the univariate statistics on material from artificial excavation units and original features are compared it turns out that pits usually contain twice as many bones in all categories as the rest of the squares opened during the excavation. The average number of skeletal elements represented in the pits on the other hand is only slightly higher. Quite a few high and significant correlations may be observed between the least variable bone groups in the overall sample. The number of elements is positively correlated with the amount of bones in the respective categories (Table 3 ). High correlations are much less frequent in the sub-divided sample (Table 3 ). The fewer number of specimens offer particularly less possibility for significant positive combinations in the area between features. It is only the correlation between large mammalian В and С bones and that between large mammal remains and the number of skeletal elements which are significant on the required level. In pits the bones of large mammals usually are significantly correlated with the majority of variables. Even the small mammal A category is correlated with the group of small mammal С bones. No differences of the mean values are significant. However, variances of large mammalian remains differ significantly in the two types of provenience. Conclusions Bone distribution patterns revealed by univariate statistics and analyses of correlation at this site suggest that large mammals were a major resource of animal protein. This is however, chiefly due to the fact that the bones of such animals have much more meat on them than identical skeletal parts of smaller mammals. Bones of this latter group may be slightly under-represented. The numeric dominance of large mammal bones may be overemphasized by the following facts : 1 Sometimes bones of smaller animals (especially those of young individuals) tend to entirely disintegrate and thus disappear. This is less likely to be the case in the loess of the site, which rather favors the formation of epigenic limestone concretions on the surface of bones. 2 Among bones which survive however, those from smalt animals tend to be preserved better. Beyond a certain size bones are more prone to fragmentation (Binford

—Bertram 1977) and thus the proportion of large animal bones inevitably increases. 236