Biology subjects

Jaimes, A.

Publications and source records attributed to Jaimes, A..

3 recordsLinked to original sources

bioRxiv2026

Revisiting Cucuruchu, a Late Pleistocene site on the Venezuelan Caribbean coast with megafauna and lithic association

The study of historical documents from the original excavation led by J. M. Cruxent, of museum collections, as well as results of fieldwork, are combined to provide a revised assessment of the archeologically site of Cucuruchu, which together with those of Taima-Taima and Muaco, document megafaunal lithic remains associated from the Late Pleistocene in northern Venezuela. Six vertebrate taxa are reported for Cucuruchu: five megaherbivores (a giant sloth, Eremotherium laurillardi.; a glyptodontid, Glyptotherium cf. G. cylindricum; a macrauchenid, cf. Xenorhinotherium bahiense; a proboscidean, Notiomastodon platensis; and an equid, Equus sp.), and a turtle (Kinosternon sp.). Two fragments of El Jobo projectiles were found in situ lying among megaherbivores bones, while another unidentified projectile fragment was also found in the same layer. Although no direct evidence of human exploitation on these megaherbivores remains has been found, the presence of El Jobo projectiles in the undisturbed fossil-bearing layer some of which are associated with fossil remains, could be a suggestive of such interaction. Two new radiocarbon dates indicate ages of at least ca. 15.3-16.0 kyr cal BP for the fossil-bearing layer, contrasting with the mid to late Holocene age originally reported. Cucuruchu is a window to the past and supports the presence of hunter-gatherers of the El Jobo culture in the region at the end of the Pleistocene.

paleontology↗

bioRxiv2026

Real Science Is Harder Than Benchmarks: Evaluating Advanced AI Frameworks on Published Studies. I. Uncertainty Quantification, ML on Therapeutic Data Commons, and Agent-Based Modeling

Artificial Intelligence (AI) frameworks for automating scientific research have shown strong performance on benchmarks, but their capacity to routinely reproduce results from multiple real-life published studies remains largely untested. We evaluated five advanced AI research frameworks (Kosmos, K-Dense, ToolUniverse, BioAgents from bio.xyz, and the AI Scientist-v2 from Sakana AI) on three real-life tasks (including two recently published papers) spanning uncertainty quantification for molecular property predictions, machine learning on Therapeutic Data Commons benchmarks, and agent-based modeling. AI frameworks demonstrated genuine strengths: generating original hypotheses, competently executing routine data acquisition and coding tasks, providing statistical measures of confidence often absent from the original papers, and producing well-formatted final reports. At the same time, our experiments revealed that real-world scientific tasks remain considerably harder than current benchmarks suggest. No AI framework matched the scope or depth of the original studies, results varied across multiple runs of the same framework with the same prompt, and we documented cases of severe hallucinations in final reports, gaps in literature coverage, and overconfident conclusions. Verification of AI outputs required substantial domain expertise. While these three tasks are only partially representative of the broader scientific landscape, they offer a starting point for developing a more rigorous methodology for evaluation of AI performance than what is currently practiced. We conclude that AI frameworks are already valuable for prototyping research directions and stress-testing completed studies, and some of the limitations documented here appear largely tractable through infrastructure improvements and continued development.

bioinformatics↗

bioRxiv2026

Diverse megamammals exploited by humans, chronology and palaeoecology at Taima-Taima, Late Pleistocene, South America

The arrival of humans to the Americas at the end of the Pleistocene has been one of the foundations of the hypothesis that hunter-gatherer groups had a direct role in the extinction of megafauna. However, empirical evidence of the exploitation of megafauna by humans and associated chronological context are still scarce, especially in South America. Using evidence of modifications on osteological remains, we provide a reassessment of the megafauna that was exploited at the classic Taima-Taima site (Venezuela), with an updated chronology and paleoenvironmental inferences. Our results show that, at least, five species of extinct megaherbivores were probably exploited at the site, expanding the previous records, restricted to the proboscidean Notiomastodon platensis and the giant armadillo Glyptotherium cf. G. cylindricum. The new data of the other exploited taxa include the terrestrial sloth Glossotherium, the macrauchenid cf. Xenorhinotherium bahiense and a toxodontid; both killing and slaughtering activities were likely practised at the site. Seven new enamel radiocarbon dates show an age of 15.3-17.8 cal kyr BP for the fossiliferous strata, although this is likely to be a minimum age given the materials propensity for modern carbon uptake. The study of phytoliths and starch grains from the dental calculus of N. platensis identified five plant taxa from the phytoliths: arboreal dicotyledons, Poaceae, Marantaceae, Asteraceae and Arecaceae. Ecometrics studies of dental diversity reveal how no other single site on the continent preserves such a diversity of grazing and browsing megaherbivores exploited by humans.

paleontology↗