the extinction of animal and plant species is one of the crises of our analog world, which is not alone cope with digital resources. However, networked computers can actually help (they are now even essential) to identify the most critical developments and – in terms of the geographic focus of nature conservation – the habitats to identify with the greatest need for protection. For years, researchers collect the necessary data, but their resources are extremely limited.

Joachim Müller-Jung

editor in the features section, responsible for the area of “nature and science”.

F. A. Z. Twitter

Almost avant-garde seems to be that what have developed an American Team to Anahi Espindola, University of Maryland, and Tara Pelletier of Radford University, with the Algorithms of a learning machine: a risk model to allow the current hazards to plants in any habitat – from the rainforest to the city Park. A kind of Red list from the machine room to a certain extent.

more species are Endangered than we thought

Indeed, of the world Union for conservation of nature, regularly updated Red list for plants is very patchy: less than 6.5% of known land plant species, one knows the risk status. The researchers have now used some of the most reliable collection and risk data pool, in order to train the Computer so that the statement have been legally determined risk parameters (distribution, biology, and temperature anomalies).

This pattern has been applied to more than 150,000 species, and the plants in the five hazard categories are sorted. Regions with significant (and expected) of extinction risks, such as the jungles of Central America, Australia, and Madagascar have been identified, but also unexpected “Hotspots”: California, for example, or the Arabian Peninsula. A good ten percent of the KI-tested species are threatened, the researchers report (doi: 10.1073/pnas.1804098115): “globally, a lot more types of plants in your inventory are at risk as was to be expected.”