Abstract
We investigated the spatial and temporal accuracy of otolith chemistry for distinguishing populations (and subpopulations) of anadromous river herring (Alewife Alosa pseudoharengus and Blueback Herring A. aestivalis) from New York watersheds. Water chemistry of selected analytes (Sr:Ca, Ba:Ca, 87:86Sr, and δ18O) differed spatially and interannually among sites within the Hudson River as well as between the Hudson River and Long Island estuaries. Otolith chemical markers for juvenile river herring differed significantly among all sampling sites, with combined elemental and isotopic ratios differing significantly between species and between years, although the effect of year did not influence assignments at broad spatial scales (i.e., 105 m). Adult river herring collected in the Hudson River were assigned to natal sites using all juvenile otolith markers except δ18O, which had methodological problems. Adult Alewives and Blueback Herring were classified to Hudson River populations at 73% and 42%, respectively. Inclusion of δ18O in the assignment model likely would have increased the percentage of adults assigned to the Hudson River, as juvenile reclassification accuracies were lower without δ18O and oxygen isotopes are strongly correlated with latitude. Otolith chemistry matched our geochemical atlas of water and distinguished among populations and subpopulations at a range of spatial scales. However, rates of reclassification are dependent on the markers included and the geographic distances among groups.
Received March 31, 2014; accepted July 16, 2014
ACKNOWLEDGMENTS
We are grateful to Molly Payne for laboratory assistance and to the New York State Department of Environmental Conservation's Hudson River Fisheries Unit, the New Jersey Department of Environmental Protection, and the Virginia Institute of Marine Sciences for fish collections. We thank Deb Driscoll (ESF), J. Blaustein (WHOI Plasma Facility), and the Stable Isotope Laboratories at the University of Michigan, University of Saskatchewan, and University of Arizona for assistance with sample analyses. Funding for the project was provided by the Hudson River Foundation and the National Fish and Wildlife Foundation. Thanks to D. Secor, E. Schultz, and an anonymous reviewer for comments on earlier drafts.