Many understory birds and other groups form genetically differentiated subspecies or closely related species on opposite sides of major rivers of Amazonia, but are proposed to come into geographic contact in headwater regions where narrower river widths may present less of a dispersal barrier. Whether such forms hybridize in headwater regions is generally unknown, but has important implications to our understanding of the role of rivers as drivers of speciation. We used a dataset of several thousand single nucleotide polymorphisms to show that seven taxon pairs that differentiate across a major Amazonian river come into geographic contact and hybridize in headwater regions. All taxon pairs possessed hybrids with low numbers of loci in which alleles were inherited from both parental species, suggesting they are backcrossed with parentals, and indicating gene flow between parental populations. Ongoing gene flow challenges rivers as the sole cause of in situ speciation, but is compatible with the view that the wide river courses in the heart of Amazonia may have driven interfluvial divergence during episodes of wet forest retraction away from headwater regions. Taxa as old as 4 Ma in our Amazonian dataset continue to hybridize at contact zones, suggesting reproductive isolation evolves at a slow pace.

We possess limited understanding of how speciation unfolds in the most species-rich region of the planet-the Amazon basin. Hybrid zones provide valuable information on the evolution of reproductive isolation, but few studies of Amazonian vertebrate hybrid zones have rigorously examined the genome-wide underpinnings of reproductive isolation. We used genome-wide genetic datasets to show that two deeply diverged, but morphologically cryptic sister species of forest understorey birds show little evidence for prezygotic reproductive isolation, but substantial postzygotic isolation. Patterns of heterozygosity and hybrid index revealed that hybrid classes with heavily recombined genomes are rare and closely match simulations with high levels of selection against hybrids. Genomic and geographical clines exhibit a remarkable similarity across loci in cline centres, and have exceptionally narrow cline widths, suggesting that postzygotic isolation is driven by genetic incompatibilities at many loci, rather than a few loci of strong effect. We propose Amazonian understorey forest birds speciate slowly via gradual accumulation of postzygotic genetic incompatibilities, with prezygotic barriers playing a less important role. Our results suggest old, cryptic Amazonian taxa classified as subspecies could have substantial postzygotic isolation deserving species recognition and that species richness is likely to be substantially underestimated in Amazonia.

No abstract

The role of historical factors in driving latitudinal diversity gradients is poorly understood. Here, we used an updated global phylogeny of terrestrial birds to test the role of three key historical factors-speciation, extinction, and dispersal rates-in generating latitudinal diversity gradients for eight major clades. We fit a model that allows speciation, extinction, and dispersal rates to differ, both with latitude and between the New and Old World. Our results consistently support extinction (all clades had lowest extinction where species richness was highest) as a key driver of species richness gradients across each of eight major clades. In contrast, speciation and dispersal rates showed no consistent latitudinal patterns across replicate bird clades, and thus are unlikely to represent general underlying drivers of latitudinal diversity gradients.

Live fences may act as tools for biodiversity conservation by providing habitat for native species and increasing connectivity in the landscape. We studied the influence of live fence characteristics on species richness and fence use by birds by examining both local and landscape factors. We studied three types of live fences: planted fences of a native tree, planted fences of an exotic, and spontaneous. They were either connected to forest fragments or isolated, and were all within a pasture matrix. Spontaneous and planted live fences maintain a diverse plant (77 shrub and tree species) and bird communities (98 species). Fence types strongly differed in vegetation composition and structure. We found that by analyzing each fence characteristic independently, there was no difference in bird richness or abundance. However, there was a significant correlation when plant richness, structure, and connectivity were analyzed together. This could be the result of some variables counterbalancing each other. Birds used fences for a variety of purposes including foraging, breeding, and moving across the landscape. Native birds and plants used live fences as habitat and refuge in a landscape where large forest tracts have been lost for decades. Live fences in conjunction with small forest fragments maintain a diverse array of plant and birds species that are a subsample of the species originally found in the landscape before extensive deforestation. We recommend the establishment of live fences, allowing growth of spontaneous understory.