Geochemical signatures in fin rays provide a nonlethal method to distinguish the natal rearing streams of endangered juvenile Chinook Salmon Oncorhynchus tshawytscha in the Wenatchee River, Washington

Highlights

• Juvenile spring Chinook Salmon from three streams were analyzed for fin ray chemical composition.

• Element/calcium and isotopic ratios in fin rays were correlated with those in the water.

• Geochemical signatures in fin rays can distinguish juvenile Chinook Salmon from different streams with high accuracy.

Abstract

Rebuilding fish populations that have undergone a major decline is a challenging task that can be made more complicated when estimates of abundance obtained from physical tags are biased or imprecise. Abundance estimates based on natural tags where each fish in the population is marked can help address these problems, but generally requires that the samples be obtained in a nonlethal manner. We evaluated the potential of using geochemical signatures in fin rays as a nonlethal method to determine the natal tributaries of endangered juvenile spring Chinook Salmon in the Wenatchee River, Washington. Archived samples of anal fin clips collected from yearling smolt in 2009, 2010 and 2011 were analyzed for Ba/Ca, Mn/Ba, Mg/Ca, Sr/Ca, Zn/Ca and ⁸⁷Sr/⁸⁶Sr by inductively coupled plasma mass spectrometry. Water samples collected from these same streams in 2012 were also quantified for geochemical composition. Fin ray and water Ba/Ca, Sr/Ca, and ⁸⁷Sr/⁸⁶Sr were highly correlated despite the samples having been collected in different years. Fin ray Ba/Ca, Mg/Ca, Sr/Ca, Zn/Ca and ⁸⁷Sr/⁸⁶Sr ratios differed significantly among the natal streams, but also among years within streams. A linear discriminant model that included Ba/Ca, Mg/Ca, Sr/Ca, and ⁸⁷Sr/⁸⁶Sr correctly classified 95% of the salmon to their natal stream. Our results suggest that fin ray geochemistry may provide an effective, nonlethal method to identify mixtures of Wenatchee River spring Chinook Salmon for recovery efforts when these involve the capture of juvenile fish to estimate population abundance.

Introduction

Efforts to rebuild fish populations of conservation concern depend on reliable estimates of juvenile and adult abundance to evaluate recovery actions. These estimates are often based on mark-recapture methods that employ physical or natural tags to distinguish mixtures of both individuals and populations (Pine et al., 2003, Skalski et al., 2009, Drenner et al., 2012). Although physical tags such as coded wire, passive integrated transponder and fluorescent elastomer remain in wide use, variable retention in small fish and low probabilities of recapture in small populations have increased interest in natural tags where each individual in the population is uniquely or commonly marked (Skalski et al., 2009). Genetic tags, for example, are routinely employed for ecological monitoring because they can be used to obtain not only estimates of abundance and distribution, but also to assess demographic and evolutionary processes (Lukacs and Burnham, 2005, Schwartz et al., 2007, Luikart et al., 2010). For some applications, however, genetic markers may be of limited value, such as determining natal origins in population mixtures where the genetic structure is weak because of gene flow from natural straying or management actions (Miller et al., 2010, Barnett-Johnson et al., 2010, Garvin et al., 2013, Rundel et al., 2013, Brennan et al., 2015).

Geochemical signatures in fish can complement and in certain cases provide an alternative to genetic tags for studies of provenance because they are independent of the requirement for reproductive isolation and maintenance of population genetic structure (Feyrer et al., 2007, Miller et al., 2010, Barnett-Johnson et al., 2010, Rundel et al., 2013). These signatures also provide an additional scale of inferential power such as describing movement (Pracheil et al., 2014), and their identification in otoliths, bone, and scales has become a valuable tool for reconstructing environmental history (Campana, 1999, Elsdon et al., 2008, Kerr and Campana, 2014). Otoliths, which are composed of calcium carbonate in a protein matrix, have been preferred for such analyses because of their chemical stability and because of the high correlations of some elements with concentrations in the environment (Wells et al., 2003b, Elsdon and Gillanders, 2004, Gibson-Reinemer et al., 2009). Elements such as strontium (Sr) and barium (Ba) for which there is limited physiological regulation accrete in otoliths through branchial and digestive uptake where they provide a permanent record that reflects the water chemistry and dietary history of the fish (Kennedy et al., 2000, Buckel et al., 2004, Walther and Thorrold, 2006, Woodcock et al., 2012). A major limitation for the use of otoliths, however, is the fish must be sacrificed to obtain the sample.

Bone and scales also record environmental life history (Wells et al., 2003a, Muhlfeld et al., 2005, Clarke et al., 2007, Hanisch et al., 2010, Wolff et al., 2013), but they are also more labile because of their role as reservoirs in calcium metabolism, and their application in the study of provenance and movement of fish has been less extensive. Scales, for example, can undergo resorption to supply calcium for gonad formation during reproduction or during periods of dietary restriction (Armour et al., 1997, Persson et al., 1998). Scale geochemistry may also be subject to post-depositional change when fish such as salmon migrate from fresh water to seawater (Ramsay et al., 2011). In other species (e.g. weakfish, Cynoscion regalis), both post-depositional change and calcium mobilization during spawning contribute to the degradation in scale elemental signature (Wells et al., 2003a). By contrast, bone structures such as fin rays also function as reservoirs in calcium metabolism, but appear to be more inert than scales during periods of increased calcium demand (Mugiya and Watabe, 1977, Persson et al., 1997), which may account for the higher correlations in elemental concentrations observed between fin rays and otoliths than between scales and otoliths (Gillanders, 2001, Clarke et al., 2007, Wolff et al., 2013). Moreover, despite more labile chemistry, fin rays and scales offer certain analytical advantages when compared to otoliths. The calcium phosphate mineralogy of fin rays and scales differs from that of otoliths, leading to higher concentrations in some elements (Wells et al., 2000, Wells et al., 2003b, Campana and Thorrold, 2001, Holá et al., 2011), which can improve detection. Fin rays and scales also provide an alternative to otoliths for nonlethal sampling, which is critical when dealing with species of conservation concern. Fin clips, for example, have long been used for marking (Mears and Hatch, 1976) and age determination (Beamish, 1981, Sikstrom, 1983) and have recently found application in elemental and isotopic studies of provenance (Clarke et al., 2007, Wolff et al., 2013), movement (Balazik et al., 2012, Phelps et al., 2012) and trophic ecology (Sanderson et al., 2009, Andvik et al., 2010, Hanisch et al., 2010) in various threatened and endangered species.

Geochemical analysis of fin rays to determine the natal origins of Pacific salmon (Oncorhynchus spp.) listed under the Endangered Species Act (ESA) may have similar application because unbiased and precise estimates of abundance for these populations are often lacking (Rawding et al., 2014) The Wenatchee River in central Washington, for example, supports one of the three remaining populations of spring Chinook Salmon (Oncorhynchus tshawytscha) in the upper Columbia River listed under the ESA in 1999. Major spawning aggregations are found in the upper basin in the White River, Chiwawa River and Nason Creek tributaries (Upper Columbia Salmon Recovery Board, 2007, Murdoch et al., 2006, Murdoch et al., 2010), and although evidence indicates the historic population structure was probably altered by dam construction, some genetic distinction either persisted or developed among the tributaries subsequent to this activity (Utter et al., 1995, McClure et al., 2008). Recent (2006–2011) estimates of juvenile emigrants have ranged from N = 5879 ± 2667 (mean ± SD) for the White River to N = 97,680 ± 20,790 for the Chiwawa River (Hillman et al., 2014, Ishida et al., 2014). These estimates have been based on mark–recapture sampling of PIT tagged fish captured in rotary screw traps within the tributaries as they migrate downstream from late summer (i.e., sub-yearling fry and parr) to spring (i.e., yearling smolt). Flow conditions, however, differ widely between these time periods (Murdoch et al., 2006, Murdoch et al., 2010, Ishida et al., 2014), which has a highly variable effect on trap efficiency and tagging patterns, and thus the accuracy and precision of the abundance estimates (personal communication, P. Graf, Grant County Public Utility District). Resource managers have attempted to improve these estimates by trapping downstream migrants in the mainstem of the Wenatchee River to remove tributary dependent effects on trap efficiency, and by genotyping juvenile salmon from each tributary using microsatellite and single nucleotide polymorphisms. However, the utility of these genetic markers for determining stream origin has yet to be established.

As an alternative, we examined the potential of using geochemical signatures in the fin rays of juvenile spring Chinook Salmon as a nonlethal method to identify their natal tributaries. The objectives of the study were to determine (1) if water chemistry varied sufficiently among streams to produce distinguishing elemental or isotopic markers in the fin rays of juvenile salmon and (2) whether the markers remained stable across years to permit stream assignment from an established baseline rather than developing cohort specific signatures.