Social Networking Principles Help In Animal Research
A team of data analysts from Oxford
University may have taken a note out of Facebook founder Mark Zuckerberg's book
when they developed a social network for a population of great tits birds.
The bird species Parus major is a common social woodland songbird in Europe and Northern Africa, and is often observed by scientists interested in analyzing animal behavior. Mating pairs will usually raise between three and four hatchlings per season and the older female siblings may stick around to help with the new members of the family. These birds do not migrate and although pairs live in their own tree hollows, populations of great tits will take up residence throughout a forest area. The social structures of the birds can be monitored by observing how far they travel from their own tree hollow.
In a scientifically monitored forest called Wytham Woods, electronic recorders track when individual birds visit particular feeding stations set up throughout the forest. That information reveals which birds live in which areas, and along with the family ancestry of the birds and personal observations of the researchers, is used to interpret the relationships between the birds. However, the amount of electronic recordings can become overwhelming and meaningless without a way of analyzing it.
To solve this problem of too much random information, a group from the Pattern Analysis and Machine Learning Research Group at Oxford University developed a computer program to turn two years worth of electronic data points into a summary of meaningful facts. Rather than a list of separate occurrences "Bird A visits feeding station 1, 10am; Bird B visits feeding station 1, 10:05am; Bird C visits feeding station 2, 11am" the computer program summarizes the information to interpret that Bird A and Bird B interact often, whereas Bird C lives elsewhere in the forest and does not interact with the other two birds. All three birds are part of the same population, but they are not in the same social group. Conceptually, the program is similar to how social networking sites can be used to differentiate between people who are members of the site and people who are close friends by tracking how often those individuals interact online.
Greater understanding of the birds' social groups is valuable for researchers interested in deciphering the role of genetics in social interaction as well as the behavioral differences between familiar and unfamiliar birds. The future plan for the new computer program is to adapt the basic principles for other research opportunities.