|
Abstract:
|
Investigating patterns in social groups is difficult if the information available on individuals is sparsely distributed in time and space. The Great Lakes Acoustic Telemetry Observation System detects fish with implanted transmitters when they swim through a network of receivers. The number of fish with transmitters is relatively small (e.g., in the hundreds) and the receivers are sparsely distributed throughout a large area (e.g., hundreds of km apart). We used randomization to look for evidence of lake trout travelling in social groups. We counted the cumulative number of distinct fish observed at increasing distances apart. The distances were randomly shuffled several times. Because the fish that travel in groups are observed repeatedly at shorter distances, the cumulative count of distinct fish at middle distances is lower than expected at random. This approach gave us valuable insight into the social behavior of lake trout that we were unable to gain using traditional social network analytical methods.
|
