Field Assessment of Colorado Pikeminnow Exposure to Mercury Within Its Designated Critical Habitat in Colorado, Utah, and New Mexico

Archives of Environmental Contamination and Toxicology https://doi.org/10.1007/s00244-018-0566-2

Field Assessment of Colorado pikeminnow Exposure to Mercury Within Its Designated Critical Habitat in Colorado, Utah, and New Mexico

Barbara C. Osmundson1,3 · Joel D. Lusk2

Received: 5 March 2018 / Accepted: 17 September 2018 © The Author(s) 2018

Abstract Mercury contamination in freshwater fsh is widespread across North America, including the western United States. Atmos- pheric mercury from both natural and manmade emissions deposits into watersheds and, through methylation and biomag- nifcation, accumulates in aquatic food webs. Highest mercury concentrations are found in predatory fsh. The endangered Colorado pikeminnow (Ptychocheilus lucius) is a long-lived, top-level piscivore endemic to the Colorado River basin. Mercury exposure to Colorado pikeminnow and another native fsh species, the roundtail chub (Gila robusta), was assessed by analyzing muscle tissues collected using a nonlethal technique. Mercury concentrations in Colorado pikeminnow > 400- mm long, captured from critical habitat throughout the species’ present range, exceeded the tissue threshold-efect level of 0.2 µg/g wet weight (WW) for whole body fsh (0.31 µg/g WW in muscle) recommended to protect fsh from injury. Mercury is a neurotoxin and endocrine disruptor, and impacts to fsh may include reduced ability to avoid predators, secure food, and reproduce. The highest mercury concentrations were found in both Colorado pikeminnow and roundtail chub collected from the White River, a tributary to the Green River. Colorado pikeminnow from the White and Green rivers had the highest mean mercury concentrations and the lowest mean relative body conditions. Exposure to high mercury concentrations may act in concert with other threatening factors to compromise Colorado pikeminnow population viability and eventual recovery.

Water quality and fsh surveys conducted during the past released from industrial and energy facilities and during 20 years revealed widespread mercury contamination, espe- mining/metal processing (Eagles-Smith et al. 2016). In cially in freshwater systems of the northern hemisphere western North America, atmospheric mercury from trans- (Schmidt and Brumbaugh 1990; Brumbaugh et al. 2001; pacifc transport (i.e., Asia) combines with local releases of Scudder et al. 2009; Cladis et al. 2014; Eagles-Smith et al. inorganic mercury (Pacyna et al. 2010; Pirrone, et al. 2010; 2014, 2015, 2016). Although mercury occurs naturally in UNEP 2013) and returns to the landscape by wet or dry the environment and can be emitted from volcanic activity, deposition. Mercury deposits enter waterways via watershed and reemitted from oceans and forest fres, anthropogenic runof, often ending up in lake bottoms and riverine wet- mercury emissions now far surpass those derived from natu- lands (Lindberg et al. 2007; Peterson et al. 2007; Eagles- ral processes (Mason and Sheu 2002; Pacyna et al. 2010; Smith et al. 2014, 2015, 2016). Driscoll et al. 2013; UNEP 2013). Inorganic mercury is After inorganic mercury enters aquatic ecosystems, conversion to methyl mercury is a key mechanism afecting bioaccumulation in the aquatic food web (Cocca * Barbara C. Osmundson [email protected] 2001; Engstrom 2007; Pasquale et al. 2009; Tsui et al. 2010; Driscoll et al. 2013; Eagles-Smith et al. 2016). * Joel D. Lusk [email protected] Fish accumulate methyl mercury mostly from their diet, and ingested mercury initially accumulates in fsh intes- 1 Colorado Ecological Services, Western Colorado Field tines but then is transferred to other tissues, including Ofce, US Fish and Wildlife Service, 445 West Gunnison blood, spleen, kidney, liver, and brain (Boudou et al. Ave., Suite 240, Grand Junction, CO 81501‑5711, USA 1991; Wiener and Spry 1996; Sandheinrich and Wiener 2 New Mexico Ecological Services, US Fish and Wildlife 2011). Skeletal muscle tissue is the primary ‘receiver’ of Service, 2105 Osuna Road NE, Albuquerque, NM 87113‑1001, USA redistributed methylmercury. Once there, mercury forms a complex with protein (Sandheinrich and Wiener 2011) 3 Present Address: 380 34 Road, Palisade, CO 81526, USA

Vol.:(0123456789)1 3 Archives of Environmental Contamination and Toxicology and is eliminated slowly, with an estimated half-life of location in southwestern Colorado as among the high- approximately 400 days (Gonzalez et al. 2005) to 2 years est in the United States. Walters et al. (2015) found high (Wiener and Spry 1996). mercury and selenium concentrations in fsh food webs Mercury in fsh muscle is predominately methylmercury, from the Colorado River in the Grand Canyon. Mercury so total mercury often is used as a surrogate measurement contamination has been found in several fsh species in the (Driscoll et al. 2013). Skeletal tissue concentrations vary upper Colorado River basin. Six of 23 roundtail chub (Gila with fsh species, location, feeding habitats, and age (Sand- robusta) collected in 1992 from the Gunnison and Colo- heinrich and Wiener 2011). With food web biomagnifca- rado rivers had concentrations > 0.3 mg/kg WW (Butler tion, fsh at higher trophic levels usually contain the greatest et al. 1994, 1995). A 2003 investigation revealed elevated mercury concentrations (Beckvar et al. 2005; Peterson et al. concentrations in some smallmouth bass (Micropterus 2007; Sandheinrich and Wiener 2011). dolomieu) (0.25 ± 0.03 µg/g WW) collected from Colo- Biotic and abiotic factors affect mercury toxicity in rado’s Yampa River and channel catfsh (Ictalurus punc- aquatic organisms. Environmental conditions (e.g., pH tatus) (0.21 ± 0.00 µg/g WW) from Utah’s Green River and temperature), sensitivities of individual species and (Hinck et al. 2006, 2007). Advisories for consumption life stages, and chemical and physical form of mercury, all of sport fsh species have been posted within Colorado afect toxicity (Wiener and Spry 1996). For several years, pikeminnow critical habitat in both Utah and Colorado. the scientifc and regulatory focus on mercury in aquatic Seven Colorado pikeminnow that died in captivity (held systems was motivated by the health risks to humans from as broodstock) 2–8 months after being taken from the consumption of mercury laden fsh (Wiener and Spry 1996; White and Colorado rivers in 1986 were frozen, and their USEPA 2001). However, several feld and laboratory studies whole bodies later analyzed for mercury. Elevated con- have demonstrated neurotoxic efects and impaired repro- centrations were found in the four from the White River duction occur in the fsh themselves at relevant dietary expo- (0.31–0.96 µg/g WW) and in the three from the Colorado sures similar to those found in the environment (Crump and River (0.28–0.52 µg/g WW) (Krueger 1988). Trudeau 2009). Some investigators have suggested that mercury toxic- The Colorado pikeminnow (Ptychocheilus lucius) evolved ity in fsh might be counteracted by selenium, especially as the Colorado River Basin’s top predatory fsh. As with when Se:Hg molar ratios exceed 1 (Peterson et al. 2009; all long-lived piscivores, Colorado pikeminnow are at risk Penglase et al. 2014). These researchers suggest that pres- of accumulating high mercury concentrations. Before the ence of selenium in fsh tissue must be considered when 1850s, they were abundant throughout warm-water reaches assessing potential toxicity of mercury concentrations. of the Colorado River Basin (Seethaler 1978; Platania 1990). Contaminant analyses in fsh tissue integrates the route, By the 1970s, all lower basin populations (downstream of duration, and magnitude of exposure, as well as chemical Glen Canyon Dam) and some upper basin populations were form, metabolic transformations, and modifying biotic and extirpated due to environmental alterations, including exten- abiotic factors. Mercury in fsh tissue can be accurately sive dam-building (Miller 1961). The species was federally assessed by using muscle biopsies (Pearson 2000; Baker listed as endangered in 1967 (USFWS 1967; Miller 1961; et al. 2004; Peterson et al. 2005). To do so nonlethally Moyle 1976; Tyus 1991; Osmundson and Burnham 1998). (i.e., for endangered species), a dermal muscle punch can Habitat considered critical to the survival and recovery be used to sample tissue for accurate assessment of mer- of Colorado pikeminnow was later designated in portions cury residues. To determine if mercury exposure could of Colorado, Utah, New Mexico, Arizona, and California be a factor contributing to the decline of this species, or (USFWS 1994) (Fig. 1). perhaps hampering recovery eforts, we assessed mercury Numerous sources of mercury emissions operate either concentrations in muscle tissue of Colorado pikemin- adjacent to or upwind of critical habitat in Colorado, Utah, now sampled from throughout their remaining range. We New Mexico, and Nevada (United States Environmental Pro- also assessed mercury concentrations in roundtail chub tection Agency [USEPA] 2017). These sources of mercury, sampled in the White River. Our objectives were: (1) to listed in USEPA’s (2017) TRI Explorer database, include: determine whether concentrations exceeded established coal-fred power plants, copper smelters, gold ore process- guidelines for fsh health; (2) to determine whether mean ing with autoclave roasters, oil refneries, and cement and concentrations varied by fsh size and by river; (3) to asphalt plants. These local emissions contribute several hun- determine Se:Hg ratios; and (4) to assess whether body dred pounds of atmospheric mercury annually in addition weight (an index of fsh health) is afected by mercury to that contributed by trans-Pacifc sources (Seigneur et al. concentrations. 2004). Weidner (2007) and Sather et al. (2013) described mercury concentrations in precipitation at a monitoring

1 3 Archives of Environmental Contamination and Toxicology

Fig. 1 Distribution and critical habitat of the Colorado pikeminnow in the Colorado River system (USFWS 2014a, b)

Materials and Methods critical habitat and represented much of the species’ remaining range. Study Area Muscle Tissue Collection Muscle tissues were collected from wild Colorado pikeminnow captured from 330 river kilometers (rkm) of the Green Muscle plugs were collected from 41 Colorado pikemin- River, 122 rkm of the Colorado River, 164 rkm of the White now during late spring 2008 and 57 during 2009 (Table 1). River, and 34 rkm of the Yampa River (Fig. 1; Table 1). In 2008, 29 samples were from upper, middle, and lower Some also were collected from individuals in the lowermost Green River reaches, 8 from the Yampa River, and 4 from 3.7 rkm of the Gunnison River (downstream of Redlands the White River. In 2009, 23 samples were collected from Diversion Dam), comprising part of the Colorado River the lower and upper Colorado River reaches, 29 from the population (Osmundson et al. 1998). Captures of stocked, San Juan River, and 6 from the White River. Muscle plugs hatchery-reared, Colorado pikeminnow were from 64 rkm also were taken from 11 roundtail chubs captured from the of the San Juan River. All reaches sampled were within White River in eastern Utah during 2009. Muscle tissue

1 3 Archives of Environmental Contamination and Toxicology

Table 1 River, river segment, river kilometer locations, segment description, and Universal Transverse Mercator (UTM) coordinates of river segment where Colorado pikeminnow were captured in 2008 and 2009 River River segment and (# River kilometer locations Segment description UTMs (Zone 12; meters East (E) and of fsh captured) (converted from river North (N) miles)

Colorado (23) Lower Colorado (12) 68.1–170.3 Below Potash to Fish Ford launch 615674 E 4255628 N–652223.5 E 4309671 N Upper Colorado (11) 275–295 Gunnison River confuence to Labor 710178.1 E 4325780 N–726547 E Camp 4330627 N San Juan (29) 170.7–235.5 Aneth to Shiprock 665398 E 4114487 N–703649 E 4074289 N White (10) 1.5–165 Close to Green River confuence to 636766 E 4433366 N–667791 E Below Taylor Draw Dam 4430780 N Yampa (8) 131.4–166.5 Upper Maybell to Morgan Gulch 748673 E 4489051 N–Zone 13 256857 E 4473396 N Green (29) Lower Green (10) 86.6–178.9 Green River, Utah 586913 E 4267030 N–584974 E 4273276 N Middle Green (9) 214.9–383.7 Close to Price River confuence 581269 E 4326476 N–631382 E Below Ouray 4429511 N Upper Green (10) 468.5–537.6 Bonanza Bridge to Island Park 631382 E 4463388 N–656798 E 4485121 N

collection followed procedures of Williamson (1992), Baker the protection of the health of fsh themselves from mer- et al. (2004) and Peterson et al. (2005) with some modifca- cury, various authors have suggested diferent tissue toxicity tions. Fishery crew members from federal and state agen- threshold-efect levels. In a review of mercury toxicity to cies, and Colorado State University, were frst instructed to freshwater fsh, Sandheinrich and Wiener (2011) reported follow standardized procedures for safe handling and length fsh-tissue mercury concentrations starting at 0.3 µg/g WW measuring of fsh and collecting and preserving muscle tis- in whole body fsh and 0.5 µg/g WW in fsh axial muscle sue. A 5-mm dermal punch was inserted 1–2 cm below the were associated with efects on biochemical processes, dam- dorsal fn, and a plug of muscle tissue was extracted with age to cells and tissues, and reduced reproduction. After a slight twisting motion. Breaking of of the tissue sample reviewing 10 mercury residue-efect studies for 8 fsh spe- was facilitated by tilting the punch during removal. A dif- cies, and using survival, growth, reproduction, and behavior ferent punch was used on each fsh and discarded after use. as toxicity endpoints, Beckvar et al. (2005) recommended A single muscle plug was taken from each fsh and placed a 0.2 µg/g mercury whole body guideline to protect juve- in an acid-washed cryogenic vial. These were kept on wet nile and adult fsh health. Dillon et al. (2010) developed ice until crews returned from the feld. Betadine was applied a mercury dose–response curve based on published fsh to the fshes’ wound to decrease the risk of infection and tissue-residue toxicity studies, in which endpoints related promote healing. Samples were frozen upon return from the to mortality (i.e., survival and reproductive success) were feld. After all samples were collected, muscle plug skins selected. They found residues of 0.1 µg/g WW associated were dissected and removed using acid-washed, stainless with a 2.8% injury rate (e.g., reduced survival) in adult fsh steel surgical equipment. Tissues were then returned to con- and a 19.8% injury rate in early life stages. Higher mercury tainers and frozen to − 20 °C. Samples were inventoried and concentrations were associated with higher levels of injury. shipped to the Trace Element Research Laboratory (TERL) Because Colorado pikeminnow are endangered, we chose in College Station, Texas, for mercury analysis using com- the more protective mercury toxicity threshold of 0.2 µg/g bustion atomic absorption spectrometry (CAAS) (USEPA WW in whole body fsh (Beckvar et al. 2005) as a standard 1998; Cizdziel et al. 2002). with which to compare our results. Regression equations developed for various fsh species Mercury Toxicity Guidelines have been used to convert mercury concentrations in skin- less muscle tissue to whole body concentrations (Peterson In 2001, the U.S. Environmental Protection Agency devel- et al. 2005). One equation developed for Northern pikemin- oped a methyl mercury water quality criterion of 0.3 mg/kg now (Ptychocheilus oregonensis), a physiologically similar wet weight (WW) in edible fsh tissue, designed for the pro- species to Colorado pikeminnow, had a slope of 0.9048 and tection of humans consuming fsh (USEPA 2001). Regarding intercept of −0.2387. We used this equation to convert the

1 3 Archives of Environmental Contamination and Toxicology recommended whole body toxicity threshold of 0.2 µg/g verted guideline for muscle tissue of 0.31 µg/g WW per WW (Beckvar et al. 2005) to a corresponding muscle tissue Beckvar et al. (2005). toxicity threshold of 0.31 µg/g WW. Results Statistical Analysis and Data Interpretation Mercury Concentrations 1. Colorado pikeminnow mercury concentrations were transformed (Log10(X + 1)) to improve normality and The mean mercury concentration in muscle plugs from 99 t tests (p < 0.05) were then used to compare mercury Colorado pikeminnow from combined river segments was concentrations among rivers 0.49 µg/g WW (95% confdence interval [CI] = 0.42–0.56), 2. Relations between body size and mercury concentration exceeding the recommended fsh muscle toxicity guideline were investigated with regression, using transformed for fsh health (0.31 µg/g WW). The mean concentration (Log 10) fsh total lengths (independent variable) and for 83 Colorado pikeminnow > 400-mm long was 0.66 µg/g transformed (Log10(X + 1)) muscle plug mercury con- WW (95% CI = 0.6–0.72) or twice that of the toxicity guide- centrations (dependent variable). line. For the 10 Colorado pikeminnow sampled from the 3. We regressed transformed (Log 10) Colorado pikemin- White River, the mean concentration (1.1 µg/g WW (95% now total length and mass measurements and used the CI = 0.79–1.31) was over three times that of the toxicity slope and y-intercept to calculate relative body condi- guideline (Table 2). This mean concentration in White River tion. samples was signifcantly higher than those means from the 4. Relative condition (Kn) of Colorado pikeminnow was Colorado and Yampa river segments and likely would have calculated using the following equations from Osmund- been signifcantly higher than that in Green River Colorado son and Burnham (1998): pikeminnow (α = 0.06) had the sample size been larger. For K M M the 11 White River roundtail chubs, the mean concentration n = 100 × o∕ e (1) (0.59 µg/g/WW [95% CI = 0.42–0.60]) also exceeded the M M where o is the observed mass (g) and e is the toxicity guideline. expected mass (g) as calculated from: For samples from combined Green River reaches, M m b log10 e = log10 (length) ∗ + (2) the mean concentration was 0.72 µg/g WW (95% CI = 0.66–0.78), for samples from combined Colorado River M was calculated from transformed lengths and e reaches, 0.6 µg/g WW (95% CI = 0.53–0.67), and from the weights of Colorado pikeminnow sampled for muscle Yampa River, 0.48 µg/g WW (95% CI = 0.43–0.53) (Table 2; tissues from all river reaches, with (m) as the slope and Fig. 2). Many of the stocked Colorado pikeminnow in the (b) as the y-intercept (developed under #3 above). San Juan River were < 400-mm long, and those that were 5. Mean K of Colorado pikeminnow was compared among n larger had concentrations exceeding the toxicity guideline. rivers and relative conditions of individual fsh (depend- Aside from means, mercury concentrations exceeded tox- ent variable) were regressed against the corresponding icity thresholds in most individual Colorado pikeminnow mercury concentrations found for those fsh (independ- sampled, particularly those from the White River. ent variable) to assess whether mercury burdens might afect fsh body condition. Fish Length and Mercury Concentration 6. Mass concentrations of selenium and mercury were converted to molar concentrations by dividing by their To determine if mercury concentrations difered among molecular weights (Se = 78.96; Hg = 200.61; Peterson river segments because of size diferences in sampled et al. 2009). We compared molar concentrations of fsh, we compared mean total lengths among fsh from Se:Hg to estimate when and where any ‘surplus’ sele- each river segment (Table 2; Fig. 2). Colorado pikemin- nium was potentially available for protection against now from the Yampa River were some of the largest fsh mercury toxicity. Mean selenium concentrations for captured but had lower mercury concentrations than those Colorado pikeminnow sampled during 1996 in the same from other rivers. Mercury concentrations were high- river segments as this study (listed by Hamilton et al. est in White River Colorado pikeminnow, but these fsh 2003) were used for comparisons with our more recently were not signifcantly larger than those captured in the observed mercury concentrations. Colorado and Yampa rivers. Those White River Colorado 7. Mercury concentrations in Colorado pikeminnow mus- pikeminnow with the highest overall mercury concentra- cle plugs were compared to those of other fsh species tions (> 1.0 µg/g WW) were not the largest fsh captured reported to have adverse biological efects (i.e., con- during the study. Hence, there appeared to be a river efect

1 3 Archives of Environmental Contamination and Toxicology

Table 2 Summary statistics for mercury concentrations (wet weight) pikeminnow. The geometric mean for fsh total length for each river in muscle plugs taken from Colorado pikeminnow (CPM) and round- segment also is provided tail chub (RTC) in river segments in critical habitats for Colorado

Matrix Statistic Green River Yampa River White River Colorado San Juan River River Colorado Roundtail < 400 mm > 400 mm pikeminnow chub total length total length

Mercury Median 0.72 0.49 1.02 0.52 0.56 0.09 0.38 Geometric 0.7 (0.66– 0.48 (0.45– 1.01 (0.88– 0.57 0.57 (0.51– 0.11 (0.09– 0.37 (0.32– mean (95% 0.74) 0.51) 1.15) (0.3–0.89) 0.61) 0.13) 0.42) CI’s) Minimum 0.32 0.39 0.43 0.11 0.31 0.03 0.31 Maximum 1.08 0.58 1.8 1.97 1.04 0.28 0.43 n 29 8 10 11 23 26 3 Total length Geometric 501 (479– 589 (552– 562 (513– 282 (246– 599 (549– 282 (269– 427 (418–442) mean (95% 525) 604) 617) 324) 649) 295) CIs) Body condi- Mean (95% 96 (90–102) 117 (108– 88 (82–94) 105 (98–111) tion CIs) 126) CPM over 28/29 8/8 10/10 6/11 23/23 0/26 2/3 400 and RTC over 250 mm Total length that exceed toxicity GL

Fig. 2 Mean (+ 95% CI) total 1.4 700 Total mercury mercury and mean fsh total 1.3 length for Colorado pikemin- Fish length now from each river segment in 1.2 600 critical habitat. Rivers include: 1.1

WR GR Mean total length (mm) White River, Green 1.0 500 River, CR Colorado River, YR Yampa River, SJR-L San 0.9 Juan River fsh > 400 mm total 0.8 400 SJR S length, - San Juan River 0.7 fsh < 400-mm total length. Horizontal line = toxicity 0.6 300 guideline (TG) of 0.31 µg/g 0.5 WW 0.4 200 TG Mean total mercury (ug/g WW) 0.3 0.2 100 0.1 0.0 0 WR GR CR YR SJR-L SJR-S River reach

separate from any length efect. However, there was a sig- (r2 = 0.8, p < 0.001) between mercury concentration and nifcant correlation (r2 = 0.65, p = 0.0000) between mer- roundtail chub length. cury concentration and total length when samples from all river segments were pooled (Fig. 3). Body Condition and Mercury Concentration Of roundtail chubs from the White River, the largest individuals (almost 400-mm long and > 400 g) contained Because most Colorado pikeminnow from the San Juan the highest mercury concentration (almost 2.0 µg/g WW). River were relatively small and recently stocked before In addition, there was a high and signifcant correlation capture, they were not included in body condition

1 3 Archives of Environmental Contamination and Toxicology

0.5 r2 r r = 0.74 trend ( = 0.26) and inverse correlation ( = − 0.51) 2 r = 0.55 between mercury concentration and relative body condi- 0.4 p < 0.0001 tion (Fig. 5). n = 99

0.3 Selenium–Mercury Interaction (Total Hg+1) Colorado pikeminnow from all river segments had higher 10 0.2 mass and molar selenium concentrations than mercury con- Log

0.1 centrations (Table 3), based on mean concentrations from two separate sampling periods. Surplus selenium, which may aford protection against mercury in Colorado pikemin- 0.0 2.42.6 2.83.0 now, was highest in the Colorado River and lowest in the Yampa and White rivers (Table 3). Log10 TL (mm)

Fig. 3 Log-transformed total mercury concentration, Discussion Log10([T-Hg] + 1), versus log-transformed total length, Log 10(TL) for all sampled Colorado pikeminnow from combined river segments’ Diference in Mean Mercury Concentrations Among Rivers analyses. Length and weight of Colorado pikeminnow from the White, Green, Yampa, and Colorado rivers Elevated mercury concentrations reported here verify not were strongly correlated variables (r2 = 0.94, p < 0.0001, only that mercury deposition occurs in watersheds contain- n = 68), as expected. Length–weight regression coef- ing Colorado pikeminnow critical habitat but also that mer- cients, used in calculating relative body condition, had cury has moved through the aquatic food web and into the a slope (m) of 3.19111 (SE = 0.104) and y-intercept (b) tissues of roundtail chub and endangered Colorado pikemin- of −5.587 (SE = 0.154). Colorado pikeminnow from the now. Sources are not defnitively known, but deposition White and Green rivers had the highest mean mercury likely includes both local and global contributions. Follow- concentrations and the lowest mean relative body condi- ing deposition, conditions favoring methylation of inorganic tions (Fig. 4), whereas those from the Yampa and upper mercury enable bioaccumulation. Our results suggest that Colorado rivers had lower mean mercury concentrations remaining populations of Colorado pikeminnow are at risk and higher mean relative body conditions. When Colorado of sufering the adverse biological efects associated with pikeminnow were pooled (n = 63) from all sites (excluding mercury exposure. the San Juan River), there was a signifcant (p < 0.0001)

Fig. 4 Mean (+ 95% CI) total 1.5 Total mercury 140 mercury and relative body 1.4 Body condition condition (Kn) for Colorado 1.3 pikeminnow sampled from the Mean relative body condition White, Green, Colorado, and 1.2 Yampa rivers. River reaches 1.1 120 include: WR White River, GR 1.0 CR Green River, Colorado 0.9 River, YR Yampa River. Hori- 0.8 zontal line = toxicity guideline 100 of 0.31 µg/g WW 0.7 0.6 0.5

0.4 80 ( K Mean total mercury (ug/g WW) 0.3 n 0.2 ) 0.1 0.0 60 WR GR CR YR

River reach

1 3 Archives of Environmental Contamination and Toxicology

150 r = -0.51 Yampa and White River basins are needed to help explain

2 ) r = 0.26 the diferences in mercury burdens noted here. n

( K p < 0.0001 n = 68 Fish Length and Mercury Concentration 100

Mercury in fsh tissue may increase most rapidly when fsh are small and growing quickly, particularly when piscivorous 50 species shift from a diet of invertebrates to one of other fsh. Mercury concentrations in fsh tissue generally rise with

Relative body condition increasing fsh age or body size because of the slow rate of elimination relative to the faster rate of uptake (Sandhein- 0 0.00.1 0.20.3 0.40.5 rich and Wiener 2011; Wiener and Spry 1996). Colorado pikeminnows switch from an insectivorous to a piscivo- Log (Total Hg+1) 10 rous diet when they exceed 40- to 100-mm long (Vanicek and Kramer 1969; Muth and Snyder 1995). Diet studies of Fig. 5 Relative body condition versus log-transformed total mercury Colorado pikeminnow in the Colorado River indicated that concentrations, Log10 ([T-Hg] +1), for Colorado pikeminnow sampled from the White, Green, Colorado, and Yampa rivers individuals 400- to 550-mm long had primarily eaten other fsh species, including roundtail chubs, and larger Colorado pikeminnow ate larger prey fsh (Osmundson et al. 1998). The Yampa and White rivers in northwestern Colorado As results from this study demonstrate, mercury muscle con- are relatively close geographically and run roughly parallel centrations increase with Colorado pikeminnow size. Large to one another. Because of this, aerial (wet and dry) depo- individuals are older and thus have accumulated mercury for sition rates into each river was expected to be similar. Yet a longer period. Also, because they eat larger prey than do Colorado pikeminnow from the Yampa River had the lowest smaller Colorado pikeminnow, larger prey fsh themselves concentrations of the study (excluding recently stocked San may have higher mercury burdens. For White River round- Juan River fsh), whereas those in the White River had the tail chub, there also was a positive and signifcant corre- highest concentrations. This suggests a localized source(s) lation between mercury concentration and fsh length, and of mercury entering the White River. Landscape modifca- the largest roundtail chubs contained concentrations as high tions, including land disturbance, can alter mercury inputs as those found in Colorado pikeminnow. Other factors, as to downstream aquatic ecosystems (Evers et al. 2007). Other yet unknown, also infuence tissue concentrations and are localized sources might include mining or other geochemi- evidently specifc to the river segment in which each fsh cal, biological, or watershed factors afecting levels of mer- primarily resides. cury bioaccumulation. Landscape characteristics infuenc- ing mercury transport to surface waters include land cover, Fish Body Condition and Mercury Concentration oxidation–reduction conditions, hydrologic fow paths, large water-level manipulations in reservoirs, and nutrient load- Body condition in fsh often is used to gauge the overall ing (Evers et al. 2007). Further detailed comparisons of the health of an individual fsh. Relative condition factor is

Table 3 Mass and molar Mean mercury concentration Mean selenium concentration Surplus Se concentrations of mercury (minimum, maximum) (minimum, maximum) µmol/g Se (Hg) and selenium (Se) and surplus selenium concentrations µg/g Hg WW µmol/g Hg WW µg/g Se WW µmol/g Se WW in Colorado pikeminnow (selenium concentrations are River Segment from Hamilton et al. 2003) Mid Green R. 0.77 0.0038 0.98 0.0124 0.0086 (0.68,0.87) (0.87, 1.08) Yampa R. 0.49 0.0024 0.62 0.0079 0.0055 (0.44, 0.53) (0.44, 0.72) White R. 0.95 0.0052 0.93 0.0118 0.0071 (0.43, 1.83) (0.64, 1.18) San Juan R. (> 400 mm) 0.236 0.0007 0.83 0.0105 0.0098 (0.199, 0.267) (0.74, 1.0) Colorado R. 0.6 0.003 1.92 0.0243 0.0214 (0.31, 1.04) (0.93, 2.16)

1 3 Archives of Environmental Contamination and Toxicology an index that relates an individual’s plumpness to that of environmental stressors) have not yet been determined for the average fsh in the population or in many populations Colorado pikeminnow. combined. Thus, it indicates the degree of diference in the In 468 fsh, representing 40 western U.S. species, 97.5% observed weight of a fsh compared with the species-spe- had a molar ratio > 1 (i.e., containing more selenium than cifc expected weight for a fsh of its length (Le Cren 1951). mercury) (Peterson et al. 2009). Fish species with Se:Hg < 1 Reduced body condition associated with mercury exposure included other pikeminnow species. Colorado pikeminnow has been found in various fsh species, including striped from the White River had the highest mercury concentra- bass (Morone saxatilis), Northern pike (Esox lucius), white tions observed during this study and therefore are at greatest sturgeon (Acipenser transmontanus), and walleye (Stizoste- risk of injury from mercury exposure. The role that sele- dion vitreum) (Sandheinrich and Wiener 2011). This inverse nium may play in counteracting these efects is currently relationship between mercury concentration and body con- unknown. dition, however, can be confounded and complicated by other co-occurring contaminants or by other environmental Mercury Levels in Colorado pikeminnow and Those factors. Osmundson et al. (1998) reported that mean Kn of in Other Species Demonstrating Toxicity Colorado pikeminnow in the Colorado River signifcantly decreased with increased fsh size in the lower river seg- More than half of the Colorado pikeminnow sampled for this ment but increased with fsh size in the upper river segment. study exceeded the 0.2 µg/g WW whole body fsh toxicity In addition, mean Kn within several 100-mm length classes threshold (0.31 µg/g WW in fsh muscle) proposed by Beck- significantly differed among four 3-year study periods var et al. (2005). A residue-based mercury dose–response (Osmundson and White 2014). Those investigators attrib- model for fsh (Dillon et al. 2010) associates a mercury uted spatial and temporal changes in mean Kn to variation residue of 0.3 µg/g WW whole body (= 0.45 µg/g WW in in food availability. Our results indicate mercury exposure is muscle) with an 8% injury rate for juvenile/adult fsh and a also afecting Colorado pikeminnow body condition. Mean 42.5% injury rate for early life stages. Comparison of mer- Kn was lowest in Colorado pikeminnow from the White and cury levels in Colorado pikeminnow to this dose–response Green rivers, where mercury residues were highest. In addi- curve suggests that many individuals probably experience tion, we found a signifcant inverse correlation between body some rate of impaired reproduction, reduced growth, and condition and mercury concentration when individuals from reduced survival. four rivers were pooled. The r2 value (0.26, p < 0.0001) was A population viability analysis recently modeled for not high, but this is expected given that food availability, Colorado pikeminnow in the San Juan River estimated that water temperature, and other environmental variables also current levels of mercury toxicity would reduce reproductive afect condition. Most importantly, reduced body condition success by 2% among newly recruited adult females (Miller associated with elevated mercury residues points to mercury 2014). As these females age, the percent injury was expected negatively impacting the health of these fsh. to increase to 5%. If mercury deposition in the San Juan River increases in the future as anticipated, injury estimates Selenium–Mercury Interaction are predicted to increase to 3.5–9%. Under this assumption, the estimated injuries to both reproductive success and age- Accumulations of selenium and mercury individually can specifc survival led to decreases in simulated population result in irreversible injury during early development of fsh growth potential (Miller 2014). Thus, anticipated mercury (Mailman et al. 2014; Walters et al. 2015). Some researchers load increases in the San Juan sub-basin are expected to have reported evidence of selenium moderating the nega- reduce the efectiveness of current recovery eforts. tive efects of mercury in freshwater fsh when accumulated Sublethal efects can be important in fsh. Crump and together (Peterson et al. 2009), but underlying mechanisms Trudeau (2009) reported that mercury accumulation in the and the amount of excess selenium needed are unclear (Khan fsh brain resulted in reduced hormone secretion, nerve dam- and Wang 2009). Kim et al. (1977) and Cuvin and Furness age, and alterations in neurotransmission. As a neurotoxin, (1988) demonstrated both antagonistic and synergistic toxic mercury can alter behavior, reducing predator avoidance interactions between selenium and mercury are possible and ability and/or the ability to secure food, leading to slow are a function of concentrations and the molar ratio of one growth and emaciation (Sandheinrich and Atchison 1990; to the other. More recently, low-level addition of selenium Webber and Haines 2003; Crump and Trudeau 2009; Sand- to a lake was found to decrease MeHg bioaccumulation in heinrich and Wiener 2011; Depew et al. 2012). Lowered fsh gonads (Mailman et al. 2014). The optimal antago- body condition, as noted, would be an expected result of nistic molar ratios for selenium and mercury in the envi- this. Mercury also can negatively afect fsh endocrine sys- ronment or in tissues (along with other contaminants and tems, resulting in diminished reproductive success (Matta et al. 2001; Hammerschmidt et al. 2002; Drevnick and

1 3 Archives of Environmental Contamination and Toxicology

Sandheinrich 2003; Tan et al. 2009; Sandheinrich and Wie- wild-produced larvae to the adult phase has not yet been ner 2011; Richter et al. 2014). Crump and Trudeau (2009) documented (Durst and Franssen 2014). suggested that the inhibitory efect of mercury on reproduc- A recent decline in wild adults has been reported for the tion in fsh may occur at multiple sites within the repro- Green River basin (Bestgen et al. 2016a, b) as well as the ductive system, including the hypothalamus and pituitary upper Colorado River (Osmundson and White 2017; Elverud in the brain, in the gonads themselves, and in other tissues and White 2017). Especially concerning is the very weak and functions of the endocrine system. Investigations into age-0 representation in the middle Green River from 1999 the efects on reproductive organs demonstrated a range of to 2013 (Bestgen and Hill 2015; Bestgen et al. 2016a). A injuries, including reduced gonad size and gamete produc- similar decline in young of the year fsh (YOY) in the Colo- tion, reduced circulating reproductive steroids, and lowered rado River from 1997 to 2013 led researchers to conclude spawning success (Crump and Trudeau 2009; Tan et al. that recruitment rates have been insufcient to ofset adult 2009; Sandheinrich and Miller 2006; Sandheinrich and mortality rates (Osmundson and White 2017). Wiener 2011). Water regulation is thought to negatively afect reproduc- Higher trophic level fsh can accumulate substantial tis- tive success and habitat suitability for young (Bestgen and sue mercury, and females can transfer it to developing eggs Hill 2015; Osmundson and White 2017), while persistently during oogenesis (Hammerschmidt and Sandheinrich 2005; high densities of nonnative predators (e.g., smallmouth bass Alvarez et al. 2006; Crump and Trudeau 2009). In addition (Micropterus dolomieu), northern pike, and walleye), par- to maternal transfer, eggs released into the aquatic envi- ticularly in the Yampa River (Johnson et al. 2008; USFWS ronment may accumulate mercury directly from the water 2014a, b, 2015), reduce survival of juveniles. In combina- (Crump and Trudeau 2009). Beckvar et al. (2005) and Dil- tion, rates of recruitment to the adult stage are depressed. lon et al. (2010) suggested that early life stage fsh have a Predation by nonnative fsh on young Colorado pikeminnow much greater sensitivity to mercury compared with juve- is an obvious, direct cause of mortality, while river regula- nile and adult fsh. Dillon et al. (2010) calculated a median tion and the pathways by which associated impacts afect efect concentration (MC50) for mercury in larval fsh to be reproduction and recruitment are more complex and there- 0.406 µg/g WW (95% CI = 0.117–1.405 µg/g WW). Behav- fore less well understood. Layered on this are the efects of ioral impairment after developmental mercury exposure has environmental contaminants. Our results, along with those been demonstrated in laboratory studies (Fjeld et al. 1998; from studies of other fsh, suggest mercury burdens are suf- Alvarez et al. 2006; Weber 2006) at mercury concentrations fciently high in Colorado pikeminnow that negative efects as low as 0.007–0.015 µg/g WW (Weber 2006). Behavioral at the population level should be expected. impairments include reduced competitive feeding ability and predator-evasion responses, which in turn may reduce survival and recruitment to endangered fsh populations. Conclusions Crump and Trudeau (2009) also discussed a mercury- altered lipid balance resulting in disruption of vitellogenesis High mercury concentrations are known to adversely afect and egg production. Richter et al. (2011) found individual reproductive output and adult survival in fshes. For Colo- genes (related to nervous system development and lipid rado pikeminnow, the high concentrations documented here metabolism) in female zebrafsh (Danio rerio) exhibiting may act multiplicatively with other threats to reduce popu- altered expression in response to methylmercury expo- lation growth rate and ultimately impact recovery poten- sure. Similar tissue-level studies are needed for Colorado tial. Mercury exposure was found in all sampled Colorado pikeminnow to assess degree of injury in individuals and pikeminnow. Those > 400-mm long contained mercury what, if any, adverse impacts might be expected at the popu- above recommended toxicity guidelines designed to protect lation level. demographic endpoints, such as reproduction and survival. Although the role that selenium may play in counteracting Status of Colorado pikeminnow mercury toxicity is unknown, the relationship we found between mercury concentration and reduced body condition The status of wild Colorado pikeminnow populations in strongly suggests that injury is occurring. Tissue-level stud- all upper basin rivers remains tenuous. The San Juan River ies are needed to better understand physiological pathways population consists almost exclusively of stocked fsh, the of impairment and quantify toxicity efects. Managers tasked last capture of a wild adult having occurred in 2000 (Ryden with restoring sustainable Colorado pikeminnow popula- 2003; Furr and Davis 2009; Durst and Franssen 2014). At tions need to consider mercury contamination as an impor- least some of the stocked fsh have survived to maturity tant threat to demographic rates and recovery of Colorado and the presence of larvae in the system verifes successful pikeminnow. Collaborative eforts with regulatory agencies reproduction has occurred. However, recruitment of these

1 3 Archives of Environmental Contamination and Toxicology are needed so strategies to reduce sources of mercury can Yampa rivers, Utah and Colorado, 1979–2012. Final report to be developed. the Upper Colorado River Endangered Fish Recovery Program, Project FW BW-Synth, Denver, CO. Department of Fish, Wildlife, Because mercury deposition is so widespread, specifc and Conservation Biology, Colorado State University, Fort Col- fsh species from throughout North America and around lins. Larval Fish Laboratory Contribution 183 the globe are at risk of toxicity efects. Such efects may Bestgen KR, Walford C, Hill A, Wilcox T, and Hawkins J (2016a) be insignifcant for short-lived fshes or those that occupy Response of the native fsh communities of the Yampa and Green Rivers to non-native fshes and fows. In: 37th Annual researchers a low trophic niche. However, for long-lived predatory fsh, meeting of the Upper Colorado River Endangered Fish Recovery concentrations are expected to be high and can have deleteri- Program and the San Juan River Basin Recovery Implementation ous efects at both the individual and population level. When Program, Fort Lewis College, Durango, CO, 12–13 January, 2016 such species are already endangered due to other impacts Bestgen KR, Walford CD, White GC, Hawkins JA, Jones MT, Web- ber PA, Breen M, Skorupski J, Howard J, Creighton K, Logan to their environment, contaminants, such as mercury, can J, Battige K, Wright FB (2016b) Population status of Colorado impede eforts to improve reproduction and recruitment and pikeminnow in the Green River Basin, Utah and Colorado, therefore should not be ignored when developing recovery 2000–2013. Colorado River Recovery Implementation Program strategies. Project Number 128, Larval Fish Laboratory Contribution 200, Larval Fish Laboratory, Department of Fish, Wildlife and Con- Acknowledgements servation Biology, Colorado State University, Fort Collins, CO This study would not have been possible with- Boudou AM, Delnomdediew D, Georgescauld F, Ribeyve F, Saouter out several individuals who agreed to collect muscle plugs while in E (1991) Fundamental roles of biological barriers in mercury the feld. These personnel represent state and Federal agencies and accumulation and transfer in freshwater ecosystems, (analysis Colorado State University, all of which were connected with the Upper at organism, organ, cell, and molecular levels). Water Air Soil Colorado River Endangered Fish Recovery Program and/or San Juan Pollut 56:807–821 River Recovery Implementation Program. Many thanks to John Hawk- Brumbaugh WG, Krabbenhoft DP, Helsel DR, Wiener JG, Echols ins, Sam Finney, Dave Beers, Dale Ryden, Doug Osmundson, Travis KR (2001) A national pilot study of mercury contamination Francis, Trina Hedrick, Paul Badame, Darek Elverud, Boyd Wright, of aquatic ecosystems along multiple gradients: bioaccumula- Scott Durst, and numerous seasonal employees who participated in tion in fsh: US Geological Survey, Biological Science Report muscle plug collections. We also appreciate the assistance of Tom USGS/BRD/BSR-2001-0009 Czapla, Kevin Bestgen, Chuck McAda, Dave Irving, Mark Fuller, Pat- Butler DL, Wright WG, Hahn DA, Krueger RP, Osmundson BC rick Goddard, Lori Martin, and Dave Campbell in the study approval (1994) Physical, chemical, and biological data for detailed study process and in coordinating sampling eforts. Thanks to John Isanhart, of irrigation drainage in the Uncompahgre Project area and in who coordinated muscle plug collections and laboratory submittal of the Grand Valley, West-central Colorado, 1991–1992. US Geo- Colorado pikeminnow and roundtail chub samples during 2009 from logical Survey Open-File Report 94-110, US Geological Survey, the White River. Thanks to Larry Gamble, Kevin Johnson and John Denver, CO Wegrzyn for reviewing the study proposal and assisting in making this Butler DL, Krueger RP, Osmundson BC, Jensen EG (1995) Recon- project competitive for funding by the Environmental Contaminants naissance investigation of water quality, bottom sediment, and Program of the US Fish and Wildlife Service. Suggestions and editing biota associated with irrigation drainage in the Dolores Project by Doug Osmundson greatly improved the fnal version of the manu- area, Southwestern Colorado and Southeastern Utah, 1990–1991. script. Thoughtful comments from four anonymous reviewers are also US Geological Survey Open-File Report 94-4041, US Geological appreciated. This article refects the views of the authors and does not Survey, Denver, CO necessarily refect the views of the USFWS. Cizdziel JV, Hinners TA, Heithmar EM (2002) Determination of total mercury in fsh tissues using combustion atomic absorp- Open Access This article is distributed under the terms of the Crea- tion spectrometry with gold amalgamation. Water Air Soil Pollut tive Commons Attribution 4.0 International License (http://creativeco 135:355–370 mmons.org/licen ses/by/4.0/ ), which permits unrestricted use, distribu- Cladis DP, Kleiner AC, Santerre CR (2014) Mercury content in tion, and reproduction in any medium, provided you give appropriate commercially available fnfsh in the United States. J Food Prot credit to the original author(s) and the source, provide a link to the 77:1361–1366 Creative Commons license, and indicate if changes were made. Cocca P (2001) Mercury maps: a quantitative spatial link between air deposition and fsh tissue. USEPA Peer Reviewed Final Report EPA-823-R-01-009, Ofce of Water, Washington, D, p 29 References Crump KL, Trudeau VL (2009) Critical review: mercury-induced reproductive impairment in fish. Environ Toxicol Chem 28:895–907 Alvarez MC, Murphy CA, Rose KA, McCarthy ID, Fuiman LA (2006) Cuvin MLA, Furness RW (1988) Uptake and elimination of inorganic Maternal body burdens of methylmercury impair survival skills of mercury and selenium by minnows Phoxinus. Aquat Toxicol ofspring in Atlantic croaker (Micropogonias undulates). Aquat 13:205–215 Toxicol 80:329–337 Depew DC, Basu N, Burgess NM, Campbell LM, Devlin EW (2012) Baker RF, Blanchfeld PJ, Paterson MJ, Flett RJ, Wesson L (2004) Toxicity of dietary methylmercury to fsh: derivation of ecologi- Evaluation of nonlethal methods for the analysis of mercury in cally meaningful threshold concentrations. Environ Toxicol Chem fsh tissue. Trans Am Fish Soc 133:568–576 31:1536–1547 Beckvar N, Dillon TM, Reads LB (2005) Approaches for linking Dillon TN, Beckvar N, Kern J (2010) Residue-based mercury dose– whole-body fsh tissue residues of mercury or DDT to biological response in fsh: an analysis using lethality-equivalent test end- efects thresholds. Environ Toxicol Chem 24(8):2094–2105 points. Environ Toxicol Chem 29:2559–2565 Bestgen KR, Hill AA (2015) Reproduction, abundance, and recruitment dynamics of young Colorado pikeminnow in the Green and

1 3 Archives of Environmental Contamination and Toxicology

Drevnick PE, Sandheinrich MB (2003) Efects of dietary methylmer- contaminants and their efects on fsh in the Colorado River basin: cury on reproductive endocrinology of fathead minnows. Environ US Geological Survey Scientifc Investigations Report 2006-5163 Sci Technol 37:4390–4396 Hinck JE, Blazer VS, Denslow ND, Echols KR, Gross TS, May TW, Driscoll CT, Mason RP, Chan HM, Jacob DJ, Pirrone N (2013) Mer- Anderson PA, Coyle JJ, Tillitt DE (2007) Chemical contaminants, cury as a global pollutant: sources, pathways, and efects. Environ health indicators, and reproductive biomarker responses in fsh Sci Technol 47:4967–4983 from the Colorado River and its tributaries. Sci Total Environ Durst SL, Franssen NR (2014) Movement and growth of juvenile Colo- 378:376–402 rado pikeminnows in the San Juan River, Colorado, New Mexico, Johnson BM, Martinez PJ, Hawkins JA, Bestgen KR (2008) Rank- and Utah. Trans Am Fish Soc 143:519–527 ing predatory threats by nonnative fshes in the Yampa River, Eagles-Smith CA, Willacker JJ, Flanagan Pritz CM (2014) Mercury in Colorado, via bioenergetics modeling. N Am J Fish Manag fshes from 21 national parks in the western United States—inter- 28:1941–1953 and intra-park variation in concentrations and ecological risk. Khan MK, Wang F (2009) Critical review: mercury–selenium com- Open-File Report 2014-1051, US Geological Survey, Corvallis, pounds and their toxicological signifcance: toward a molecular OR, USA understanding of the mercury–selenium antagonism. Environ Eagles-Smith CA, Ackerman JT, Willacker JJ, Tate MT, Lutz MA, Toxicol Chem 28:1567–1577 Fleck JA, Stewart AR, Wiener JG, Evers DC, Lepak JM, Davis Kim JH, Birds E, Heisinger JF (1977) Protective action of selenium JA, Pritz CF (2015) Spatial and temporal patterns of mercury against mercury in northern creek chubs. Bull Environ Contam concentrations in freshwater fsh across the western United States Technol 17:132–136 and Canada. Sci Total Environ. https://doi.org/10.1016/j.scito Krueger RP (1988) Heavy metals analysis of seven Colorado squaw- tenv.2016.03.229 fsh from the Colorado and White rivers. U.S. Fish & Wildlife Eagles-Smith CA, Wiener JG, Eckley CS, Willacker JJ, Evers DC, Service, Grand Junction Marvin-DiPasquale M, Obrist D, Fleck JA, Aiken GR, Lepak JM, Le Cren ED (1951) The length-weight relationship and seasonal Jackson AK, Webster JP, Stewart AR, Davis JA, Alpers CN, Ack- cycle in gonad weight and condition in the perch (Perca fuvia- erman JT (2016) Mercury in western North America: a synthesis tilis). J Anim Ecol 20:201–219 of environmental contamination, fuxes, bioaccumulation, and risk Lindberg S, Bullock R, Ebinghaus R, Engstrom D, Feng XB, Fitzger- to fsh and wildlife. Sci Total Environ. https://doi.org/10.1016/j. ald W, Pirrone N, Prestbo E, Seigneur C (2007) The Madison scitotenv.2016.05.094 declaration on mercury pollution. Ambio 36:62–65 Elverud D, White G (2017) Status of Colorado pikeminnow in the Mailman M, Bodaly RA, Paterson MJ, Thompson S, Flett RJ (2014) Colorado River. In: 38th Annual Researchers Meeting, Upper Low-level experimental selenite additions decrease mercury in Colorado River Endangered Fish Recovery Program, San Juan aquatic food chains and fsh muscle but increase selenium in fsh River Recovery Implementation Program, January 10–11, 2017 gonads. Arch Environ Contam Toxicol 66:32–40 Engstrom DR (2007) Fish respond when the mercury rises. In: Pro- Mason RP, Sheu GR (2002) Role of the ocean in the global mercury ceedings of the National Academy of Science of the United States cycle. Glob Biogeochem Cycles 16:40-1–40-14 of America 104:16394–16395 Matta MB, Linse J, Cairncross C, Francendese L, Kocan RM (2001) Evers DC, Han YJ, Driscoll CT, Kamman NC, Goodale MW, Lambert Reproductive and transgenerational efects of methylmercury KF, Holsen TM, Chen CY, Clair TA, Butler T (2007) Biological or Aroclor 1268 on F. heteroclitus. Environ Toxicol Chem mercury hotspots in the northeastern United States and south- 20:327–335 eastern Canada. Bioscience 57:29–43. https://doi.org/10.1641/ Miller RR (1961) Man and the changing fsh fauna of the American B570107 southwest. Pap Mich Acad Sci Arts Lett 46:365–404 Fjeld E, Haugen TO, Vollestad LA (1998) Permanent impairment in Miller PS (2014) A population viability analysis for the Colorado the feeding behavior of grayling (Thymallus thymallus) exposed pikeminnow (Ptychocheilus lucius) in the San Juan River. Con- to methylmercury during embryogenesis. Sci Total Environ servation Breeding Specialist Group Report to BHP Billiton 213:247–254 New Mexico Coal, Apple Valley, MN Furr WD, Davis JE (2009) Augmentation of Colorado pikeminnow Moyle PB (1976) Inland fshes of California. University of California (Ptychocheilus lucius) in the San Juan River: 2008 Interim Pro- Press, Berkeley, pp 193–195, 229–231 gress Final Report Muth RT, Snyder DE (1995) Diets of young Colorado squawfsh and Gonzalez P, Dominique Y, Massabuau JC, Boudou A, Bourdineaud other small fsh in backwaters of the Green River, Colorado and JP (2005) Comparative efects of dietary methylmercury on gene Utah. Great Basin Nat 55(2):95–104 expression in liver, skeletal muscle, and brain of the zebrafsh Osmundson DB, Burnham KP (1998) Status and trends of the endan- (Danio rerio). Environ Sci Technol 39:3972–3980 gered Colorado squawfsh in the upper Colorado River. Trans Hamilton SJ, Holley KM, Buhl KJ, Bullard FA, Weston LK, McDonald Am Fish Soc 127:957–970 SF (2003) Evaluation of fushing of a backwater channel: con- Osmundson DB, White GC (2014) Population structure, abundance centrations of selenium and other inorganic elements in water, and recruitment of Colorado pikeminnow of the Upper Colorado sediment, invertebrates, forage fsh, and Colorado pikeminnow. River, 1991–2010. Final Report. U.S. Fish and Wildlife Service, Final Report to the Recovery Implementation Program for the Grand Junction, CO Endangered Fishes of the Upper Colorado River Basin, Denver, Osmundson DB, White GC (2017) Long-term, mark-recapture moni- CO. Yankton, SD: U.S. Geological Survey toring of a Colorado pikeminnow (Ptychocheilus lucius) popu- Hammerschmidt CR, Sandheinrich MB (2005) Maternal diet during lation: assessing recovery progress using demographic trends. oogenesis is the major source of methyl mercury in fsh embryos. Endanger Species Res 34:131–147. https://doi.org/10.3354/ Environ Sci Technol 39:3580–3584 esr00842 Hammerschmidt CR, Sandheinrich MB, Wiener JG, Rada RG (2002) Osmundson DB, Ryel RJ, Tucker ME, Burdick BD, Elmblad WR, Efects of dietary methylmercury on reproduction of fathead min- Chart TE (1998) Dispersal patterns of subadult and adult Colo- nows. Environ Sci Technol 36:877–883 rado squawfsh in the Upper Colorado River. Trans Am Fish Hinck JE, Blazer VS, Denslow ND, Gross TS, Echols KR, May TW, Soc 127:943–956 Orazio CE, Coyle JJ, Tillitt DE (2006) Biomonitoring of envi- Pacyna E, Pacyna J, Sundseth K, Munthe J, Kindbom K, Wilson S, ronmental status and trends (BEST) Program: environmental Steenhuisen F, Maxson P (2010) Global emission of mercury to

1 3 Archives of Environmental Contamination and Toxicology

the atmosphere from anthropogenic sources in 2005 and projec- 1998-2005. U.S. Geological Survey Scientifc Investigations Report tions to 2020. Atmos Environ 4:2487–2499 2009-5109, Reston, Virginia Pasquale MWD, Lutz MA, Brigham ME, Krabbenhoft DP, Aiken Seethaler K (1978) Life history and ecology of the Colorado squawfsh GR, Orem WH, Hall BD (2009) Mercury cycling in stream (Ptychocheilus lucius) in the upper Colorado River Basin. Thesis, ecosystems. 2. Benthic methylmercury production and bed Utah State University, Logan, UT sediment-pore water partitioning. Environ Sci Technol Seigneur CK, Vijayaraghavan K, Lohman Karamchandani P, Scott C 43:2726–2732 (2004) Global source attribution for mercury deposition in the Pearson E (2000) The analysis of mercury in fsh tissue plugs for the United States. Environ Sci Technol 38:555–569 purpose of evaluating a potentially non-lethal sampling method. Tan SW, Meiller JC, Mahafey KR (2009) The endocrine efects of mer- North Dakota Department of Health Report. North Dakota Dept. cury in humans and wildlife. Crit Rev Toxicol 39:228–269 of Health. Bismarck, ND Tsui MTK, Finlay JC, Balogh SJ, Nollet YH (2010) In Situ production Penglase S, Hamre K, Ellingsen S (2014) Selenium and mercury of methylmercury within a stream channel in Northern California. have synergistic negative efect on fsh reproduction. Aquat Environ Sci Technol 44:6998–7004 Toxicol 149:16–24 Tyus HM (1991) Ecology and management of Colorado squawfish Peterson SA, Van Sickle J, Hughes RM, Schacher JA, Echols SF (2005) (Ptychocheilus lucius). In: Minckley WL, Deacon S (eds) Battle A biopsy procedure for determining flet and predicting whole-fsh against extinction: management of native fshes in the American mercury concentration. Arch Environ Contam Toxicol 48:99–107 southwest. University of Arizona Press, Tucson, pp 379–402 Peterson SA, Van Sickle J, Herlihy AT, Hughes RM (2007) Mercury UNEP (United National Environmental Programme) (2013) Global mer- concentration in fsh from streams and rivers throughout the western cury assessment 2013—sources, emissions, releases and environ- United States. Environ Sci Technol 41:58–65 mental transport. UNEP Chemicals Branch, Geneva Peterson SA, Ralston NVC, Peck DV, Van Sickle J, Robertson JD, Spate USEPA (1998) US Environmental Protection Agency Mercury in solids VL, Morris JS (2009) How might selenium moderate the toxic and solution by thermal decomposition, amalgamation, and atomic efects of mercury in stream fsh of the Western US. Environ Sci absorption spectrophotometry, draft method 7473. US Environmen- Technol 43:3919–3925 tal Protection Agency, Washington Pirrone N, Cinnirella S, Feng X, Finkelman RB, Friedli HR, Leaner J et al USEPA (2001) US Environmental Protection Agency Water quality crite- (2010) Global mercury emissions to the atmosphere from anthropo- ria: notice of availability of water quality criterion for the protection genic and natural sources. Atmos Chem Phys 10:5951–5964 of human health: methylmercury. Federal Register 66(5) Jan 8 2001 Platania SP (1990) Biological summary of the 1987–1989 New Mexico- USEPA (2017) US Environmental Protection Agency Toxics Release Utah ichthyofaunal study of the San Juan River. Unpublished report Inventory (TRI) Program. http://www.epa.gov/toxics-release-inven to the New Mexico Department of Game and Fish, Santa Fe, and tory-tri-program/tri-data-and-tools. Accessed 2017 the U.S. Bureau of Reclamation, Salt Lake City, Utah, Cooperative USFWS (1967) US Fish and Wildlife Service Endangered species list. Agreement 7-FC-40-05060 Fed Reg 32:4001 Richter CA, Garcia-Reyero N, Martyniuk C, Knoebl I, Pope M, Wright- USFWS (1994) US Fish and Wildlife Service Endangered and threatened Osment MK, Denslow ND, Tillit DE (2011) Gene expression wildlife and plant; determination of critical habitat for the Colorado changes in female zebrafsh (Danio rerio) brain in response to acute River endangered fshes: razorback sucker, Colorado squawfsh, exposure to methylmercury. Environ Toxicol Chem 30(2):301–308 humpback chub, and bonytail chub. Fed Reg 59:13374–13399 Richter CA, Martyniuk CJ, Annis ML, Brumbaugh WG, Chasar LC, USFWS (2014) US Fish and Wildlife Service 2014. Colorado pikemin- Denslow ND, Tillitt DE (2014) Methylmercury-induced changes now (Ptychocheilus lucius) draft recovery plan. U.S. Fish and Wild- in gene transcription associated with neuroendocrine disruption in life Service, Mountain-Prairie Region 6, Denver, CO largemouth bass (Micropterus salmoides). Gen Comp Endocrinol USFWS (2014) US Fish and Wildlife Service 2013–2014 assessment 1:215–224. https://doi.org/10.1016/j.ygcen.214.03.029 of sufcient progress under the Upper Colorado River Endangered Ryden DW (2003) An augmentation plan for Colorado pikeminnow in the Fish Recovery Program in the Upper Colorado River Basin, and San Juan River. Final report. US Fish and Wildlife Service, Grand of implementation of action items in the January 10, 2005, fnal Junction, CO programmatic biological opinion on the management plan for endan- Sandheinrich MB, Atchison GJ (1990) Sublethal toxicant efects on fsh gered fshes in the Yampa River Basin, Memo from Noreen Walsh, foraging behavior: empirical versus mechanistic approaches. Envi- Regional Director, Region 6, September 10, 2014 ron Toxicol Chem 9:107–119 USFWS (2015) US Fish and Wildlife Service Draft 2014–2015 assess- Sandheinrich MB, Miller KM (2006) Efects of dietary methylmercury on ment of sufcient progress under the Upper Colorado River Endan- reproductive behavior of fathead minnows (Pimephales promelas). gered Fish Recovery Program in the Upper Colorado River Basin, Environ Toxicol Chem 25:3053–3057 and of implementation of action items in the December 20, 1999, Sandheinrich MB, Wiener JG (2011) Methylmercury in freshwater fsh: 15-Mile Reach Programmatic Biological Opinion and December recent advances in assessing toxicity of environmentally relevant 4, 2009, Gunnison River Basin Programmatic Biological Opinion. exposures, pp 169–190. In: Beyer WN, Meador JP (eds) Environ- Draft memo from Noreen Walsh, Regional Director, Region 6. mental contaminants in biota: interpreting tissue concentrations, 2nd August 4, 2014 edn. CRC Press, Boca Raton, p 751 Vanicek CD, Kramer RH (1969) Life history of the Colorado squawfsh Sather ME, Mukerjee S, Smith L, Mathew J, Jackson C, Callison R (Ptychocheilus lucius) and the Colorado chub (Gila robusta) in the (2013) Gaseous oxidized mercury dry deposition measurements in Green River in Dinosaur National Monument, 1964–1966. Trans the Four Corners area and Eastern Oklahoma, USA. Atmos Pollut Am Fish Soc 98(2):193–208 Res 4:168–180 Walters DM, Rosi-Marshall E, Kennedy TA, Cross WF, Baxter CV Schmidt CJ, Brumbaugh WG (1990) National Contaminant Biomonitor- (2015) Mercury and selenium accumulation in the Colorado ing Program—concentrations of arsenic, cadmium, copper, lead, River food web, Grand Canyon, USA. Environ Toxicol Chem mercury, selenium, and zinc in U.S. freshwater fsh, 1976–1984. 34(10):2385–2394 Arch Environ Contam Toxicol 19:731–747 Webber HM, Haines TA (2003) Mercury efects on predator avoidance Scudder BC, Chasar LC, Wentz DA, Bauch NJ (2009) Mercury in fsh, behavior of a forage fsh, golden shiner (Notemigonus crysoleucas). bed sediment, and water from streams across the United States, Environ Toxicol Chem 22:1556–1561

1 3 Archives of Environmental Contamination and Toxicology

Weber DN (2006) Dose-dependent efects of developmental mercury Wiener JG, Spry DJ (1996) Toxicological signifcance of mercury in exposure on C-start escape responses of larval zebrafsh Danio rerio. freshwater fsh. In: Beyer WN, Heinz GH, Redmon-Norwood AW J Fish Biol 69:75–94 (eds) Environmental contaminants in wildlife: interpreting tissue Weidner K (2007) Investigating the efects of mercury emissions in the concentrations. Lewis Publishers, Boca Raton, pp 297–339 Four Corners Area on local deposition levels and ambient concen- Williamson JH (1992) Colorado pikeminnow genetic survey-tissue trations. Masters Project Report, Duke University, Durham, NC sampling protocol. US Fish and Wildlife Service, Denver http://dukespace.lib.duke.edu/dspace/bitstream/handle/10161/407/ MP_klw20_a_200709.pdf?sequence=1. Accessed 2008

1 3