Friday, August 7, 2009

PS 77-23: Fatty acid signatures reveal spatial, temporal, and age-related differences in the diet of Caspian terns Hydroprogne caspia in the Columbia River Basin

Christina Maranto and Julia Parrish. University of Washington

Background/Question/Methods We studied the diet of Caspian terns Hydroprogne caspia in the Columbia River Basin using fatty acid (FA) analysis.  The objectives of the study were to determine whether there were (1) spatial, (2) temporal, and (3) age-related differences in diet based on FA signatures among and within three populations of Caspian terns: a coastal population at East Sand Island and two inland populations at Rock Island and the Potholes Reservoir.  Additionally, we compared gut content and FA signature analyses to determine whether they provided similar diet results.  We analyzed differences in FA composition between locations, time periods, and age-classes using Nonmetric Multidimensional Scaling (NMDS) and tested the statistical significance of differences among groups using Analysis of Group Similarities.  Comparisons between FA and gut contents were conducted using the Mantel test. 

Results/Conclusions

Results show that FA signatures were statistically different between spatial locations and temporal periods within all locations.  Caspian terns collected on East Sand Island were characterized by long chain monounsaturated FAs signifying they foraged on marine derived prey.  Inland populations at the Potholes Reservoir and Rock Island expressed predominantly polyunsaturated FAs, a freshwater signature.  The NMDS ordination of FA signatures between adults and chicks at the Potholes Reservoir suggests a difference in diet, although they are not statistically significant, most likely due to sample size.  There was a weak, but significant, correlation between the FA and gut content dissimilarity matrix, indicating that both techniques may provide similar results with ample sample sizes.  Gut contents are able to provide species level diet information but FA analysis is a powerful tool when sample sizes are limited.