The paradox of a wide nasal aperture in cold-adapted Neandertals: a causal assessment
Introduction
The cold, dry conditions of late Pleistocene Europe are thought to have placed strong selective pressures on Neandertals living at more northerly latitudes, resulting in a suite of morphological features that have been explained as the result of thermoregulatory adaptation. The Neandertal thoracic region has been described as barrel-chested (Schaafhausen, 1858, Boule, 1911–1913, Hrdlička, 1930, Howell, 1952, Howell, 1957, Coon, 1962, Smith, 1976, Franciscus and Churchill, 2002) and is associated with a wide bi-iliac breadth (Ruff, 1991), both of which are consistent with the expectations of Bergmann's rule in a cold-adapted endothermic species. Similarly, Neandertal distal limb proportions are predictably short relative to modern humans from more tropical environments following expectations from Allen's rule (Trinkaus, 1981, Holliday, 1997a, Holliday, 1997b).1 The Neandertal cranium has further been characterized as exhibiting cold climate adaptations with regard to neurocranial shape (e.g., Beals et al., 1983), aspects of paranasal size and shape (Sergi, 1960, Badoux, 1965, Coon, 1962), interorbital distance (Sergi, 1958, Sergi, 1962), and internal nasal size and shape (Coon, 1962, Trinkaus, 1987a, Dean, 1988, Franciscus and Trinkaus, 1988, Yokley, 2006). However, other researchers have suggested that many aspects of the Neandertal midface are due to combinations of retained plesiomorphies and uniquely derived traits stemming from random genetic drift rather than to specialized adaptation to cold or other factors per se (e.g., Howell, 1952, Howell, 1957, Antón, 1994, Hublin, 1998, Maureille and Houët, 1998, Franciscus, 1999, Franciscus, 2002, Franciscus, 2003a, Weaver et al., 2007). Nonetheless, at least some aspects of the Neandertal nose, such as narrow superior internal nasal dimensions in specimens deriving from glacial periods, are consistent with the predictions of cold climate adaptation generated from empirical observation and clinical and mathematical modeling of nasal function in recent humans (e.g., Franciscus, 2003b, Roseman and Weaver, 2004).
Surprisingly, one feature that deviates strongly from a well-established pattern of cold climate morphology is the characteristically wide Neandertal nasal aperture (Coon, 1962, Wolpoff, 1968, Wolpoff, 1999, Smith, 1983, Franciscus, 1995), a feature that is usually associated with populations adapted to more tropical climates. Nasal breadth, as one component of the nasal index, exhibits a strong ecogeographic clinal distribution with recent modern humans indigenous to higher latitudes exhibiting, on average, narrower nasal breadths than recent modern humans found in equatorial regions. This adaptive trend is a reflection of respiratory physiology dynamics associated with varying levels of humidity and temperature. A narrower nasal aperture functions better in cold, dry environments by moistening and warming air as it is inhaled as well as recapturing moisture during exhalation. In contrast, a wide nasal breadth functions to dissipate heat in more tropical environments (Thomson and Buxton, 1923, Davies, 1932, Weiner, 1954, Franciscus and Long, 1991, Franciscus, 1995, Yokley, 2006).
Given the relationship between climate and nasal breadth, one would predict that Neandertals, associated with increasingly colder, drier conditions, should exhibit relatively narrower nasal apertures. To explain this paradox, various researchers have argued that nasal breadths in Neandertals, which would have otherwise narrowed through climatic selection, were ontogenetically constrained by the plesiomorphic retention of large anterior teeth and broad palates that characterized their Pleistocene ancestors (e.g., Coon, 1962, Wolpoff, 1968, Wolpoff, 1999, Smith, 1983). That is, the wide Neandertal nose is the result of a large anterior dentition with increased growth of the anterior palate in effect pulling apart the edges of the nasal aperture (e.g., Maureille and Houët, 1993). A significant correlation between nasal breadth and maxillary intercanine breadth in recent modern human adult samples documented by Schwalbe (1887) and Glanville (1969) has often been cited as support for this explanation.
Inherent to this idea is an assumption of causality; namely, that the breadth of the anterior palate and dentition actively constrain nasal breadth dimensions to a significant degree during growth rather than the reciprocal. To a certain extent, this assumption appears logical. The functional significance of mastication and its influence on peripheral facial regions and the entire cranium, both in terms of evolutionary change through natural selection and in terms of idiosyncratic developmental plastic (i.e., nongenetic) response, has been an essential principle in the interpretation of primate skull form (see references in Spencer, 1999). Moreover, the potential role of masticatory behavior combined with paramastication has also heavily influenced the interpretation of facial form in hominid evolution (see references in O'Connor et al., 2005). Nonetheless, the correlation between intercanine breadth and nasal breadth by itself does not preclude the possibility that nasal breadth, responding to some other nondentognathic factor or factors, might actually constrain intercanine breadth instead. Moreover, a simple correlation between nasal breadth and intercanine breadth does not take into account more complex aspects of craniofacial growth dynamics. That is, during growth, increased transverse facial dimensions are associated with an increase in facial dimensions in the sagittal plane (e.g., Edwards et al., 2007) and it is unknown to what degree the breadth of the Neandertal nose may be a function of facial prognathism rather than the breadth of the anterior dental arch.
The only explicit examination of the potential role of the dentognathic region and facial prognathism on nasal morphology was conducted by Glanville (1969) using a sample of recent European and African modern human crania. He found that nasal breadth was significantly correlated with basion-nasion length and basion-prosthion length, although the highest correlation was with intercanine breadth (r = 0.477 for the African sample and r = 0.635 for the European sample). Moreover, when examining the combined effects of facial prognathism and intercanine breadth on the breadth of the nasal aperture, only intercanine breadth was retained as a statistically significant variable within each group. Nevertheless, according to Glanville's results, intercanine breadth only explained 23–40% of the variance in nasal breadth. Moreover, using regression analysis, Glanville was able to explicitly reject the hypothesis that nasal breadth is causally determined by intercanine breadth. However, he was unable to reject the reciprocal hypothesis that nasal breadth causally determines intercanine breadth. This result is interesting given indications that proper eruption and occlusion of the permanent maxillary canines is dependant in part on the growth of the subnasal region (Coulter and Richardson, 1997).
Nevertheless, Glanville's results, including the nonsignificance of facial prognathism in determining nasal breadth, may be, in part, due to his focus on variation within his European and African modern human samples. Arguably, examining variation between European and African modern human samples is more salient regarding the relationships among nasal breadth, intercanine breadth, and facial prognathism given that variation in nasal breadth is arrayed primarily latitudinally across broad ecogeographic regions as a function of major differences in temperature and humidity gradients (Thomson and Buxton, 1923, Davies, 1932, Weiner, 1954, Franciscus and Long, 1991, Franciscus, 1995, Yokley, 2006). In addition, Glanville acknowledged that his results were based solely on adult crania and that demonstration of a developmental causal relationship between nasal breadth and intercanine breadth would require cross-sectional or longitudinal ontogenetic data. Finally, while his analysis was conducted only on recent modern humans, Glanville also recognized the potential importance of his findings with respect to fossil hominids, including Neandertals.
The present analysis is designed to further explore the relationships among nasal breadth, intercanine breadth, and facial prognathism. This elaborates on Glanville's (1969) original study by examining interpopulation variation in European and African modern human samples, including early to late Pleistocene fossil material, and examining the ontogenetic development of nasal breadth and intercanine breadth using longitudinal growth data. We test four hypotheses:
- 1)
We test the hypothesis that intercanine breadth (ICB) explains a significant amount of the variance in nasal breadth (NLB) in recent modern humans, as well as across genus Homo. In addition to ascertaining whether we can replicate Glanville's (1969) results on an independent recent modern human sample, we further test whether Glanville's relatively low coefficients of determination between intercanine breadth and nasal breadth in recent humans may be a function of intrapopulation variation which underestimates a potentially stronger relationship across a larger gradient of facial size and temporal variation in Homo.
- 2)
We use causal modeling (path analysis) to test the main null hypothesis that variation in nasal breadth in our recent modern human and combined genus Homo sample is statistically causally determined by intercanine breadth (e.g., Coon, 1962, Wolpoff, 1968, Wolpoff, 1999, Smith, 1983) after holding the potential causal effects of basion-nasion length (BNL) and basion-prosthion length (BPL) constant. We predict that if ICB is a primary causal determinant of NLB, it should exhibit the greatest path coefficient relative to BPL and BNL.
- 3)
We also test the alternative hypothesis that variation in ICB is statistically causally determined by the breadth of the nose after holding the potential causal effects of BNL and BPL constant. Given that Glanville (1969) was unable to reject this hypothesis within the limits of his study, we revisit this possibility with our wider sampling strategy and a more explicit analytical model. Similarly, we predict that if NLB is a primary causal determinant of ICB, it should exhibit the greatest path coefficient relative to BPL and BNL.
- 4)
Since biological correlation and causality among variables are ultimately tied to growth dynamics, and because statistical causality via path analysis does not necessarily demonstrate developmental causality, we also test the hypothesis that there is an ontogenetic relationship between intercanine breadth and nasal breadth in recent modern humans using longitudinal growth data. Should a direct developmental relationship exist between intercanine breadth and nasal breadth it is predicted that these two variables will exhibit similar growth trajectories. In contrast, dissimilar growth trajectories would suggest that these variables are responding to different growth parameters.
Section snippets
Materials and methods
To test the first three hypotheses we examined both recent modern human and fossil Homo samples. Our recent modern human mixed-sex sample (n = 231; Table 1) consisted of warm-adapted Bantu sub-Saharan African crania (n = 119) and colder-climate Western European crania (n = 112) both with approximately equal numbers of males and females. These two modern human samples were selected because they maximize nasal breadth variation from a wider compilation of Old World recent modern human samples.
Results
The correlation coefficients (r) and coefficients of determination (r2) for NLB, ICB, BPL, and BNL in the aggregate modern human and combined genus Homo samples are listed in Table 2. While all correlations were statistically significant at the p < 0.05 level or greater, they nevertheless all tended to reflect moderate rather than strong associations. Correlations between NLB and ICB were r = 0.504 and r = 0.682 for the modern human and combined genus Homo samples, respectively. As such, 26–47% of
Discussion and conclusions
The hypothesis that intercanine breadth explains a significant amount of the variance in nasal breadth in our recent, fossil, and combined samples is rejected by this analysis. Although we documented a statistically significant correlation between the breadth of the nasal aperture and ICB this does not necessarily indicate a biologically significant correlation as ICB explains less than 50% of the variance in NLB. In the modern human sample, around 75% of the variance in NLB is left unexplained
Acknowledgments
We are indebted to all of the individuals who provided access to the recent human collections and fossil hominids curated in Europe, Israel, Africa, and the United States: B. Arensburg, G. Avery, A. Barzilay, M. Bellatti, M. Beruer, J. Brauer, D. Buisson, J.F. Bussire, N. Cameron, S. Condemi, G. Commerford, H. de Lumley, R. Foley, O. Giuggiolá, D. Grimaud-Herv, M. Guerri, M. Henneberg, I. Hershkovitz, J.J. Hublin, H.E. Joachim, G. Koufos, R. Kruszynski, E. Ladier, A. Langaney, S.
References (78)
Ontogenetic integration of the hominoid face
J. Hum. Evol.
(2005)- et al.
The Sima de los Huesos crania (Sierra de Atapuerca, Spain). A comparative study
J. Hum. Evol.
(1997) Another look at the nose and the functional significance of the face and nasal mucous membrane for cooling the brain in fossil hominids
J. Hum. Evol.
(1988)- et al.
Longitudinal study of facial skeletal growth completion in three dimensions
Am. J. Orthod. Dent. Orthop.
(2007) - et al.
The costal skeleton of Shanidar 3 and a reappraisal of Neandertal thoracic morphology
J. Hum. Evol.
(2002) - et al.
The premaxilla in Neandertal and early modern children: ontogeny and morphology
J. Hum. Evol.
(1999) The Neanderthal: a new look at an old face
J. Hum. Evol.
(1986)Climate, body size and body shape in hominid evolution
J. Hum. Evol.
(1991)- et al.
The plan of the human face: a test of three general concepts
Am. J. Orthod.
(1984) - et al.
The effect of lower limb length on the energetic cost of locomotion: implications for fossil hominins
J. Hum. Evol.
(2004)