Article: How AI is Battling the Spread of Invasive Species

Photo Source: Flickr

Recently, in the UK, researchers collaborated with naturalists and wildlife conservationists to pilot a new artificial intelligence-powered tool to curtail the spread of the non-native grey squirrel. The eastern grey squirrel (Sciurus carolinensis), found natively in North America's eastern and midwestern regions, lives primarily in trees and serves as a forest regenerator. Characterized by its large, bushy grey tail and white underbelly, this rodent will be familiar to many in the US in both urban and rural areas. Grey squirrels are considered scatter-hoarders because they bury nuts and seeds in little caches in the ground or hide them in trees for later retrieval. Over the course of a season, one grey squirrel will make several thousand caches of food for later consumption. Their acute memory and spatial reasoning using landmarks help them find their hidden food later in the year to keep them and their newborns, called kits, kittens, or pups, in a steady food supply.

Interestingly, researchers have observed squirrels using deceptive behavior when hiding food to throw off competitors such as other squirrels or birds. The squirrels have been observed digging a hole and pretending to bury a nut while keeping it in their cheek. Such behavior indicates a theory of mind that squirrels apply to their competition. In other words, squirrels anticipate the thinking and behavior of different animals, and squirrels modify their behavior in anticipation of danger to their cache. The nut and seed caching also helps forests to regenerate. If a squirrel forgets about a cache of nuts or the squirrel dies before it retrieves all of its hidden food, the seeds and nuts may germinate and further the spread of nut and seed-bearing plants such as walnut, hazelnut, and beech trees, among many others.

In the United Kingdom, the non-native grey squirrel received nuisance pest designation in 1940 after naturalists concluded it caused tree damage by eating the bark. The grey squirrel in the UK and Europe has been considered an invasive species. Additionally, the grey squirrel has devastated the native red squirrel population due to the grey squirrel out-competing the red squirrel for food and introducing squirrel pox. The grey squirrel has more immunity to squirrel pox than the red squirrel, leading to a growing population disparity.

Human intervention imported the grey squirrel to the UK, particularly the English nobleman Herbrand Arthur Russell, the 11th Duke of Bedford (1858 - 1940), who brought the grey squirrel from the United States as an ornamental species for his estate in 1876. (bbc.com) He also sent grey squirrels to various friends around England. Without substantial predators, the grey squirrel has spread widely across England and Wales since 1876, with the red squirrel only holding on in Scotland and Ireland. Since its designation as a nuisance, a number of grey squirrel eradication programs have unsuccessfully taken place in the UK, including in the 1950s free cartridges and six pence bounty per squirrel tail. Later, trapping and poisoning efforts failed to eradicate the gray squirrel from the UK.

The most recent effort to reduce the population of grey squirrels in the UK employs artificial intelligence and birth control. A new system called SquirrelAgent, developed by Genysys Engine, uses an AI system called deep learning that is trained on thousands of pictures of grey and red squirrels. The AI helps control a feeding system with two doors that differentiate the red and grey squirrels with 97% accuracy. (bbc.com) A camera in the system examines squirrels to determine the red ones from the grey ones. The system automatically opens a door to give food to red squirrels. Still, grey squirrels get access to a different door containing tasty gel containing an oral contraceptive. The system will favor red squirrels and impose strict family planning on the grey squirrels. This automated system hopes to save the red squirrel in England and Wales. AI makes it possible to provide strict differentiation between the two species, which would not be possible without automation. Such a non-lethal technique may benefit other ecosystems with invasive species competing with native ones.

Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high-level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

You can buy his book on Amazon in paperback and in kindle format here.