Use of single nucleotide polymorphisms identifies backcrossing and species misidentifications among three San Francisco estuary osmerids

Abstract

Two threatened osmerid species native to the San Francisco Estuary (SFE)—Delta Smelt (Hypomesus transpacificus) and Longfin Smelt (Spirinchus thaleichthys)—are subject to broad human influence, including significant habitat alteration and the presence of the introduced osmerid, Wakasagi (Hypomesus nipponensis). The identification of these closely related species and their hybrids is difficult in field collected specimens which are subject to damage through handling and may be difficult to identify morphologically, especially when young. In addition, it is known that these three species hybridize, but the extent and effect of hybridization is difficult to quantify and monitor. We developed assays for 24 species-specific single nucleotide polymorphisms (SNPs) that identify whether a sample is a pure species (Delta Smelt, Longfin Smelt, or Wakasagi), a first generation (F1) hybrid, or a backcross. We used this SNP panel to genetically identify wild osmerids collected in Yolo Bypass from 2010 to 2016 and detected nine Delta Smelt × Wakasagi F1 hybrids and two Wakasagi × (Delta Smelt × Wakasagi) backcross hybrids; all assayed hybrids had Wakasagi as the maternal parent. The backcrossing into Wakasagi suggests that hybridization may only occur in one direction and thus preclude introgression to Delta Smelt. We also found substantial morphological field misidentifications (32.7%) in the Yolo Bypass samples resulting in more Wakasagi and fewer Delta Smelt than previously recorded when based on morphology. The SNP panel described in this study constitutes a valuable resource for monitoring hybridization in the SFE and assigning species identifications with accuracy and efficiency.