Sunday, May 14, 2017

Evolutionary Processes Shaping Diversity Across The Homo Lineage

Fig. 3. Principal component plots of PC1 and PC2 for a subset of Generalized Procrustes analyses (GPA). The remaining principal components plots are illustrated in SOM Fig. S2. A summary of all GPA results is given in Table 3. The percentage of variance explained by each principal component is displayed on each plot. (a) GPA 1 – mandible. Species convex hulls are separated along PC1. Most Pleistocene Homo specimens fall within the H. erectus convex hull, with the exception of H. rudolfensis specimens and KNM-ER 1802. KNM-ER 60000 is an outlier along PC2. (b) GPA 2 – mandible. There is a fair amount of overlap between species convex hulls. All Pleistocene Homo specimens are contained within the convex hull of H. erectus, with the exception of LD 350-1 which falls just outside of the range. D2600 is an outlier along PC2. (c) GPA 5 – upper face. Most specimens fall within the H. sapiens range, except for Ndutu, SK 847, D2700 and KNM-ER 3732. (d) GPA 6 – maxilla. Dmanisi H. erectus shows the most variability along PC1. A.L.666-1 is closely associated with D2282 and Stw 53 in shape space, and OH 65 along PC1. The H. habilis convex hull is enclosed within the H. sapiens range. (e) GPA 8 – temporal. Most specimens are contained within the H. sapiens convex hull, with the exception of OH 24, DH3, OH 9, KNM-BC 1, Tuinplaas 1 and KNMES 11693. (f) GPA 11 – neurocranium. H. erectus is most variable along PC2. DH2 falls just outside the convex hull of H. erectus along PC 1.


Evolutionary Processes Shaping Diversity Across The Homo Lineage
Lauren Schroeder, Rebecca Rogers Ackermann
This manuscript is currently under review in the Journal of Human Evolution
bioRxiv preprint first posted online May. 10, 2017

From the paper:


The results of our analyses indicate that morphological relationships among Homo taxa are complex, and suggest that diversification may be driven primarily (though not exclusively) by neutral evolution. Multivariate and geometric morphometric results were generally consistent and highlighted the large amount of morphological diversity within Homo, especially within H. erectus, a geographically and temporally widespread species. Other interesting patterns also emerged. First, the spatial relationships among specimens differed depending on the morphological region analyzed. For example, Mahalanobis’ distances between H. erectus specimen KNM-ER 3883 and other Pleistocene Homo are significantly different for the temporal region (Fig. 2c), but not for the face (Fig. 2a) and neurocranium (Fig. 2d). Second, the Dmanisi hominins and specimens of H. rudolfensis are consistently different from each other and from other taxa. Third, the oldest Homo specimen, LD 350-1, is significantly different from all other specimens for calculations of Mahalanobis’ distances, except for H. erectus specimen KNM-BK 8518 and H. sapiens specimen Tuinplaas 1. This specimen also falls within, or on the boundary of, the H. erectus convex hulls in principal component plots of Procrustes shape coordinates (Figs 3a-b), lending support to the initial diagnosis of this specimen as Homo (Villmoare et al., 2015). Finally, it is worth noting that there is a close association between H. naledi and H. erectus in both cranial and mandibular analyses (e.g. similar to what has been shown in Dembo et al., 2016; Laird et al., 2016; Schroeder et al., 2016), as well as between ~2.4 Ma early Homo specimen A.L.666-1, South African specimen Stw 53, and H. habilis specimen KNM-ER 1813. The results of these metric analyses confirm the complexity of the phenotypic variation within Homo and the difficulty faced when trying to identify potential evolutionary relationships, especially given the possibility multiple lineages within our genus.

What has produced this diversity? Our results indicate that for 95% of taxon comparisons (51% when a conservative estimate of statistical power is used), across the entire skull (face, maxilla, neurocranium, temporal, mandible), the null hypothesis of genetic drift cannot be rejected. This indicates that of the majority of the cranial and mandibular phenotypic diversity within Homo, from ~2.8 Ma-0.0117 Ma, is consistent with random genetic drift. This is particularly striking for the neurocranium where all three analyses comprising 39 different comparisons are shown to be consistent with drift, even when including very small-brained H. erectus (Dmanisi) and H. naledi(South Africa). What this indicates is that the relative size and shape variation that exists between taxa is proportional to that seen within taxa (here based on the Homo sapiens model). In other words, although morphological divergence is occurring among species, it happens consistently across the phenotype in a manner that does not change the relative relationships among parts. For the neurocranium, this is true despite considerable brain size differences between Homo taxa. In this light, recent suggestions that brain size and shape differences may poorly define Homo (Spoor et al., 2015) are intriguing, because they have arisen in the context of an increased understanding of comparable magnitudes and patterns of variation within taxa. It may be more difficult to delineate taxa under a model of drift, as opposed to a model of selection, which drives changes in the relative relationships among traits. However, it is important to remember that the neurocranial analyses in particular, due to a dearth of available homologous landmarks, did not capture all aspects of brain shape but rather gross shape/size. Nonetheless, based on these results it is necessary to re-consider the traditional view that selection was the main evolutionary process driving changes in the neurocranium, and most other cranial regions, within Homo, and consider the implications of that for our understanding of how and why our lineage evolved.

For the remaining cases, where drift was rejected, three primary patterns can be observed. First, adaptation played a role in driving the evolution of differences between the Dmanisi hominins and other early Homo specimens across both the face and mandible. Interestingly, even though the Dmanisi group itself is hugely diverse, we found that this rejection of drift is consistent across all of the Dmanisi specimens, regardless of the specimen or combination of specimens included in each analysis, confirming that this result was not just a product of intra-group variability. The Dmanisi hominins were the first of our lineage to leave Africa, and our results indicate that selection played an important role in that dispersal, resulting in significant morphological changes (and a different covariance structure) as these hominins adapted to new environmental contexts. Second, although drift was the primary force implicated in neurocranial change, selection repeatedly acted to shape maxillary and mandibular diversity among Homo groups. This result suggests that the evolution of Homo is characterized by adaptive diversification in masticatory systems among taxa, which may be related to dietary change, possibly as a result of environmental change (Vrba, 1985, 1995, 1996, 2007; Cerling, 1992; Stanley, 1992; deMenocal, 1995; Reed, 1997; Bobe and Behrensmeyer, 2004; Wynn, 2004), environment variability (Potts, 1998), and/or shifts to new foraging strategies (Stanley, 1992; Braun et al., 2010; Lepre et al., 2011; Potts, 2012; Ferraro et al., 2013). Third, the mandibular morphology of H. rudolfensis consistently emerges as being adaptively different from other Homo taxa, including the earliest Homo specimen, LD 350-1. This result implies a potentially divergent and distinct evolutionary trajectory for this taxon, possibly signifying a branching event, supporting the distinctiveness of this taxon, and providing an adaptive explanation for divergence in sympatry with other Homo taxa (i.e. H. habilis). However, despite these instances where drift was rejected, we reiterate that, for the majority, selection was not detected. For some cases, this lack of selection is surprising. For example, we do not see a massive adaptive change occurring between 2.7 and 2.5 Ma as per Vrba’s 1985 turnover-pulse hypothesis (Vrba, 1985), nor do we see the expected correspondence between most major cultural transitions and changes in skull morphology.

Interestingly, we also do not detect major selective pressure acting to differentiate Homo sapiens from Middle Pleistocene Homo. This result parallels the findings of Weaver et al. 2007 who show that genetic drift can account for the cranial differences between Neanderthals and modern humans. It also provides further evidence for a “lengthy process model” of modern human origins (Weaver, 2012), supporting the theory of morphological continuity from the later Middle Pleistocene, ~400 000 years ago, to the appearance of anatomically modern humans. While it is important to note that these analyses were only performed on crania and mandibles, these results are nonetheless significant given the emphasis placed on cranial and mandibular material for alpha taxonomy.

There is a fundamental disconnection between the realization that molecular change over evolutionary timeframes occurs predominantly through neutral processes (Kimura, 1968, 1991), and the dominant interpretation (explicitly or implicitly) that morphological change in human evolution is primarily adaptive and directional. The results of this study lend further support to the notion that random change has played a major role in human evolution (see also Ackermann and Cheverud, 2004; Weaver et al., 2007; Schroeder et al., 2014). The detection of widespread genetic drift acting on all aspects of skull morphology during the evolution of our genus is likely to be due, in part, to small population sizes of groups in isolation. This could also be correlated with a purported population bottleneck at ~2.0 Ma (Hawks et al., 2000). Because the emergence and evolution of Homo and the appearance and proliferation of stone tools roughly correspond, and continue to co-evolve, it is also possible that hominins were increasingly reliant on cultural adaptations – as opposed to biological adaptations – to manage environmental changes (Schroeder et al., 2014; Ackermann and Cheverud, 2004; Lynch, 1990). Continued investigation into evolutionary process is necessary – especially for anatomical regions such as the postcranium which remain largely unexplored (but see Grabowski and Roseman, 2015) – in order to provide further insight into how and why the human lineage evolved.

No comments: