This year, for the first "real" lecture of my evolutionary biology class, I gave an overview of the history of the Earth, from the Big Bang to the present. It went fast, and I only had a couple of slides at the end for one of the geological processes most responsible for current patterns of biodiversity: the climate cycles of the Pleistocene. Periods of warming and cooling, and accompanying changes in sea level and glacial coverage, were engines of diversification, subdividing species’ ranges into refugia, then allowing species pushed towards the equator by advancing ice sheets to expand towards the poles again. These patterns are evident all over terrestrial temperate regions today, and a paper published over the summer in The Molecular Ecologist shows how the impacts of Pleistocene climate change extended beyond land, into marine communities.
Brendan Reid and coauthors from New York, Brazil and Australia use mitochondrial DNA sequence data to take a look at the demographic history of sea turtles since the last glacial maximum. There are seven recognized extant species of sea turtle, many of which are found in both the Atlantic and Indo-Pacific Oceans — and the mitochondrial data show signs of isolation between those large regions, with haplotypes sorted into either the Atlantic basin or the Indo-Pacific. All oceans are linked, but there are limits: the Isthmus of Panama, and cold currents around the southern tip of Africa. Sea turtles are also limited by a critical connection to land, their nesting beaches. Between changing land and ocean temperatures associated with glacial cycles, it makes sense to expect that sea turtles probably retreated towards the tropics during the last ice age, and have expanded their ranges north and south since the last glacial maximum, which ended about 20,000 years ago.
And that’s what the mitochondrial DNA suggests. Reid et al used Approximate Bayesian Computation based on coalescent simulations to evaluate whether each of the seven sea turtles have had population expansions, and the most likely timing of that expansion. A demographic model including population expansion was the best fit, compared to models with no change in population size, or a decline, for a majority of the mitchondrial lineages represented in the dataset, and the estimated timing of that expansion was consistent across lineages — about 11,000 years ago.
The authors also used Bayesian model averaging to see whether differences in geography and life-history traits among the seven sea turtle species could explain which had expanded their ranges most. Where turtles lived — in the Atlantic or Indo-Pacific, and their overall range size within those basins — had more impact on their population history since the LGM than differences in their body size, generation time, or dietary specialization. Turtles in the Atlantic were more likely to have expanded their ranges, possibly because the Pacific has had a larger climactically stable area during the Pleistocene.
The climate shifts of the Pleistocene occurred much more slowly than contemporary climate change driven by fossil fuel pollution, but these results do suggest something of how sea turtles will cope (or not) with warming global climate. The authors suggest that if the Pacific basin has created a refuge of relatively stable climate in the past, it could do so in the future — and that would mean protection of these regions will be critical to the survival of species that can shelter in them.
Reid, BN, E Naro-Maciel, A Torres Hahn, NN FitzSimmons, and M Gehara. 2019. Geography best explains global patterns of genetic diversity and post-glacial co- expansion in marine turtles. Molecular Ecology doi: 10.1111/mec.15165
Zeugner S and M Feldkircher. 2015. Bayesian model averaging employing fixed and flexible priors: The BMS package for R. Journal of Statistical Software, 68(4): 1-37. doi: 10.18637/jss.v068.i04