Development of a Simple Morphometric Model to Identify Sex in Chinook Salmon Returning to Spawn in the Yukon River

Abstract

Chinook Salmon Oncorhynchus tshawytscha support major fisheries in the Yukon River in Canada and the United States. Demographic data, including length, age, and sex composition of annual runs underpin management programs tasked with meeting escapement goals and providing for various fisheries within the drainage. Numerous sampling projects along the river collect these data annually, most without sacrificing fish. However, substantial error can occur when assigning sex based on observations of secondary sexual characteristics, particularly early in the spawning migration. We sought to develop a quantitative method of analyzing morphology of Chinook Salmon in the Yukon River that would achieve ≥95% accuracy of sex assignment with live fish. We collected snout lengths (SL) and mideye-to-fork lengths (MEF) from 2,342 known-sex fish from six regions of the drainage ranging from 40 to 2,000 km upstream from the sea, including a collection from postspawning fish. We fit three logistic regression models to the data using MEF, the ratio of SL to MEF, and regional group as covariates, including interactions and quadratic terms. Model selection was made using Akaike information criterion and cross-validation of prediction performance. Accuracy rates of the final quantitative model ranged from 79% for a lower Yukon River group to 92% for the postspawning group. Though short of our accuracy goal, the quantitative method may provide an improvement over existing methods of sex assignments in some regions of the drainage. Lower assignment rates in the lower river may reflect the mixed-stock composition of the fishery, where samples include fish with both short and long migrations ahead of them. However, we speculate that Chinook Salmon migrating to upper Yukon River spawning areas minimize the expression of secondary sexual characteristics, such as SL in males, and instead allocate their energy reserves to accomplishing the long migrations.

Validating morphometrics as a nonlethal tool to determine Arctic Grayling sex

“…2022) and carries through to other salmonids such that morphometrics could be used for noninvasive sex discrimination, understanding habitat selection, or determination of natal streams (Oke et al. 2019; Bradley and Brown 2021). One study conducted a review of sexual dimorphism among salmonid species (Englmaier et al.…”

“…Future research could replicate this study with larger sample sizes gathered from a broad range of locations across North America. Approaches similar to those discussed by Reed and McCann (1973) and Bradley and Brown (2021) could be used to compare patterns of sexual dimorphism and general morphology in the Arctic Grayling throughout its range.…”

“…Note that the male fish that were misclassified (bottom left) all had relatively small FLs, indicating that the model was reliable above a minimum FL threshold. (Baerum 2008;Englmaier et al 2022) and carries through to other salmonids such that morphometrics could be used for noninvasive sex discrimination, understanding habitat selection, or determination of natal streams (Oke et al 2019;Bradley and Brown 2021). One study conducted a review of sexual dimorphism among salmonid species (Englmaier et al 2022), providing a foundation for managers to identify important morphometrics for each species, which complements the methods proposed here.…”

“…However, these procedures can be costly, time‐ and skill‐intensive, and invasive to the fish. Alternatively, morphometrics have been used to nonlethally and noninvasively determine sex in species that exhibit sexual dimorphism (Merz and Merz 2004; Bradley and Brown 2021). Furthermore, morphometric data from digital images collected in the field are simple to add to ongoing studies, require little equipment, and are cost‐efficient compared to alternative methodologies (e.g., endoscopy and ultrasound).…”

“…Nonlethal determination of fish sex can be challenging, especially when working with small or vulnerable populations that cannot withstand invasive sampling techniques (e.g., Bradley and Brown 2021). There are various methods with which to nonlethally assign fish sex, including genetic analysis (Cavileer et al 2015), ultrasound techniques (Novelo and Tiersch 2012), and endoscopy (Swenson et al 2007).…”