Comments on ‘U-Pb dated flowstones restrict South African early hominin record to dry climate

Volume 116| Number 3/4 March/April 2020 Research Letter https://doi.org/10.17159/sajs.2020/7094 © 2020. The Author(s). Published under a Creative Commons Attribution Licence. Comments on ‘U-Pb dated flowstones restrict South African early hominin record to dry climate phases’ (Pickering et al. Nature 2018;565:226–229) AUTHORS: Dominic Stratford1 Laurent Bruxelles1,2,3 J. Francis Thackeray4 Travis R. Pickering5 Sophie Verheyden6


HOW TO CITE:
Stratford D, Bruxelles L, Thackeray JF, Pickering TR, Verheyden S. Comments on 'U-Pb dated flowstones restrict South African early hominin record to dry climate phases' (Pickering et   from eight Pliocene and Pleistocene South African Cradle of Humankind cave sites to propose that biases were created within the fossil record due to absent clastic sedimentation phases during wet periods, when caves were closed and only speleothems accumulated. Such a scenario has significant implications for our understanding of variability in hominin mobility, resource exploitation, functional repertoires and interactions with competitors in changing environmental and ecological contexts. We find considerable issues with the article. First, Pickering et al.'s contribution omits crucial fossil evidence from various stratigraphic units of the Sterkfontein Caves that indicates conditions were not always arid when the caves were open and sediments were deposited. Second, Pickering et al.'s proposal that clastic and speleothemic deposits (including faunal and floral material) form mutually exclusively is an overly simplified, binary depositional (and in this case environmental) framework that demonstrates an inherent bias in the sampling of cave deposits for dating. This creates the impression that either speleothems or clastic sediments are deposited and does not take into account the full spectrum of sedimentary complexity in karst caves. Third, closure of the caves across the Cradle of Humankind landscape during wet periods is not substantiated geomorphologically or speleologically; identification of the responsible process is critical to the proposed infilling scenario.

Significance:
• We propose that Pickering et al.'s interpretation of the environmental context of the South African early hominin record is problematic in that it omits crucial faunal and floral fossil evidence associating hominins with non-arid climates, is geomorphologically unsupported, and perpetuates biases against temporally and climatically representative clastic sediments due to challenges related to their dating.

Pickering et al. (2018) 1 utilise calcium carbonate flowstone deposits from eight Pliocene and Pleistocene South African
Cradle of Humankind (also referred to as 'the Cradle') cave sites to propose that biases are created within the fossil record due to absent clastic sedimentation phases during wet periods. They propose a close correlation between cave closure, flowstone growth and 'phases of increased effective precipitation', suggesting that clastic sedimentation and calcium carbonate precipitation are mutually exclusive processes, with clastic deposits (including faunal and floral material) forming only during arid periods. This scenario has significant implications for our understanding of variability in hominin mobility, resource exploitation, functional repertoires and interactions with predators in changing environmental and ecological contexts. We find this proposition to be problematic as a result of oversimplification.
Reducing environmental conditions to a 'binary' framework of such extreme end-member states of wet or dry does not realistically reflect the majority of time represented by intermediate, moderately variable climatic conditions during which deposits also accumulate. Reducing sediment accumulation to one end of the climatic spectrum limits the validity of nuanced interpretations of hominin ecological relationships represented by the diverse fossil assemblages interred in the extensive clastic deposits. Specifically, we find the following issues with the article: 1. Published evidence clearly associating hominins in part with wet ecological conditions, was not considered seriously enough in the article. For example, the authors omit crucial fossil evidence from the Sterkfontein Caves of liana vines that indicate the presence of ancient dense woodland or forest at ~2.5 million years ago. 2 The authors also omit important faunal data, notably the presence of fossil colobine monkeys, Alcelaphini bovids such as Damaliscus sp. and Megalotragus sp. (which are dependent on savanna grassland), and other broken-open country antelopes like Tragelaphus sp. and Antidorcas from various Sterkfontein stratigraphic levels. [3][4][5][6] These reflect at least some degree of savanna woodland, which, in turn, contradicts inferences of dry conditions. Similarly, micromammal samples from Sterkfontein 7 demonstrate a mosaic environment (that importantly include moist, woodland settings as suggested by the presence of Elephantulus fuscus fossils), and represent interglacial (comparatively warm and wet) conditions which prevailed during sedimentation and fossil accumulation. Further, stable carbon isotope data derived from tooth enamel of a range of taxa, including hominins 8-10 , relate to diets that reflect consumption of a significant amount of C3 vegetation, indicating, in turn, that conditions were certainly not always arid at times when the caves were open and sediments were being deposited.
2. Generally, dates for clastic sediment deposits are ascertained through the application of U-Pb or palaeomagnetic dating of interstratifying (or assumed to be interstratifying) flowstones. 11 The clastic sediments (collectively known as breccias) themselves are not often suitable for dating using the comparatively broadly applied palaeomagnetic seriation and U-Pb methods. [12][13][14] Consequently, often only speleothems are sampled, creating the impression that speleothem or clastic sediments are mutually exclusively deposited. Clastic sediments deposited nearby at the same time will largely remain undated and therefore overlooked.  19 makes application of a uniform Cradle-scale cave closure process challenging and in this case unsupported. While speleothem deposition seems to have a strong relationship with climate, more specifically with water and CO 2 availability 20 , the opening and closing of a cave system seems to be primarily determined by erosional cycles (and tectonic changes), and is therefore potentially strongly influenced at the local scale and only indirectly influenced by climate. While little literature exists on the mechanisms of cave entrance opening and closure in palaeokarst, authors describe vadose zone openings as enlarging during wetter periods due to increased recharge flow rates. 21,22 These processes result in a complex succession of sedimentation processes 23 and sedimentary deposition and speleothem deposition should not be considered mutually exclusive phenomena.
In summation, faunal and floral evidence clearly indicate that hominin fossils are not necessarily always associated with arid climates and that fauna and flora accumulate in open caves through a broad range of environmental conditions. Pickering et al.'s methodology for sampling cave deposits for dating, which focuses on the application of U-Pb or palaeomagnetic techniques to interstratifying (or assumed to be interstratifying) flowstones, produces ages (and by proxy cave sedimentation conditions) only for speleothems. The proposed correlation of speleothem formation with cave closure creates a convenient but largely unsupported deterministic interpretation of palaeoenvironmental conditions that necessarily excludes the significant clastic deposit record. Their resulting interpretation of the environmental context of the South African hominin record consequently perpetuates biases against temporally and climatically representative clastic sediments. Ideally, palaeoenvironmental reconstructions should be developed that integrate clastic and speleothemic evidence in coherent sequences prior to isolating and excluding evidence.