species diversity

The role of dispersal in Neotropical avian diversity

In a paper published in Nature last month by Brian T. Smith (American Museum of Natural History) and collaborators argue that the strongest predictors of avian speciation in the Amazon are the amount of time a species lineage has endured in the landscape, and how well a bird can move through that landscape. Their results suggest that the dispersal abilities of the birds and how long their lineage has persisted are important drivers of the high biodiversity in the Amazon.

The authors start the introduction by reminding us that we, scientists, usually link the biodiversity of the Neotropics to two major hypotheses:

1) large-scale landscape changes that generate bio-diversification by population fragmentation followed by isolation, and

2) the formation of a geographically structured landscape matrix on which diversification occurred.

The first, commonly known as vicariance, involves reconfigurations of the landscape, such as the separation of continents by plate tectonics, the uplift of mountains or the formation of large rivers. Since the study involves the avian fauna of the Neotropic region, the large-scale events considered by the authors that could drive biodiversity patterns are the Andean mountain uplift, and the formation of the (very large) Amazonian rivers. This first hypothesis is easier to understand: big mountain or rivers separate populations, which can no longer exchange genes and start differentiating from one another to the point where the different sides will have completely separate evolutionary futures.

The second hypothesis involves organisms’ ability to persist in a structured landscape, which does not necessarily need to change. In this case, allopatric speciation would follow dispersal events, and thus, organism-specific abilities to persist and disperse in the landscape are the principal drivers of speciation. Species with lower dispersal abilities have a lower chance of navigating the landscape and, therefore, tend to accumulate higher genetic differentiation between populations. Higher differentiation, in turn, leads to higher speciation rates.

Figure 1 from Smith et al. 2014. Main landscape barriers and data points in the Neotropics.

Figure 1 from Smith et al. 2014. Main landscape barriers and data points in the Neotropics.

To test these two hypotheses, the authors used 2,500 individuals from 27 widespread bird lineages in the Neotropics. To prevent biases of current taxonomic limitations, authors considered lineages instead of species, i. e., they used monophyletic groups as their definition of a lineage instead of going by current taxonomic nomenclature.

They looked at relatively recently diversified lineages that have their distribution interrupted by the Andes, the Isthmus of Panama and large rivers of the Amazon Basin (the Amazon, Madeira and Negro rivers).

To get around hypothesis 1, the authors tested whether the timing of divergence events were congruent with a single episode of vicariance associated with barrier formation, the Andean uplift. To test hypothesis 2, they compared the different dispersal abilities of lineages to their diversification rate. The idea being that species with lower dispersal abilities accumulate higher genetic differentiation between populations, which, in turn, leads to higher speciation rates. The measures of dispersal are based on “foraging stratum (a measure of dispersal ability linked to the behavior of birds: canopy, high dispersal ability or understory, low dispersal ability) and niche breadth (an indirect measure of dispersal ability based on habitat preference)”.

Birds included in the study. Bird drawings from Smith et al. (2014), originally from del Hoyo et al. (2013) Handbook of the Birds of the World.

Birds included in the study.
Bird drawings from Smith et al. (2014), originally from del Hoyo et al. (2013) Handbook of the Birds of the World.

What their genetic data indicate is that there was not a single divergence event, but rather between 9 and 29, and the timing of these events were not synchronous. Most of the species diversity originated during the Pleistocene, i.e. after the Neogene formation of the landscape matrix. If any of the vicariance events predicted to affect speciation (Andean uplift, Isthmus of Panama, Amazonian rivers formation) had been the source of the diversification, the lineage divergence time would be synchronous, since they were being affected by the same event, considering these are relatively recently divergent species. However, wouldn’t only older divergence events be affected by old vicariance events? How well we can test this is entirely dependent on how well the old phylogenetic node divergences can be estimated. In the paper, the authors acknowledge that they “… do not reject the possibility that the initial geographical isolation of populations at deeper phylogenetic scales was due to vicariance associated with the Andean orogeny or with the emergence of other landscape features”.

“Although highly suggestive of multiple dispersal events, this variation could be explained by a single vicariant event associated with the Andean uplift if the dispersal restrictions imposed by the barrier were heavily dependent on dispersal ability, such as was reported for a taxonomically diverse group of marine organisms isolated by the formation of the Isthmus of Panama. In a similar fashion, the emerging Andes could have first become a barrier for bird lineages with low dispersal abilities, with fragmentation of the distributions of more dispersive lineages occurring later. However, we detected no significant associations between dispersal abilities and divergence times across the Andes and the Isthmus of Panama that would support a model of ecologically mediated vicariance for these barriers.”

What about hypothesis 2? They found that whether a bird lineage inhabits canopy or understory affected the species diversity of that lineage. Since they used foraging strata as a proxy for dispersal ability, this result corroborates with the idea that dispersal-limited lineages (occupying forest understory) are significantly more diverse. The longer a lineage has persisted through time was also a good predictor of species diversity, i.e., older lineage accumulated more differentiation.

“The accumulation of bird species in the Neotropical landscape occurred through a repeated process of geographical isolation, speciation and expansion, with the amount of species diversity within lineages influenced by how long the lineage has persisted in the landscape and its ability to disperse through the landscape matrix.”

All in all, the paper doesn’t refute the vicariance hypothesis, but highlights the role of dispersal. These findings add to the ever-increasing pile of possible explanations for the higher diversity of the tropics and its heated discussion.

Smith, Brian Tilston, John E McCormack, Andrés M Cuervo, Michael J Hickerson, Alexandre Aleixo, Carlos Daniel Cadena, Jorge Pérez-Emán, et al. 2014. “The Drivers of Tropical Speciation.” Nature, September. doi:10.1038/nature13687.