TESTING THE CONCEPT OF SIMPSONIAN ADAPTIVE ZONES IN NEOGENE PLANKTONIC FORAMINIFERA

The concept of adaptive zones was developed by Simpson (1944) to describe how species affect each other’s evolution through deep time. It predicts how communities and clades will evolve: species in a full zone compete in a zero-sum evolutionary arms race; species colonising empty zones will be able to radiate taxonomically and ecologically. Adaptive zones therefore provide a way in which factors intrinsic to a clade can interact with extrinsic factors to shape macroevolution. Species’ interactions within zones determine changes in abundance, phenotype, persistence and proliferation, while extrinsic forces – from other biota or the abiotic environment – cause the zones themselves to arise, wax, wane or disappear. Despite the conceptual importance of adaptive zones to theories of Red Queen evolution and adaptive radiation, their importance has proved hard to demonstrate, largely because few clades provide a sufficiently detailed fossil record.

Planktonic foraminifera have perhaps the best fossil record of any group, allowing evolution to be analysed at unusually fine spatial and taxonomic scales. In this analysis we compare samples taken from two ODP sites: 872C in the stable centre of a sub-tropical gyre and 926B on the less stable gyre edge. For samples at 1Ma intervals, we estimated species’ relative abundances by fully counting splits of approximately 300 individuals and characterised morphology using repeatably-measured traits.

We use two approaches to look for evidence of adaptive zones. First, we test whether negatively correlated abundance trajectories are more common between species having similar depth and trophic ecologies than between species that make their livings in different ways. Second, we test whether ecomorphological clusters above the species level are more temporally persistent in the climatically stable gyre centre than in the more variable gyre edge. We also investigate the importance of sample size for the detectability of these patterns.