Were Semi-Arid Areas Wetter or Drier during the Pliocene?

Abstract:

Drying is projected for many semi-arid areas under global warming scenarios. The Pliocene provides a natural analog for global warmth, allowing us to test the terrestrial response to past warming. Today North East Africa (O to 12 N, 38 to 50 E) includes C3 shrubs in arid areas, C4 grasslands in semi-arid areas and C3 trees in the moister uplands. Here we take a biomarker approach to reconstructing terrestrial vegetation and hydrological changes in NE Africa during the Pliocene and develop complementary records of sea surface temperature (SST) changes in the Western Indian Ocean. We find the plant leaf wax carbon isotopic negative anomaly to be ambiguous. Woody cover interpretations suggest afforestation with C3 trees. But an alternative explanation, an increase of dry C3 shrublands, appears more likely based on drying evidence from hydrogen isotopes and pollen for an expansion of shrub taxa. Either change implies a contraction of the Miocene C4 grasslands. To explore the mechanisms for drying we look to the Western Indian Ocean and find biomarker trends show a coincident cooling (in contrast to Mg/Ca), providing a mechanism to reduce moisture supply to the continent. We are adding to the resolution and coverage of the d13C, dD, pollen and TEX86 records in order to assess whether this Pliocene scenario is robust. Specifically we are testing the hypothesis that a cooling trend in the Western Indian Ocean is coincident with a drying and C3 shrubland expansion. Our work to resolve these quite different scenarios (afforestation or aridification on land) is key, since model simulations have offered up both wetter and drier regional responses to Pliocene global warmth and questions about the direction of SST changes (warming or cooling Indian Ocean across the Pliocene) forms part of the larger present debate about the fidelity of Pliocene SST proxies.