An Argentine scientist, co-author of a method to predict which plants are in danger

A group of  scientists  from the United States, including an Argentine one, developed, through an artificial intelligence system, a classification methodology that will allow us to  predict whether or not a  plant  is in  danger of extinction  . The International Union for Conservation of Nature (IUCN) is continuously working on the creation of a so-called  Red List , which analyzes species and classifies them into categories ranging from 'minor concern' to 'extinct', to intermediate categories.  These classifications are used to define conservation actions , so they are of “enormous utility”, but “for species to be included they have to be individually evaluated, requiring a predefined protocol, available funds and the presence of specialists who perform the evaluation, making it a slow process,” said  Anahí Espindola , Argentina's co-author of the study. “The base method we used was 'random forest', which is known for its ability to classify and predict data,” he said and explained that “in this case we try to predict the probability that a species is endangered or not, using data related to characteristics of its range of distribution, their preferred climatic conditions and some morphological characteristics”.

Espindola, professor of entomology at the University of Maryland in the United States, explained that “this method allows to use all species that have already been evaluated by IUCN to train and  create our random forest classification , using the characteristics of the species as predictor variables”. “Once we have obtained a sufficiently accurate classification model, we can use that same model on species for which we know the characteristics used in the model (range of distribution, preferred climatic conditions and morphology) but for which we do not know the level of risk of extinction”. In this regard, the co-author of the article published in the specialized journal PNAS stated that the use and classification of these data will allow “to calculate the likelihood that these species not yet evaluated by the IUCN Red List are at risk.” The “extremely useful” thing about this system is that it is “relatively accurate, and also  analysable without having access to important computing resources “, it also has the “advantage” of being entirely based on “public data” (open access), is say that anyone can perform these analyzes and use their results. “In addition, this method can be adapted to any geographical or taxonomic scale of interest, as it can also be used at national, regional, or local levels and allows to assist in the identification of species that should be prioritized by IUCN,” he pointed out. Oacute; Spindola and qualified this tool as “useful and complementary to these assessments”. The specialist said that of the 150,000 species analyzed, “about 10% (15,000) have a high probability of belonging to conservation categories other than 'minor concern'”. “From a global point of view, we identify regions that are more likely to have endangered species, such as some Andean regions in northern South America, or the  atl forest  . aacute; brazilian ntico. These regions are characterized by a high level of endemism, and the presence of many rare species,” he concluded.

Source: Telam