Integration and Evolvability in Primate Hands and Feet

Abstract

Morphological integration theory predicts that sets of phenotypic traits that covary strongly due to developmental and/or functional connections between them eventually co-evolve because of a coordinated response to selection, and accordingly become less independently evolvable. This process is not irreversible, however, and phenotypic traits can become less integrated, and hence more independently evolvable, in the context of selection for divergent functions and morphologies. This study examines the reciprocal relationship between shared function, integration and evolvability by comparing integration patterns among serially homologous skeletal elements in the hands and feet of a functionally diverse sample of catarrhine primates. Two hypotheses are tested:(1) species in which the autopods are functionally more similar (e.g. quadrupedal monkeys) have more strongly integrated autopods than species in which the autopods are functionally divergent (e.g. gibbons, humans) and (2) the latter have autopods that are more evolvable, collectively and independently. Morphometric data from selected hand and foot digital rays were used to derive phenotypic variance/covariance matrices. The strength of integration among autopods was quantified using eigenanalysis and a measure of trait variational autonomy. Evolvability was estimated by subjecting phenotypic variance/covariance matrices to simulated random selection gradients, and comparing evolutionary responses among species. Results indicate that integration decreases as hands and feet become functionally divergent, and that the strongly integrated hand and foot skeletons of quadrupedal monkeys respond to selection in a highly collinear manner, even when simulated selective pressures acting on each autopod are in opposite directions in phenotypic space. Results confirm that the evolvability of morphological traits depends largely on how strongly they covary with other traits, but also with body size. The role of pleiotropy as a developmental mechanism underlying integration and evolvability is also discussed.

Selection, Morphological Integration, and Strepsirrhine Locomotor Adaptations

“…Our results indicate that a reduction in limb integration is also found in non-AQ strepsirhines and suggests that, across primates, selection for divergent limb function reduces between-limb integration. This general tendency for a decrease in integration of primate limb elements has also previously been associated with changes in selection pressures related to hand function (Rolian 2009), and suspensory behaviors (Young 2004). In fact, we hypothesize that, across tetrapods, a decrease in integration may be broadly expected in any instance where selection for quadrupedality is relaxed.…”

“…Therefore, one way to measure evolvability is to quantify the relationship between the vectors for selection (b) and response (Dz). Note that we substituted the phenotypic covariances for the genotypic covariances, which are unavailable for this data set; phenotypic covariance matrices have been effectively used as substitutes (see Cheverud 1988;Porto et al 2009;Rolian 2009). We subjected the covariance matrix for each sample to 10,000 randomly generated selection vectors, and measured the angle between the selection vector and the response vector.…”

“…Evolvabilities of the AQ taxa will be compared to those of the non-AQ group, with the expectation that if genetic integration is largely driving the relationship of limb lengths in the species, and selection related to locomotion has little or no effect, there should be no difference in evolvability between the two groups. See Rolian (2009) for a similar exercise on the hands and feet of catarrhines.…”

The relative correspondence of cranial and genetic distances in papionin taxa and the impact of allometric adjustments

“…In contrast, the hominoid postcranial bauplan is highly variable reflecting a range of diverse locomotory repertoires include terrestrial knucklewalking, brachiation, and bipedalism. A similar pattern of strong integration in the cercopithecoid autopod, compared with a less integrated pattern in hominoids has also been shown (Rolian, 2009). Therefore, many of the differences we observe here in terms of patterns of morphological-genetic matrix correlations could be due to overall different levels of phenotypic integration and consequent evolvability in cercopithecoids and hominoids, although this remains to be explicitly tested for the cranium.…”

Multivariate Selection Theory in Primatology: An Introduction to the Concepts and Literature

“…As such, phenotypic integration involves an investigation of the G and P matrices. Cheverud [56], among others, showed that there is a strong correlation between G and P. Due to this, many studies of phenotypic integration only examine patterns of phenotypic covariation (e.g., [71][72][73]). Ackermann and Cheverud [74] provide an excellent review of phenotypic integration studies in primates, and methodological and theoretical approaches are covered in [75][76][77].…”