REASSESSING OUR EXPECTATIONS FOR MARINE LATITUDINAL BIODIVERSITY GRADIENTS IN MODERN AND ANCIENT SYSTEMS

Systematically decreasing taxonomic richness from the equator to the poles has become the first-order prediction for the geographic distribution of global biodiversity throughout Earth history. Many constant geophysical properties of the tropics - e.g., warm temperatures, high insolation, and a large areal extent - potentially facilitate the accumulation of biodiversity. However, several recent analyses of large empirical datasets have challenged the ubiquity of this pattern in both modern and ancient marine systems, observing bimodal distributions or extra-tropical peaks in richness. If these contrarian findings are correct, then latitudinal biodiversity gradients should not necessarily be considered the null hypothesis.

Here, we turn to forward modelling to ask whether classical latitudinal biodiversity gradients (unimodal richness distributions centred near the equator) are even expected in modern marine settings given the Earth's current geophysical state. We use the Huisman-Olff-Fresco response curve models in conjunction with empirically-derived thermal tolerances of skeletonized marine invertebrate genera, modern sea-surface temperature satellite data, and estimates of continental shelf area to generate hypothetical latitudinal biodiversity gradients.

We find that even if distributions are biased such that >90% of generated taxa are set to most prefer tropical temperatures (>20C), the predicted latitudinal biodiversity gradient is bimodal. This is true whether biodiversity is measured at the alpha- or gamma-levels. The location of these peaks are comparable to empirical patterns in the Ocean Biogeographic Information System and recent meta-analysis of the literature under a variety of model parameterizations. Our findings suggest that even if we assume that taxa strongly prefer tropical temperatures, a classic latitudinal biodiversity gradient is not necessarily expected in modern or ancient marine systems.