Geologic variability underlying stream catchment areas correlates with fish otolith microchemistry across disparate glacial till depths

Abstract

Abstract A primary application of otolith microchemistry is distinguishing natal origins of individual fish within a mixed-stock fishery. Stocks must be distinguishable according to stock-specific microchemistry patterns, with accurate stock assignment contingent upon microchemistry assessment of all sources contributing to the mixed-stock fishery. However, otolith microchemistry signatures of individual fish, upon which classification models are built, likely represent only a portion of the variability that exists for the stocks corresponding to each natal source. To statistically infer expected otolith microchemistry patterns among unsampled catchment areas proximal to sampled areas, we tested the hypothesis that variation in catchment geology among 35 stream sites across the Lake Michigan basin is correlated with the variation in otolith microchemistry signatures of age-0 steelhead collected at those sites. Matrices of Mahalanobis distances between all pairs of individual fish were calculated for each of the following: (1) assignment scores from discriminant function analysis of the variation among sites based on otolith microchemistry, and (2) the geology underlying the catchments upstream of each of the sites where fish were sampled. Based on Mantel tests, these matrices were found to be significantly correlated, indicating that age-0 steelhead that exhibit greater differences in otolith microchemistry signatures tended to come from sites exhibiting greater differences in catchment geology. Surficial geology alone was more correlated with otolith microchemistry than bedrock age, bedrock lithology, or any combinations of the three geological datasets. The significant relationship between geology and otolith microchemistry, although weak, supports tenuous hydrologic and geologic bases for delineating natal source geographic boundaries.