PageRank Algorithm Used for the Ecosystem

Researchers at the University of Chicago had the idea of adapting the PageRank algorithm to the interactions in nature to predict failures in the ecosystem.

Nature, if we include the fauna and flora, is a large web where each eats other, so more or less depend heavily on the existence of another species of animal or plant.

The dependence of a species, for example Panda to bamboo is as links from the latter to the former. Food is the link between species. Pandas are very popular with bamboos as a site is popular among other sites and get lots of links from them.
Suppress the popular site and all sites receiving links from it lose their ranking and therefore their traffic.

Researchers have needed only minor modifications to the algorithm to adapt it to the ecosystem, but the principle has been reversed.
On the Web, a page is important if many pages point to it. In nature, conversely, a species is important if it points to important species. In our example, bamboo has many links to pandas.

The supply is still very complex. The death of a creature, the researcher notes, brings his body to the compost that nourishes others.

By comparing the PageRank algorithm adapted to the nature, and other computational methods already employed in the field, much more difficult to implement, the researchers found they gave the same results.

PageRank is expected to help save species endangered by a better understanding of the ecosystem and greater predictability.

References

PLoS Computational Biology. Dr. Stefano Allesina and Dr. Mercedes Pascual.