Stable isotope variation in tooth enamel from Neogene hippopotamids: Monitor of meso and global climate and rift dynamics on the Albertine Rift, Uganda

The Neogene was a period of long-term global cooling and increasing climatic variability. Variations in African-Asian monsoon intensity over the last 7 Ma have been deduced from patterns of eolian dust export into the Indian Ocean and Mediterranean Sea as well as from lake level records in the East African Rift System (EARS). However, lake systems not only depend on rainfall patterns, but also on the size and physiography of river catchment areas. This study is based on stable isotope proxy data (¹⁸O/¹⁶O, ¹³C/¹²C) from tooth enamel of hippopotamids (Mammalia) and aims in unravelling long-term climate and watershed dynamics that control the evolution of palaeolake systems in the western branch of the EARS (Lake Albert, Uganda) during the Late Neogene (7. 5 Ma to recent). Having no dietary preferences with respect to wooded (C₃) versus grassland (C₄) vegetation, these territorial, water-dependant mammals are particularly useful for palaeoclimate analyses. As inhabitants of lakes and rivers, hippopotamid tooth enamel isotope data document mesoclimates of topographic depressions, such as the rift valleys and, therefore, changes in relative valley depth instead of exclusively global climate changes. Consequently, we ascribe a synchronous maximum in ¹⁸O/¹⁶O and ¹³C/¹²C composition of hippopotamid enamel centred around 1. 5-2. 5 Ma to maximum aridity and/or maximum hydrological isolation of the rift floor from rift-external river catchment areas in response to the combined effects of rift shoulder uplift and subsidence of the rift valley floor. Structural rearrangements by ~2. 5 Ma within the northern segment of the Albertine Rift are well constrained by reversals in river flow, cannibalisation of catchments, biogeographic turnover and uplift of the Rwenzori horst. However, a growing rain shadow is not obvious in ¹⁸O/¹⁶O signatures of the hippopotamid teeth of the Albertine Rift. According to our interpretation, this is the result of the overriding effect of evaporation on ¹⁸O/¹⁶O responding to aridification of the basin floor by a valley air circulation system through relative deepening of the valley. On the other hand, a synchronous arid pulse is not so clearly recorded in palaeosol data and mammalian fauna of the eastern branch of the EARS. This discrepancy indicates that rift mesoclimates may represent an underestimated aspect in previous palaeoclimate reconstructions from rift valley data and represent a clear limitation to attempts at global climate reconstructions. The results of this study also suggest that using ¹⁸O/¹⁶O data as a proxy to rain shadow evolution must take into account relative basin subsidence to properly document mountain range uplift.