Developing Stature Estimation Regression Formulae for the Arikara of the Larson and Leavenworth Sites

Stature is an important descriptive characteristic of an individual. Living stature cannot be measured directly in archaeological populations and thus must be estimated by bone. Current stature estimation formulae cannot be used with archaeological populations because the relationships between stature and the length of the various bones used in estimation differ among races and populations. Similarly, secular change in stature makes the use of formulae derived from modem populations on archaeological groups problematic.

Anatomical methods of estimating stature account for the skeletal elements that contribute to an individual's height and provide an estimate of the soft tissue component of stature. These methods give results very close to properly recorded living statures. In populations for which recorded statures or cadaver lengths are unavailable, anatomically estimated statures can be substituted for living statures for the purpose of creating stature estimation formulae.

Living stature was estimated for 48 individuals (25 females, 23 males) excavated from the South Dakota Larson (39WW2) and Leavenworth (39C09) sites using Fully's anatomical method. These sites represent two Arikara villages from the Great Plains protohistoric and historic phases, respectively. The Fully method estimates were treated as living statures for the purpose of regression. Linear regression formulae were calculated using the Fully estimates and basion-bregma height, bicondylar length of the femur, condylo-malleolar length of the tibia, vertebral column height, height of the articulated talus, and calcaneus, maximum humerus length, and maximum radius length.

The regression formulae were calculated in two ways. In the first manner, classical calibration, the bone or bones are regressed onto stature and the equation is solved for stature. This method creates general formulae that should be applicable to many American Indian and other indigenous American populations. In the second manner, inverse calibration, stature is regressed onto the bone or bones. This method ideally gives more precise but population specific formulae.

To illustrate the importance of population specificity, the long bone measurements for the Arikara were used in stature estimation formulae developed from closely related groups including Illinois Middle Mississippians, Ohio Valley natives from the archaic through protohistoric periods, modem Mesoamericans, and modern "Mongoloid" and Mexican populations.

The goals of this study are 1) to develop population specific formulae for use in estimating stature in incomplete skeletal remains of the Arikara and closely affiliated groups; 2) to develop general stature estimation formulae for potential use in other indigenous America groups; and 3) to argue for the importance of using population specific formulae when estimating stature.

The results of this study show that for this population, the best estimator of living stature in females is bicondylar femur length, followed closely by the combined femur and tibia lengths. For males, the best estimator was the combination of the femur and tibia lengths with the next best estimator being femur length alone. In both sexes the vertebral column and humerus and radius estimate well. Basion-bregma height did not work at all in females and only marginally in males. Articulated talus-calcaneus height produced regression formulae of questionable utility.