Where to Dig for Fossils: Combining Climate-Envelope, Taphonomy and Discovery Models

Challenges of classifying and mapping perennial freshwater systems within highly variable climate zones: A case study in the Murray Darling Basin, Australia

Perennial freshwater systems are valuable natural resources that provide important ecological services globally. However, in highly variable climates, such as Australia, water availability in rivers and streams can vary greatly from year to year and from decade to decade. Further, across Australia and many other regions, perennial river systems are projected to decrease because of anthropogenic climate change, placing the ecosystems they support under additional pressure. Quantifying the potential impacts of climate change on perennial freshwater systems requires robust databases of existing water features with accurate classifications. This is a challenge for rivers that display a high degree of interannual variability since the river classification can be dependent on the period of available data. In this study, we carry out a regional scale comparison of three different spatial databases commonly used in environmental and ecological assessments of perennial systems of Australia, namely Geodata, Geofabric and Water Observations from Space (WOfS). Focusing on the southern Murray Darling Basin (MDB), due to its national and international significance and its highly variable flow regimes, we show that no single spatial database is reliable by itself in terms of perennial water classification, with notable differences likely arising from variations in the periods analysed and methods used to classify the systems. Further, an analysis of high-quality gauged streamflow data (with approximately 40-year daily records) for four sub-catchments, and long-term simulation data (>100 years) for two sub-catchments in the lower MDB, confirm that flow persistence can be non-stationary through time, with some 'perennial' systems exhibiting sustained periods of cease to flow (i.e. becoming non-perennial) during prolonged droughts. This study demonstrates that due consideration is required in developing baseline classification of perennial freshwater systems for assessing future changes and measuring adaptive capacity.

Climatic and tectonic drivers shaped the tropical distribution of coral reefs

Today, warm-water coral reefs are limited to tropical-to-subtropical latitudes. These diverse ecosystems extended further poleward in the geological past, but the mechanisms driving these past distributions remain uncertain. Here, we test the role of climate and palaeogeography in shaping the distribution of coral reefs over geological timescales. To do so, we combine habitat suitability modelling, Earth System modelling and the ~247-million-year geological record of scleractinian coral reefs. A broader latitudinal distribution of climatically suitable habitat persisted throughout much of the Mesozoic-early Paleogene due to an expanded tropical belt and more equable distribution of shallow marine substrate. The earliest Cretaceous might be an exception, with reduced shallow marine substrate during a 'cold-snap' interval. Climatically suitable habitat area became increasingly skewed towards the tropics from the late Paleogene, likely steepening the latitudinal biodiversity gradient of reef-associated taxa. This was driven by global cooling and increases in tropical shallow marine substrate resulting from the tectonic evolution of the Indo-Australian Archipelago. Although our results suggest global warming might permit long-term poleward range expansions, coral reef ecosystems are unlikely to keep pace with the rapid rate of anthropogenic climate change.

Unsupervised learning of satellite images enhances discovery of late Miocene fossil sites in the Urema Rift, Gorongosa, Mozambique

Background

Paleoanthropological research focus still devotes most resources to areas generally known to be fossil rich instead of a strategy that first maps and identifies possible fossil sites in a given region. This leads to the paradoxical task of planning paleontological campaigns without knowing the true extent and likely potential of each fossil site and, hence, how to optimize the investment of time and resources. Yet to answer key questions in hominin evolution, paleoanthropologists must engage in fieldwork that targets substantial temporal and geographical gaps in the fossil record. How can the risk of potentially unsuccessful surveys be minimized, while maximizing the potential for successful surveys?

Methods

Here we present a simple and effective solution for finding fossil sites based on clustering by unsupervised learning of satellite images with the k-means algorithm and pioneer its testing in the Urema Rift, the southern termination of the East African Rift System (EARS). We focus on a relatively unknown time period critical for understanding African apes and early hominin evolution, the early part of the late Miocene, in an overlooked area of southeastern Africa, in Gorongosa National Park, Mozambique. This clustering approach highlighted priority targets for prospecting that represented only 4.49% of the total area analysed.

Results

Applying this method, four new fossil sites were discovered in the area, and results show an 85% accuracy in a binary classification. This indicates the high potential of a remote sensing tool for exploratory paleontological surveys by enhancing the discovery of productive fossiliferous deposits. The relative importance of spectral bands for clustering was also determined using the random forest algorithm, and near-infrared was the most important variable for fossil site detection, followed by other infrared variables. Bands in the visible spectrum performed the worst and are not likely indicators of fossil sites.

Discussion

We show that unsupervised learning is a useful tool for locating new fossil sites in relatively unexplored regions. Additionally, it can be used to target specific gaps in the fossil record and to increase the sample of fossil sites. In Gorongosa, the discovery of the first estuarine coastal forests of the EARS fills an important paleobiogeographic gap of Africa. These new sites will be key for testing hypotheses of primate evolution in such environmental settings.

Extinction of eastern Sahul megafauna coincides with sustained environmental deterioration

Explanations for the Upper Pleistocene extinction of megafauna from Sahul (Australia and New Guinea) remain unresolved. Extinction hypotheses have advanced climate or human-driven scenarios, in spite of over three quarters of Sahul lacking reliable biogeographic or chronologic data. Here we present new megafauna from north-eastern Australia that suffered extinction sometime after 40,100 (±1700) years ago. Megafauna fossils preserved alongside leaves, seeds, pollen and insects, indicate a sclerophyllous forest with heathy understorey that was home to aquatic and terrestrial carnivorous reptiles and megaherbivores, including the world’s largest kangaroo. Megafauna species diversity is greater compared to southern sites of similar age, which is contrary to expectations if extinctions followed proposed migration routes for people across Sahul. Our results do not support rapid or synchronous human-mediated continental-wide extinction, or the proposed timing of peak extinction events. Instead, megafauna extinctions coincide with regionally staggered spatio-temporal deterioration in hydroclimate coupled with sustained environmental change.

The causes of the Upper Pleistocene megafauna extinction in Australia and New Guinea are debated, but fossil data are lacking for much of this region. Here, Hocknull and colleagues report a new, diverse megafauna assemblage from north-eastern Australia that persisted until ~40,000 years ago.

Application of artificially intelligent systems for the identification of discrete fossiliferous levels

The separation of discrete fossiliferous levels within an archaeological or paleontological site with no clear stratigraphic horizons has historically been carried out using qualitative approaches, relying on two-dimensional transversal and longitudinal projection planes. Analyses of this type, however, can often be conditioned by subjectivity based on the perspective of the analyst. This study presents a novel use of Machine Learning algorithms for pattern recognition techniques in the automated separation and identification of fossiliferous levels. This approach can be divided into three main steps including: (1) unsupervised Machine Learning for density based clustering (2) expert-in-the-loop Collaborative Intelligence Learning for the integration of geological data followed by (3) supervised learning for the final fine-tuning of fossiliferous level models. For evaluation of these techniques, this method was tested in two Late Miocene sites of the Batallones Butte paleontological complex (Madrid, Spain). Here we show Machine Learning analyses to be a valuable tool for the processing of spatial data in an efficient and quantitative manner, successfully identifying the presence of discrete fossiliferous levels in both Batallones-3 and Batallones-10. Three discrete fossiliferous levels have been identified in Batallones-3, whereas another three have been differentiated in Batallones-10.

FosSahul 2.0, an updated database for the Late Quaternary fossil records of Sahul

The 2016 version of the FosSahul database compiled non-human vertebrate megafauna fossil ages from Sahul published up to 2013 in a standardized format. Its purpose was to create a publicly available, centralized, and comprehensive database for palaeoecological investigations of the continent. Such databases require regular updates and improvements to reflect recent scientific findings. Here we present an updated FosSahul (2.0) containing 11,871 dated non-human vertebrate fossil records from the Late Quaternary published up to 2018. Furthermore, we have extended the information captured in the database to include methodological details and have developed an algorithm to automate the quality-rating process. The algorithm makes the quality-rating more transparent and easier to reproduce, facilitating future database extensions and dissemination. FosSahul has already enabled several palaeoecological analyses, and its updated version will continue to provide a centralized organisation of Sahul's fossil records. As an example of an application of the database, we present the temporal pattern in megafauna genus richness inferred from available data in relation to palaeoclimate indices over the past 180,000 years.

Mudslide and/or animal attack are more plausible causes and circumstances of death for AL 288 ('Lucy'): A forensic anthropology analysis

Following a global morphological and micro-CT scan examination of the original and cast of the skeleton of Australopithecus afarensis AL 288 ('Lucy'), Kappelman et al. have recently proposed a diagnosis of a fall from a significant height (a tree) as a cause of her death. According to topographical data from the discovery site, complete re-examination of a high-quality resin cast of the whole skeleton and forensic experience, we propose that the physical process of a vertical deceleration cannot be the only cause for her observed injuries. Two different factors were involved: rolling and multiple impacts in the context of a mudslide and an animal attack with bite marks, multi-focal fractures and violent movement of the body. It is important to consider a differential diagnosis of the observed fossil lesions because environmental factors should not be excluded in this ancient archaeological context as with any modern forensic anthropological case.