Despite being among the most researched animals on Earth, there are currently hundreds of mammals in the wild waiting to be discovered. This is what a new study says that has used a prediction model based on machine learning or automatic learning.
The results of machine learning suggest that most of these unknown animals are small, such as bats, rodents and shrews . Their size has probably made it difficult for experts to identify morphological differences, which means that some species have been grouped together, taxonomically speaking.
“Small, subtle differences in appearance are harder to notice when looking at a tiny animal weighing 10 grams than when looking at something human-sized,” explains biologist Bryan Carstens of The Ohio State University, and one of the authors of the research published in PNAS .
Scientists call these hidden species “biodiversity wildcards” because until they are proven to exist, they cannot be included in evolutionary theory, food webs, or conservation work.
According to the most recent prediction models, more than 80% of mammals have probably received a formal classification . With more than 6,400 species of mammals described in the records, that would mean that there are still more than a thousand unknown species pending classification.
Using machine learning to analyze the genetic sequences and geographic and biological data of more than 4,000 mammals, researchers have identified which taxa (groups of related organisms, which have been grouped together in a given classification) are most likely to harbor hidden species . By figuring out the habitat of these taxa, scientists have helped highlight ecosystems for future taxonomic research.
For example, Southeast Asia is projected to be a hub for unidentified mammal species. Both evolutionary and genetic models estimate that this region of the world contains the highest proportion of diversity wildcards relative to species richness.
A study in Indonesia recently identified 14 new species of shrews , the largest set of new mammals detailed in a single paper since 1931.
The current model found that, worldwide, most undescribed mammals belong to orders that include bats, rodents, and shrews and hedgehogs. These orders also tend to be found in larger geographic areas with high temperature and rainfall variability, such as tropical rainforests .
The results are what taxonomists have long suspected. Tropical rainforests, in general, harbor the greatest diversity of mammals in the world, and as of 1992, most of the newly described mammals are small in size, occur over large areas, and are adapted to habitats that experience daily and seasonal oscillations in temperatures. precipitation and temperature.
“Our study reinforces existing calls for more investment in taxonomic research, especially in understudied and undescribed taxa facing silent extinction ,” the authors write. The research “suggests that hidden species exist in predictable places, awaiting formal description.”
History shows us that, when the human being proposes it, the tree of life grows with new species. One of the main reasons why more species of mammals have been described than other orders is simply the size of these animals. And it is that, the larger the size, the more easily they are seen . In addition, humans, in general, tend to be more interested in the species that are more closely related to us. All this also means that the resources devoted to mammalian research are greater than, for example, those directed to invertebrates.
Considering what can be achieved with the right approach, the authors of the recent study call for renewed funding and interest in taxonomic research to bridge the gap between known and unknown mammalian species.
“That knowledge is important for people doing conservation work. We can’t protect a species if we don’t know it exists ,” says Carstens.
Reference:
J. Parson et al. 2022. Analysis of biodiversity data suggests that mammal species are hidden in predictable places. PNAS. https://doi.org/10.1073/pnas.2103400119
