Scientists have mapped the evolutionary relationships among all 9,993 of the world's known living bird species. The study is an ambitious project that uses DNA-sequence data to create a phylogenetic tree — a branching map of evolutionary relationships among species — that also links global bird speciation rates across space and time.
“This is the first dated tree of life for a class of species this size to be put on a global map,” says study co-author Walter Jetz, an evolutionary biologist at Yale University in New Haven, Connecticut.
But the endeavour is also controversial, owing to the large number of species for which no sequence data are available. “This is a conceptually brilliant attempt to link space with time while crafting a complete phylogeny,” says Trevor Price, an evolutionary biologist at the University of Chicago in Illinois. “But there are almost certainly introduced artefacts by lacking one-third of the sequences used to create it.”
Jetz and his colleagues built on an extensive phylogenomic study, published in 2008, to divide bird species into 158 clades, well-established groups believed to have evolved from a common ancestor2. Using ten fossils, the researchers dated and anchored that backbone, and placed all the living species on the tree, starting with the roughly 6,600 for which genetic information was available. For the remaining 3,330 species for which no genetic data were available, the researchers used specific constraints — such as membership in the same genus — to identify where species would most likely be placed in the tree. They then created thousands of possible tree configurations and modeled estimates of speciation and extinction rates for each one to account for the uncertainty. The researchers found that although rapid radiations have occurred throughout time and space, the rate of speciation has sharply increased over the past 40 million years.
Some scientists question the finding. “For a tree this size, any small systematic biases in assumptions, integrated over 10,000 species, may result in the detection of trends that simply didn’t exist,” says Mark Pagel, an evolutionary biologist at the University of Reading, UK. But when the researchers repeated the analysis using only species for which genetic data exists, they saw roughly the same pattern.