Genomic determinants of speciation

Abstract

Studies of life rely on classifying organisms into species. However, since Darwin, there is no agreement about how to separate species from varieties. Contrary to a frequent belief, quantitative standards for species delineation are lacking and overdue, and debates about species delimitation create obstacles for conservation biology, agriculture, legislation, and education. To tackle this key biological question, we have chosen butterflies as model organisms. We sequenced and analyzed transcriptomes of 188 specimens representing pairs of close but clearly distinct species, populations, and taxa that are debated among experts. We find that species are robustly separated from populations by the combination of two measures computed on Z-linked genes: fixation index that detects hiatus between species, and the extent of gene flow that quantifies reproductive isolation. These criteria suggest that all 9 butterfly pairs that caused experts' disagreement are distinct species, not populations. When applied to Homo, our criteria agree that all modern humans are the same species distinct from Neanderthals, suggesting relevance of this study beyond butterflies. Furthermore, we found that proteins involved in interactions with DNA (including proteins encoded by trans-regulatory elements), circadian clock, pheromone sensing, development, and immune response recurrently correlate with speciation. A significant fraction of these proteins is encoded by the Z-chromosome, which appears to be resistant to introgression. Taken together, we find common speciation mechanisms in butterflies, reveal the central role of Zchromosome in speciation, and suggest quantitative criteria for species delimitation using genomic data, which is vital for the exploration of biodiversity.

Influence of the large–Z effect during contact between butterfly sister species

“…At the same time, the fraction of divergent positions in butterfly sister genomes is 61 roughly the same for autosomes and the Z chromosome (see Fig. 5A of reference [2]). Correlation between fixation indices for the Z chromosome, F z st , and the autosomes, F a st .…”

“…In 34 the present context, stars in a galactic cluster might be likened to pairs of interbreeding 35 species belonging to a particular family of organisms, and the pattern of data exhibited by 36 such pairs could provide similar insight into the conditions that regulate species integrity. 37 Recently, Cong et al have collected exactly this sort of data in their study of but-38 terfly sister species across suture zones in North America [2]. To determine the status of 39 populations, they likewise employed two basic indicators: (i) The index of gene flow, I gf , 40 a measure of their own invention, which describes the fraction of genes in a population 41 imported from its sister population, and (ii) the fixation index, F st , a standard genetic 42 measure of population difference [5].…”

“…A species can be defined as a population of organisms that maintains its genomic 19 integrity despite interbreeding, and exchange of genetic material with other populations 20 [1]. In practice, species are delineated by comparing genomic sequences using simple 21 measures of genetic distance; Genomes sampled from a population describe members 22 of the same species if they are significantly more similar to each other than they are to 23 genomes from closely related populations [2]. A large amount of effort has been devoted to 24 the improvement of methods for comparing and characterizing natural populations [3][4][5][6][7].…”

“…Recent studies comparing divergent populations of butterflies have revealed elevated levels of divergence on the Z chromosome relative to autosomes [1,2]. To explain these observations, it was suggested that the observed patterns of divergence result from the accumulation of postzygotic incompatibilities, obstructing introgression on the Z chromosome in hybrids (see, for example, Fig.…”

“…Because the probability of obtaining a genome comparison with G min ≤ 0.25 will depend 302 on the number genomes sampled, we compute I gf by sampling only 10 genomes per 303 population [3], about twice the number sampled from butterfly populations by Cong et 304 al. [2].…”