THE COLLECTION AND PRESENTATION OF TWO AND THREE DIMENSIONAL BONE ORIENTATION DATA

Orientation data should be collected for fossils as part of routine quarry documentation, and as a means of potentially determining assemblage taphonomic history (e.g. fluvial or human transport). Orientation data should be recorded prior to excavation and removal of objects because removing material from quarries permanently destroys the relative and absolute orientations of skeletal material.

Since bones are three dimensional, at least three pieces of information are necessary to constrain their orientations. For irregularly shaped bones (e.g. broken vertebrae or skulls), three mutually orthogonal descriptors can be used. For elongate bones (e.g. femora or humeri), the trend and plunge can be measured and the polarity (which end faces which direction) reported. Consistent polarity conventions should be used and reported so other investigators can interpret the data. Similarly, the strike and dip of flat bones (e.g. scapulae or neural spines) can be measured with a description of which side faces up or down.

Rose diagrams should not be used to illustrate bone orientation data since only two dimensions can be presented out of the three required to constrain bone orientation (e.g. plunge and dip data are impossible to plot). Since data amalgamation is required to construct a rose diagram, it is difficult to present bone polarity data and subdivide the data for finer scale interpretation (e.g. plot orientations of just femora or humeri). Similarly, changing the number of degrees in a petal and starting location alters the final figure, thus rose diagrams are subjective data presentation tools that can cause an investigator to reach erroneous conclusions.

Stereonets should be used to present bone orientations because they plot data in three dimensions (including plunge and dip). If an assemblage has little dip or plunge (> ~10^o), then a corona dot diagram or any of its derivatives can be used. Both stereonets and corona dot diagrams can plot individual data points without binning data, consequently they produce a consistent unique data representation. Presenting polarity data and illustrating subdivided data are easy with stereonets and corona dot diagrams by changing symbol shapes and colors. A 360^o plot should be used instead of a half diagram; and data contouring can provide a more elegant and informative presentation.