Evolution of tropical watersheds and continental hydrology during the Late Cretaceous greenhouse; impact on marine carbon burial and possible implications for the future
Regional climate modelling for the Late Cretaceous greenhouse and high-resolution marine stratigraphic records from both sides of the low latitude Atlantic show that tropical South American and African hydrology and watersheds had a strong effect on freshwater transfer into the Equatorial Atlantic and subsequently the marine carbon record. This conclusion is derived from new detailed geochemical records from Demerara Rise off Suriname drilled at Ocean Drilling Program (ODP) Site 1261 combined with frequency analyses and climate simulations providing evidence for mainly eccentricity-driven changes in carbon burial in the western tropical Atlantic. Shorter orbital frequencies, in particular precession, clearly dominating black shale cycles off tropical Africa (ODP Site 959), are far less dominant at Demerara Rise despite comparable time resolution of the geochemical records. We suggest that these different frequency patterns in carbon burial were related to the regional evolution of Cretaceous watersheds and hydrology in tropical South America and Africa. River discharge deduced from simulations indicates higher and less variable discharge from South America compared to western Africa at that time. This runoff pattern would have supported more permanent anoxic conditions off South America compared to Africa, at least indirectly, and caused the lack of strong higher frequency geochemical cycles in the western sector of the Equatorial Atlantic. Furthermore, climate simulations show a general switch of primary runoff from either side of the Cretaceous Equatorial Atlantic every half precession cycle (i.e. every ~ 10 kyr). Similarities between the developments of Cretaceous and Holocene hydrology in the tropical Atlantic area imply that orbital-scale evolution of watersheds is a robust feature through time that is independent from the mean global climate state. Based upon the comparison of Cretaceous and Holocene trends in hydrology we infer that future hydrology in the study region may develop in a comparable direction to the one observed in the Cretaceous. If true, this suggests that the modern Amazon rain forest could shrink over the next millennium due to a ~ 30% loss of moisture while the Congo rain forest in Africa is likely to expand in response to a 14% gain in moisture.