RELAXING SELECTIVE PRESSURES ON DEVELOPMENTALLY COMPLEX INTEGUMENTARY STRUCTURES: FEATHER VANE SYMMETRY EVOLVES IN ADDITION TO BODY MASS AND WING LENGTH AFTER FLIGHT LOSS IN RECENT BIRDS

Abstract

Feathers are complex integumentary structures with high diversity across species and within plumage and have varied functions (e.g., thermoregulation, flight). Flight is lost in many crown lineages, and frequently occurs in island founding or semiaquatic context. Different extant lineages lost flight across at least three orders of magnitude of time ([~]79.58 Ma-15 Ka). Flight losss effect on sensory capacity, brain size, and skeletomusculature have been studied, but less work exists on relations between flightlessness and feathers. To understand how flight loss affects feather anatomy, we measured 11 feather metrics (e.g., barb length, barb angle) from primaries, tertials, rectrices, and contour feathers on skins of 30 flightless taxa and their phylogenetically closest volant taxa, supplemented with broader sampling of primaries across all orders of volant crown birds. Our sample includes 27 independent losses of flight; the sample contains nearly half the extant flightless species count and matches its [~]3:2 terrestrial:semiaquatic ratio. Vane symmetry increases in flightless lineages, and these patterns are strongest in flight feathers and weakest in coverts. Greatest changes in feathers are in the oldest flightless lineages like penguins, which show robust filaments (rachis, barbs, and barbules) on small feathers, and ratites, which show high interspecific diversity with plumulaceous filaments and/or filament loss. Phylogenetic comparative methods show that some of these microscopic feather traits, such as barb/barbule length and rachis width, are not as dramatically modified upon flight loss as are body mass increase and relative wing and tail fan reduction, whereas the effect on vane symmetry is more easily detected. Upon relaxing selection for flight, feathers do not soon significantly modify many of their flight adaptations, although increased vane symmetry is likely the most detectable shift. Feathers of recently flightless lineages are in many ways like those of their volant relatives. Feather microstructure evolution is often subtle in flightless taxa, except when flight loss is ancient, perhaps because developmental constraints act upon feathers and/or selection for novel feather morphologies is not strong. Changes in skeletomusculature of the flight apparatus are likely more evident in recently flightless taxa and may be a more reliable way to detect flight loss in fossils, with increased vane symmetry as potentially a microscopic signal. Finally, we see an intriguing, reversed pattern in feather evolution after flight loss from the pattern proposed in popular developmental models of feathers, with the later stages of feather development (asymmetric displacement of barb loci) being lost more readily, while early stages of development (e.g., differentiated barb ridges on follicle collar) are only lost after many millions of years of flightlessness.