Hybrid zones, where distinct populations meet and interbreed, give insight into how differences between populations are maintained despite gene flow. Studying clines in genetic loci and adaptive traits across hybrid zones is a powerful method for understanding how selection drives differentiation within a single species, but can also be used to compare parallel divergence in different species responding to a common selective pressure. Here, we study parallel divergence of wing colouration in the butterflies Heliconius erato and H. melpomene, which are distantly related Müllerian mimics that show parallel geographic variation in both discrete variation in pigmentation, and quantitative variation in structural colour. Using geographic cline analysis, we show that clines in these traits are positioned in the roughly the same geographic region for both species, which is consistent with direct selection for mimicry. However, the width of the clines varies markedly between species. This difference is explained in part by variation in the strength of selection acting on colour traits within each species, but may also be influenced by differences in the dispersal rate and total strength of selection against hybrids between the species. Genotyping-by-sequencing also revealed weaker population structure in H. melpomene, suggesting the hybrid zones may have evolved differently in each species; which may also contribute to the patterns of phenotypic divergence in this system Overall, we conclude that multiple factors are needed to explain patterns of clinal variation within and between these species, although mimicry has probably played a central role.
Digital images of butterfly wings were taken with a Nikon D7000 DSLR camera fitted with an AF-S DX Micro NIKKOR 40 mm f/2.8G lens (Nikon UK Ltd., Surrey, UK), mounted on an adjustable platform. Standardised lighting conditions were achieved using two external natural daylight fluorescent lights, mounted to illuminate at 45 degrees from incident, to maximise brightness of observed iridescent colour. Photographs were taken with a shutter speed of 1/60 sec and an aperture of f/10. Each sample was photographed with an X-Rite colorchecker passport (X-Rite, Inc., MI, USA) in shot. The Nikon raw (.NEF) image files were converted to standard raw files (.DNG) using Adobe DNG converter (Adobe Systems Inc., USA). The RGB channels in the images were then linearized using the neutral grey scale on the colorchecker using GNU Image Manipulation Program, v2.8.
The mean RGB (red, green, blue) values from regions in the discal cell on the right forewing and the Cu2 cell on the right hindwing were measured (SI Figure S1A) from photographs. If the wings on the right-hand side showed damage, wings on the left-hand side were used. Wing regions were selected using the polygon selection tool in ImageJ, version 1.50b (Abràmoff, Magalhães, & Ram, 2004), and mean RGB scores were measured using the Color Histogram plugin for ImageJ. This plugin outputs the mean, mode and standard deviation of each colour channel (red, green and blue) in the selected area.
Each row corresponds to an individual. Some individuals have two entries, as they were photographed and measured twice to test for repeatability of the measurements. WR1 describes the forewing wing region which was measured, WR2 described the hindwing region which was measured. The mean, mode, and standard deviation of each colour channel (R, G & B), in each wing region are reported in separate columns.
Natural Environment Research Council, Award: NE/K008498/1
Departamento Administrativo de Ciencia, Tecnología e Innovación (COLCIENCIAS), Award: FP44842-5-2017