Generation of Quantitative Polymerase Chain Reaction Detectability Half-Lives and Comparison of Sampling Protocols for Genetic Diet Studies of San Francisco Estuary Fishes

Abstract

Several factors affect the probability of genetic analyses to detect prey in predator gut contents, including biological differences in the prey and predator species as well as differences in sampling and laboratory methodologies. Understanding these biases allows researchers to more appropriately put genetic prey detections in an ecological context. In this study, we determined the detectability half-lives of DNA from two prey species in the guts of two predators. The half-life detectability of juvenile Chinook Salmon Oncorhynchus tshawytscha in Striped Bass Morone saxatilis was 66.2 h, and that of larval Delta Smelt Hypomesus transpacificus in Mississippi Silverside Menidia audens was 26.4 h. Additionally, we performed a series of laboratory trials to examine the effects of variables in sample collection and preservation methodologies on the detectability of prey. Differences between methodologies were minimal, providing confidence that laboratory analyses will not be greatly affected by inconsistencies in field sampling procedures. Injecting a 95% solution of ethanol into the stomach via the esophagus immediately following collection and placing the fish on ice in the field prior to freezing at –20°C in the laboratory is a protocol readily applied in the field that will provide consistent results.

Received August 19, 2015; accepted December 3, 2015 Published online April 11, 2016

Acknowledgments

We would like to thank CABA at UC–Davis, the staff that helped with the feeding trials (Naoaki Ikemiyagi, Jared Frantzich, Lynn Takata, Haley Carlson, Jasmine Shen, Phil Choy, Shaun Rohrer, and Brett Harvey), the Freshwater Fish Co. for Striped Bass, as well as the California Department of Fish and Wildlife and the Feather River Hatchery for Chinook Salmon.