**Notes:** Results from one-way ANOVA with Tukey's Honestly Significant Difference post-hoc comparisons for those variables with equal variances and Games-Howell post-hoc comparisons for those variables with unequal variances. Because orbit height, orbit area, and malar height all scale allometrically, the most meaningful comparisons are among taxa of similar body size. The estimated mass for *P*. *angusticeps* averages ~21 kg for males and ~15 for females [39]. *P*. *izodi* is estimated at ~20 kg for males and ~15 for females [39]. The most similar extant taxon in terms of body mass is *P*. *h*. *cynocephalus*, ~23 kg for males and 12.5 kg for females, which is why *P*. *h*. *cynocephalus* is used in the above comparisons. All specimens were pooled regardless of sex in order to increase sample size. For sex-specific values, see Table 3. n.s. = non-significant. Note that *P*. *angusticeps* and *P*. *h*. *cynocephalus* are both significantly different from *P*. *izodi*, but not from each other. Results for all comparisons are the same if UW 88–886 is included in the *P*. *angusticeps* sample. Orbit height defined as the maximum distance between the inferior and superior orbit borders. Orbit width defined as the maximum distance between the lateral and medial orbit borders. Orbit area is defined as orbit width x orbit height. Malar height defined as the distance between orbitale inferior/zygoorbitale and zygomaxillare inferior. Relative malar height defined as malar height/orbit height. *P*. *angusticeps* specimens include CO 100, CO 135A/B, CO 101, GV 4040, and HGD 1249. *P*. *izodi* specimens include TP 12, SAM 11728, T10, T13, UCMP 125854, UCMP 125855, UCMP 125856, STS 262, T89-11-1, and SWP UN-2. Values for each taxon represent averages. Numbers in parentheses represent estimates. For boxplots with ranges, see also Fig 4 and Table A in S1 Dataset.

Comparison of selected morphological features in "small-bodied" *Papio* species.