Note from the General Editor

Nota del Editor General

In the last years, Artificial Intelligence (AI) has revolutionized the methodologies of scientific research and the writing of scientific papers, as well as the editorial management of scientific journals and academic books. This technology -which allows computers to simulate the human intelligence and capacities for problem resolution- involves the development of algorithms modelled from the processes of decision making of the human brain that can “learn” from the available data to make classifications and predictions.

The benefits of using AI are clear, among them, the increasing efficiency of the processes of investigation, analysis and interpretation of results, and decision making. In the international scientific community, notwithstanding, concerns have been raised on the use of Large Language Models (LLMs) -such as ChatGPT, Google Bard, or Bing– in the activities that are their own. These concerns are principally centered in ethical aspects, and in the authenticity and integrity of the scientific publications due to the possible fraudulent or malicious use of the tools. In fact, challenges have been posed to avoid the damages that can be derived from the use of AI such as reinforcement of biases, lack of data privacy (particularly important in research with human beings), perpetuation of inaccuracies, and the potential to reduce critical thinking due to overconfidence in the tools. The reduction in critical thinking, particularly among the youngest researchers who are more prone to use AI due to their familiarization with technologies, might have undesirable consequences on the advancement of the scientific knowledge and its applications.

For the previously mentioned, the need has arisen to develop guides or protocols for the use of AI in scientific research and writing, and in editorial management to insure an ethical and responsible application. In relation to this point, minimum consensus have been reached by the international community, namely: (a) AI used for the development of a work or the elaboration of a manuscript cannot be cited as an author because only human beings can be responsible for the content of a scientific article; (b) not being legal entities, AI models cannot make statements regarding conflicts of interest or manage copyrights or licenses of use; (c) editors and reviewers are responsible for the evaluations, opinions, and decisions on the manuscripts handled by them. In short, the various actors of the scientific system are fully responsible of their actions -including those performed with AI- and, thus, of any unethical breach.

It is fundamental, then, that users of AI choose the tools according to the benefits that could be obtained from them, but in full knowledge of the limitations and possible damages that might be derived from their improper use. Likewise, it is a must that the international community continue the work on the development of guides and protocols to insure the responsible and ethical use of AI in the scientific field.

Elsa L. Camadro

ARTICLE 1 – research

Bessega C., Pometti C., Saidman B.O., Fortunato R., Santoro C.M., Mcrostie V., Vilardi J.C.

Genealogical relationships among DNA lineages considering their current geographic distributions are useful to infer historical events that have shaped the contemporary distributions of species and their genetic variation. In this study we analyzed the variation of the nadhF-rpl32 intergenic spacer in Prosopis alba Griseb. samples (algarrobo) collected in Chile, Argentina and Bolivia in order to contribute to our understanding of the evolutionary history of this species in southern South-America. We assessed the influence of environmental conditions on the demographic history of them by using a Bayesian ecological clustering (BPEC) approach and simulations based on the theory of coalescence. The results obtained allowed us to identify nine haplotypes. The Tajima (TD= -1.35) and Fu (Fs= -2.36) tests were non-significant, suggesting absence of selection. On the other hand, the disparity between sequences or raggedness (rg=0.021) was also non-significant, compatible with the population expansion. The coalescent analysis using MCMC indicated that the best fit demographic model was the linear growth one, with a time to the most recent common ancestor, for the haplotypes sampled in the present analysis, TMRA=0.0072, that is, roughly 7,000 generations. The BPEC analysis identified two clusters whose distribution partially overlaps in the Atacama Desert (Chile) and allows us to postulate that the species would have expanded to the north and west from the Chaqueña Region in Argentina. The comparison of scenarios by means of ABC (Approximate Bayesian Computation) analyses is in accordance with this result as the cases where the East cluster or the Argentinean samples were postulated as ancestral, yielded the higher posterior probabilities. The analysis performed contributed to the P. alba historical reconstruction throwing light on the trans Andean movement considering direction, time and natural- and human-mediated dispersal agents.

Key words: algarrobo, Atacama Desert, Bayesian analysis, coalescent models, nadhF-rpL32 intergenic spacer, population genetics
Language: English