Allometry, sexual selection and evolutionary lines of least resistance shaped the evolution of exaggerated sexual traits within the genusTyrannus

Abstract

Variational properties hold a fundamental role in shaping biological evolution, exerting control over the magnitude and direction of evolutionary change elicited by microevolutionary processes that sort variation, such as selection or drift. We studied the genus Tyrannus as a model for examining the conditions and drivers that facilitate the repeated evolution of exaggerated, secondary sexual traits in the face of significant functional limitations. In particular, we explore the role of allometry, sexual selection and their interaction, on the diversification of tail morphology in the genus, assessing whether and how they promoted or constrained phenotypic evolution. Non‐deep‐forked species tend to show reduced sexual dimorphism and moderate allometric variation in tail shape. The exaggerated and functionally constrained long feathers of deep‐forked species, T. savana and T. forficatus, which show both marked sexual dimorphism and allometric tail shape variation, independently diverged from the rest of the genus following the same direction of main interspecific variation accrued during the evolution of non‐deep‐forked species. Moreover, the latter direction is also aligned with axes summarising sexual dimorphism and allometric variation on deep‐forked species, a feature lacking in the rest of the species. Thus, exaggerated tail morphologies are interpreted as the result of amplified divergence through reorientation and co‐option of allometric variation by sexual selection, repeatedly driving morphology along a historically favoured direction of cladogenetic evolution.

Diversification of the ruminant skull along an evolutionary line of least resistance

“…Schluter ( 7 ) tested for the presence of an LLR by successfully predicting that the direction in which species diverge from their ancestors should be biased toward the principal direction of intraspecific variation [PC1 of an additive genetic covariance matrix, G , or a phenotypic covariance matrix, P , used as substitute for G ; ( 76 )] and that species with greater alignment between the direction of their variation and divergence should diverge farther. Statistically testing the congruence between evolutionary divergence and PC1 of G or P matrices has remained the standard means for documenting LLRs since Schluter’s seminal work [e.g., ( 77 – 80 ), although see ( 81 )]. Considering PC1 as the de facto LLR is intuitive because, in the presence of a strong LLR, responses to selection should be channeled into the direction of the LLR, consequently forming the major axis of interspecific phenotypic variation.…”

Functional, genetic, and structural constraints on the exaggeration and diversification of male genital morphology inOhomopterusground beetles

“…Comparative analyses of related species with and without exaggerated sexual traits are expected to be useful to evaluate the tradeoff between constraints and exaggeration (e.g. Satomi et al 2021;Fasanelli et al 2022).…”

Innovation and elaboration on the avian tree of life

“…However, at the macroevolutionary scale, the major axis of phenotypic variation is an emergent property of broader set of genetic and developmental constraints interacting with selection on a moving adaptive landscape ( 24 , 25 ). At this scale, the major axis of phenotypic variation among species ( 15 , 26 ) can be estimated from a matrix of divergences in species mean phenotypic traits [e.g. the D matrix of ( 21 , 27 )].…”