Morphological integration in the hominid midfoot

Foot morphology is reflective of locomotor behavior and has been the focus of extensive research in evolutionary anthropology as it pertains to the evolution of bipedalism (Morton, 1935; Susman, 1983; Ward et al., 2012; DeSilva et al., 2019). Previous analyses have emphasized that there is considerable morphological variation in the foot bones of extinct hominins, and that there were multiple forms of bipedal adaptations throughout the Plio-Pleistocene (Zipfel et al., 2011; Haile-Selassie et al., 2012; Clarke, 2013; McNutt et al., 2021). Some authors have suggested that this morphological diversity is more pronounced in the medial elements of hominin feet compared to the lateral elements (e.g., Haile-Selassie et al., 2012; Clarke, 2013; DeSilva et al., 2019). The lateral elements (i.e., the cuboid, and fourth and fifth metatarsals [Mts4–5]) are further hypothesized to have evolved adaptations for midfoot stiffness earlier than the medial elements (i.e., the cuneiforms and Mts1–3), and to have played a significant role in early modes of bipedal locomotion (Kidd, 1999; Lovejoy et al., 2009; Fernández et al., 2018; McNutt et al., 2018; DeSilva et al., 2019). However, given the patchy nature of the fossil record, this remains difficult to prove. In addition, there is some evidence of lateral foot morphological diversity based on the dorsoplantarly convex Mt4 proximal facet of Australopithecus sediba (Zipfel et al., 2011), and some authors have argued that the Burtele Mt1 is not as different from the Australopithecus afarensis Mt1 as initially proposed (Prang et al., 2021).

Nonetheless, there is considerable evidence that not all elements of the midfoot evolved at a similar rate, or in the same morphological direction among contemporaneous taxa (Haile-Selassie et al., 2012; Clarke, 2013; DeSilva et al., 2019). It is also unclear which evolutionary processes contributed to this morphological diversity. Changes to hallucal morphology have been attributed to directional selection for increasingly bipedal and terrestrial locomotion (Rolian et al., 2010), and aspects of Mts4–5 proximal epiphyseal shape have been clearly linked to midfoot mobility and stiffness (DeSilva, 2010). However, the midfoot region is composed of 10 closely packed elements (i.e., the cuboid, navicular, cuneiforms, and proximal metatarsals), which largely act as a functional unit. For this reason, we expect that the morphological variation across these elements is correlated, and that elements will likely vary in their ability to respond to selection independently. Furthermore, it is also possible that morphological diversification of certain elements was not directly caused by selection, but rather by selection on other correlated traits or elements. Previous studies have suggested that many morphological traits did not evolve independently (Lande, 1979; Cheverud, 1982; Lovejoy et al., 1999; Marroig et al., 2009; Young et al., 2010), but to date, this has not been investigated in the midfoot.

Morphological integration is a central concept in evolutionary biology that refers to the coordinated variance of phenotypic traits due to common functional, developmental, or genetic factors (Olson and Miller, 1958; Hallgrímsson et al., 2009). Traits that covary because of these common factors may become subjected to long-term coevolution (Cheverud, 1996). Within an organism, traits often do not show uniform magnitudes of integration; when a subset of traits are highly integrated with one another to the exclusion of other traits, the organism is said to have a modular structure (Klingenberg, 2008, 2013). Both morphological integration and modularity may influence a population's evolvability, which is defined as its ability to respond to selection (Hansen and Houle, 2008). Specifically, morphological integration can either constrain or facilitate a population's ability to evolve in the direction of selection (Hansen, 2003; Hansen and Houle, 2008). Highly integrated traits can influence a species' evolutionary trajectory when selection on one trait leads to a correlated response in other associated traits (Grabowski et al., 2011; Rolian, 2014; von Cramon-Taubadel, 2019). In particular, high levels of integration among traits may limit their ability to respond in the direction of selection and facilitate a response along the direction of size-related variation (Marroig, 2009; Porto et al., 2009; Grabowski et al., 2011). A highly modular structure may lead to different evolutionary trajectories between parts of an organism, particularly when traits experience differing or conflicting functional demands (Raff, 1996; Wagner and Altenberg, 1996; Gerhart and Kirschner, 1997; Kirschner and Gerhart, 1998; Klingenberg, 2013). Determining the genetic variance/covariance matrix (G-matrix) for morphological characters is not straightforward, therefore it has been commonplace in the literature to substitute this matrix with the phenotypic variance/covariance matrix (P-matrix) following Cheverud's conjecture (Cheverud, 1988). This conjecture is based on studies showing that these two matrices are nearly proportional across several species, allowing us to use phenotypic traits to make inferences about evolutionary processes, as well as patterns and magnitudes of morphological integration in biological structures (Cheverud, 1988; Roff, 1995, 1996; Marroig et al., 2009; Porto et al., 2009). Although this substitution has been criticized by some (e.g., Kruuk et al., 2008), recent studies have provided further support for the conjecture (e.g., Sodini et al., 2018).

Previous research on postcranial morphological integration across anthropoids has focused on the long bones, the spine, and the pelvis (Rolian, 2009; Young et al., 2010; Grabowski et al., 2011; Arlegi et al., 2018; Conaway et al., 2018), and has emphasized the importance of functional differences in altering covariational structure. For example, humans have lower levels of morphological integration in the pelvis than other hominids, particularly among traits that are functionally important for bipedalism (Grabowski et al., 2011). Furthermore, all primate taxa with divergent forelimb and hindlimb function show reduced integration between the limbs than primates with similar forelimb and hindlimb function (Young et al., 2010; but see Diogo et al., 2018).

It is less clear how functional differences might influence patterns and magnitudes of integration within the hands and feet. Williams (2010) found that elements associated with the ‘knuckle-walking complex’ (i.e., the capitate and third ray) do not show different patterns of integration in knuckle-walking vs. non-knuckle-walking species, suggesting that differences in hand function may not result in considerable differences in patterns of integration in these elements. Within the rearfoot, Robinson (2018) demonstrated that humans and African apes share similar patterns and magnitudes of integration among the navicular, talus, and calcaneus despite their locomotor differences, cautioning against the assumption of an influence of function on morphological integration. However, gorillas, chimpanzees, and humans all broadly possess adaptations for terrestrial heel strike plantigrady in the rearfoot (Gebo, 1992), which could result in similar patterns and magnitudes of integration across the analyzed elements. Furthermore, Robinson (2018) uses a sample size of 20 per species, which is low when considering results from Cheverud (1988) and Jung and colleagues (2020), who found that a sample size of at least 40 [individuals] is needed to accurately estimate the structure of P-matrices used in certain measures of integration. Others have also shown that for some measures of integration and evolvability, even 40 may be too low (Grabowski and Porto, 2017).

In contrast, elements of the midfoot show highly pronounced morphological and functional differences (e.g., hallucal abduction), and as a result, may display different patterns and magnitudes of morphological integration. This suggestion is supported by Diogo et al. (2018), who use anatomical network analyses to reveal that the human hallux represents an independent musculoskeletal module from all other digits, whereas the chimpanzee hallux does not. Although anatomical network analyses do not make use of variance-covariance matrices, making these results not directly comparable to most studies of morphological integration across skeletal elements, they do emphasize that hallucal abduction may influence integration and modularity in the midfoot.

In addition, integration between traits may occur as a result of pleiotropic genetic effects or shared developmental pathways (Cheverud, 1996; Grabowski et al., 2011), but relatively little is known about the genetic underpinnings of foot evolution. However, by comparing humans with nonhuman apes, we may be able to provide insights into how magnitudes of morphological integration are influenced by differences in biomechanical function. We can also determine if there are any elements that are consistently more or less integrated with one another across the hominid lineage, which may have implications for midfoot evolution across a macroevolutionary scale.

The morphology of the midfoot reflects the foot's overall range of motion, as well as its biomechanical effectiveness in propulsion and/or grasping (Elftman and Manter, 1935; Schultz, 1963; Susman, 1983; Aiello and Dean, 1996). To reconstruct the evolution of bipedalism in hominins, we must understand the midfoot's structure and how it varies throughout the hominid lineage. Here, we quantify magnitudes of integration in the midfoot of humans, western lowland gorillas, chimpanzees, and orangutans, and aim to determine how functional differences might alter their covariational structure. We also aim to determine whether magnitudes of integration between elements are driven by joint congruence, as inferred by magnitudes of articular surface trait covariance. We quantify magnitudes of integration within individual elements of the midfoot, between adjacent elements, and between a selection of articulating facets in the midfoot. We make three major predictions:

1.

Chimpanzees will show similar magnitudes of morphological integration between elements of the midfoot because of their wide locomotor repertoire.

Both gorillas and chimpanzees are considered semiterrestrial quadrupeds, using terrestrial quadrupedalism, vertical and horizontal climbing, and occasional bouts of bipedalism (Tuttle and Watts, 1985; Remis, 1995; Doran, 1996; Thorpe and Crompton, 2006). Although African apes are more terrestrial than arboreal, they have a wider locomotor range than humans and orangutans. Chimpanzees generally practice more arboreal locomotion than gorillas and utilize a wider range of branch diameters (Hunt, 1992; Kano, 1992; Jabbour, 2008). As a result, they likely experience a wider range of foot postures. Given the shared function of all elements of the midfoot depending on the locomotor mode, we predict that all tarsals and metatarsals will show similar magnitudes of integration with one another.

2.

Species with an abducted hallux will show a lower magnitude of morphological integration between the medial cuneiform and the Mt1, and between the Mt1 and Mt2 compared to humans.

Elements with divergent functions tend to be less integrated with one another (Rolian, 2009; Young et al., 2010). Although all taxa with an abducted hallux use the Mt1 in combination with Mts2–5 for grasping, the Mt1 is more independently moveable, and has a wider range of mobility than in humans (Morton, 1935; Latimer and Lovejoy, 1990). In humans, the Mt1 is less mobile, and along with the other metatarsals, functions as part of the lever arm of the foot during bipedal locomotion (Schultz, 1963). We predict that all species with an abducted hallux will show lower levels of integration between the Mt1 and the medial cuneiform, and between the Mt1 and the Mt2 than in humans. Conversely, we predict that humans will show higher levels of morphological integration between these two pairs of elements.

3.

Humans will show relatively high magnitudes of integration between the articulating facets of the medial elements, whereas apes will show relatively similar or lower magnitudes of integration across all pairs of articular facets of the midfoot.

The human midfoot is distinct from that of other apes in that the medial elements (particularly Mt1 and Mt2) experience relatively high levels of compressive loading compared to the lateral elements (Mts4–5 and the cuboid), because of their role in propulsion during bipedal locomotion (Hutton and Dhanendran, 1981; Katoh et al., 1983; Hills et al., 2001; Pataky et al., 2008). Because of the high loading the medial elements incur, we predict that the tarsometatarsal joints of the first three rays and the intertarsal joints of the navicular and cuneiforms will be more congruent than in the lateral midfoot, as indicated by higher magnitudes of integration between articulating facets. In contrast, apes incur more variable foot loading patterns (Wunderlich, 1999; Vereecke et al., 2003; Wunderlich and Ischinger, 2017), and we predict that all articular surfaces with the exception of the first tarsometatarsal joint will show similar levels of morphological integration.

留言 (0)

沒有登入
gif