Determination of River Herring Natal Origin Using Otolith Chemical Markers

Young of year river herring (i.e. alewife, Alosa pseudoharengus, and blueback herring, A. aestivalis) and water samples were collected from multiple locations in the Hudson River watershed and four Long Island estuaries in 2009 and 2010. We hypothesized that 1) otolith chemistries can be used to accurately assign fish to natal locations, and the spatial scale to which fish can be accurately assigned depends on the spatial scale at which local water chemistries vary, and 2) that interannual variation will affect natal assignment, but the effect will diminish at broader spatial scales. The markers used were Sr:Ca, Ba:Ca, ^87:86Sr, and δ¹⁸O. MANOVA with Wilks test found water and otolith chemistries differed significantly among sites (p < 0.0001), year had a significant effect (p < 0.03), and species differences were significant (p < 0.001). Linear discriminant analysis found water sample reclassifications had an apparent error rate of 0.265. The relationships between water and otolith elemental and isotopic ratios (for both species) were significant (excepting Sr:Ca for blueback herring). Linear discriminant analyses were performed for each species in each year; reclassifications were 100% correct for alewives in both years, and 89.6% and 95.8% correct for blueback herring in 2009 and 2010, respectively.