Skip to main content

Table 9 Pros and cons of various methods of body mass estimation discussed in this paper

From: Occipital condyle width (OCW) is a highly accurate predictor of body mass in therian mammals

Variable

Pros

Cons

Pros and cons of various methods of body mass estimation discussed in this paper

Dental variables

(e.g., molar row length)

1. Abundant in the fossil record and known for almost every fossil mammalian taxon

2. Easy to obtain large sample sizes

3. Dimensions of permanent teeth in mammals always reflect adult size rather than growth stage due to diphyodonty

1. Extremely morphologically variable and often cannot be applied across taxa (e.g., edentulous taxa)

2. Little reason to believe they necessarily correlate better with mass than other variables

3. Cannot be used to reconstruct growth patterns in fossil mammals

4. May be biased by head size

Cranial variables

(e.g., skull length)

1. Does not require associated postcrania

2. Slightly less biased by dietary habits than dental variables

1. Can be highly influenced by ecology and non-isometric allometry (e.g., rostrum length)

2. Often fails to produce accurate estimates if head is disproportionately large/small relative to body size

OCW

1. Highly conserved across Mammalia

2. Strong theoretical reasons to believe OCW correlates with spine size, and therefore body size

3. Unlike other craniodental variables is not biased by skull size

4. Only requires skull rather than partial skeleton

1. Not as well studied, so potential confounding factors and interspecific variation less understood

2. Sample sizes can be small because skulls may have damaged occiputs

3. Potential concerns with covariation with brain size

Head-body length

1. Probably the strongest theoretical reasoning as to why this variable should most closely approximate mass [66, 161]

2. Extremely large sample of published extant comparative data with associated mass

1. Requires rare, nearly-complete specimens, much less tolerant of missing data than any other method

2. Head-body length highly subject to subjective decisions and almost always requires some degree of estimation, even near-complete specimens are often missing vertebrae that must be filled in from close relatives

Limb elements

1. Strong biomechanical reason to believe they correlate with body mass [87, 162]

1. Limb bones rarely associated with diagnostic specimens

2. Can be biased when phylogenetic/paleoecological signal overrides size-based signal (e.g., fossoriality or differences in weight distribution between clades [31, 87];)