Human Adaptation to Tropic Environments: A Comparison of West African and Melanesian Nasal Morphology

• Published in: The FASEB Journal
• Main Authors: Cooper, Amber R., Maddux, Scott D.
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2020
• Subjects: Arctic
• Online Access: http://dx.doi.org/10.1096/fasebj.2020.34.s1.04248

It is generally accepted that humans indigenous to hot‐humid equatorial environments exhibit superoinferiorly shorter, mediolaterally wider, and more anteroposteriorly attenuated nasal passages compared to counterparts from colder and/or drier environments. This relatively short/wide/attenuated morphology has been argued to functionally minimize nasal airflow resistance in tropical environments that require little intranasal air‐conditioning (i.e., exchange of heat and moisture transfers between nasal mucosa and respired air), and/or to facilitate increased respiratory heat‐shedding in high humidity environments where sweating is a less effective thermoregulatory mechanism. However, studies of climatic adaptation in human nasal morphology have often employed West Africans as representative of all equatorial populations. Yet, the assumption that West African nasal morphology is actually comparable to non‐African equatorial populations has not been firmly established. In order to test this hypothesis, 68 three‐dimensional biometric landmarks of the nasal cavity and surrounding cranium were collected from CT scans of mixed‐sexed samples of modern humans from West Africa (n=14) and Papua New Guinea (n=17). Data were also collected on crania from the Arctic Circle (n=17) to serve as a non‐equatorial comparator. A total of 45 linear measurements were subsequently calculated from these landmark data using the Euclidean distance formula. Multivariate Analyses of Variance (MANOVA) was subsequently conducted to test for regional differences while accounting for the potential influence of sexual dimorphism. MANOVA results indicate significant differences in nasal morphology between the three regional samples (Wilk’s λ = 0.008, p = 0.012) with no evidence of significant sexual dimorphism (Wilk’s λ = 0.111, p = 0.104) or region‐sex interactions (Wilk’s λ = 0.023, p = 0.122). Results of subsequent univariate ANOVAs for each linear measurement indicate no significant differences between any of the three samples for dimensions ...