Both ubiquity and percentage by weight were used in the above analysis to interpret the relative importance to site inhabitants of various kinds of plant foods. Each of these comparative techniques is useful in different contexts. Ubiquity, for instance, ignores mass in favor of frequency of occurrence and can be used for comparing plant remains that have different probabilities of being preserved in large quantities or that vary greatly in mass and food-to-waste ratio. Therefore, it might be expected that the two methods would yield very different rankings of resources. However, this is not the case for the resources categorized separately by weight (e.g., hickory, acorn, walnut, corn, peach, bean, and cucurbit) (see graph). For this illustration, ubiquity rankings have been determined with respect only to the taxa used in the figure, and not to seed taxa, which are not itemized by weight. The close correspondence between ranks using these two methods could mean that interpretations based on them are likely to be more secure than if they were widely divergent.
But what if ubiquity and percentage ranks are similar simply because the same biases have affected both cases? It is apparent that in most cases the most highly ranked species are also ones with remains that have relatively high preservability (e.g., hickory shell, walnut shell, corn). However, there are notable exceptions. Acorn rank is high by both methods, despite its low preservability relative to, say, walnut. Grape ranks as high as walnut shell by ubiquity (Table 30), although seeds are in general less preservable than thick nutshell (although fruit-drying over a fire, if practiced, would make grape seeds more likely to be preserved). Also, most of the highly ranked and highly preservable plant types are those mentioned as important foods ethnohistorically. Thus, the close correspondence of rankings lends some strength to the interpretations presented above. But if correspondence is this close, what is the utility of using both methods? The biggest advantage of using both methods is that ubiquity allows for comparison of classes (e.g., seeds and nutshell) that are quantified differently (by count versus by weight) because of their very different physical characteristics.
