Comparison of Body Composition and Fatty Acid Profiles between Wild and Cultured Rio Grande Silvery Minnows

Hatcheries generally use pelleted or flake diets to intensively feed fish to maximize growth and prerelease survival (as opposed to semi-intensive or extensive feeding). However, intensively feeding fish raised for conservation can be counterproductive for several reasons. First, intensive feeding increases survival of subviable individuals (genotypes), which leads to development of domesticated populations (Tave 1993, 1999; Tave et al. 2011). Second, intensive feeding can affect the development of natural foraging and predator-avoidance behaviors in some fish (Ellis et al. 2002) and can also produce maladaptive behavior, such as surface hovering (Tave et al. 2011).

Current conservation hatchery management strategies seek to create an environment that is as similar to the natural habitat as practicable (Maynard et al. 1996; Braithwaite and Salvanes 2005). In this regard, allowing fish to forage on natural food organisms rather than supplying them with manufactured fish feed advances this idea (Flagg and Nash 1999; Tave et al. 2011). This is a key component of conservation aquaculture management because fish that can selectively forage take better advantage of their native nutritional environment, which should increase the likelihood that they will survive after stocking and ultimately leave behind offspring with these same traits that can be passed on (Raubenheimer et al. 2012).

The Los Lunas Silvery Minnow Refugium (hereafter refugium) is a conservation fish hatchery that extensively propagates the federally endangered (ESA 1973, as amended) Rio Grande Silvery Minnow, Hybognathus amarus (RGSM), using conservation aquaculture management (Tave et al. 2011). The overarching goal of conservation aquaculture is to raise fish under conditions that mimic those of the target environment, and in doing so minimize domestication (Tave et al. 2011; Hutson et al. 2012). A key component of this management is to avoid the use of formulated feed in a naturalized outdoor refugium. Instead, the refugium was designed to mimic the natural environment of RGSM in the Rio Grande (Tave et al. 2011). Research has shown RGSM to be grazers with an omnivorous diet composed of benthic algae, diatoms, invertebrates, and crustaceans (Bixby and Burdett 2014). Comparison of extensively raised RGSM in the refugium with wild RGSM would assess similarities in selective foraging and the natural food available in the refugium environment to that of the Middle Rio Grande, a 280-km stretch of river from Cochiti Reservoir to Elephant Butte Reservoir in New Mexico, containing surviving remnants of the natural population (U.S. Fish and Wildlife Service 2010). To further compare fish raised under conservation aquaculture management, RGSM were also raised in tanks at Los Lunas where they were fed RGSM flake feed (Caldwell et al. 2010) with or without access to natural food organisms. Similarities and differences among these groups as a result of diet can be assessed through examination of whole-body proximate composition and fatty acid profiles.

The polyunsaturated fatty acids linoleic acid (18:2n6) and α-linolenic acid (18:3n3) are essential dietary components in fish. These polyunsaturated fatty acids are further modified into biologically active, highly unsaturated fatty acids namely, arachidonic acid (20:4n6), docosahexaenoic acid (22:6n3, DHA), and eicosapentaenoic acid (20:5n3, EPA) (Sargent et al. 2002; Tocher 2003). Analysis of polyunsaturated fatty acidand highly unsaturated fatty acid profiles provides information about lipid components of tissues and, by extension, the fish's diet, since fatty acid composition in tissues reflects fatty acid composition of the diet (Corraze 2001). Fatty acid profiles also provide information on variation in foraging and diversity of dietary items ingested (Higgs et al. 1995). Additionally, fatty acid analysis can be used to show changes in diet as fish utilize seasonal forage or change from juvenile to adult diets (Iverson et al. 2002). Comparing fatty acid profiles between postrelease and wild fish can indicate whether postrelease fish have successfully adapted to foraging in the environment (Powell et al. 2010). Finally, analysis of fatty acids can also show where potential problems may arise with deficient diets. For instance, larval diets deficient in DHA can lead to visual or neural impairment (Bell et al. 1995; Shields et al. 1999), affect innate and adaptive immunity (Waagbo et al. 1995; Thompson et al. 1996), or decrease resistance to pathogenic challenges (Fracalossi and Lovell 1994).

Our objective was to determine how closely the fat content and lipid profiles of RGSM cultured in the Los Lunas Silvery Minnow Refugium compare with wild RGSM living in the Middle Rio Grande River. We also compared both groups with RGSM raised in conventional tank systems and fed an artificial diet. Whether RGSM in the refugium approximate wild fish in fat content and fatty acid composition or may alternatively need to be supplemented with an artificial diet is addressed. A sample of RGSM feed was also analyzed for fatty acid composition to provide information to compare it with fatty acid profiles of each RGSM sample.

The RGSM used in this study were produced by a natural spawn in the Los Lunas Silvery Minnow Refugium in April 2012 (Tave and Hutson 2012). Culture of the offspring in the refugium is described by Hutson et al. (2012) and Tave and Hutson (2012). No manufactured feed was provided to the fish (hereafter referred to as the “Refugium” group). Inorganic and organic fertilizers were used to stimulate natural food production in the refugium, and this natural food was the only source of nutrition for the fish.

One-hundred 1-mo-old RGSM in the refugium were seined on June 13 and transferred to one of two tank systems: an indoor recirculating aquaculture system or outdoor tanks. Fish were not measured or weighed before stocking, as growth was not a component of this study because of differences in stocking densities and initial starting size (refugium fish started as eggs and tank fish started as 1-mo-old fry). Fifty fish were stocked in a 1.83-m-diameter 2,220-L tank in the indoor hatchery. This tank was one of five in a recirculating aquaculture system (Tave et al. 2011). These fish were the only fish in this system during the study and were fed RGSM flake feed (Table 1) to apparent satiation twice daily for 117 d (hereafter referred to as the “Hatchery” group). Another 50 fish were stocked into each of two 2.44-m-diameter 3,666-L outdoor tanks (Tave et al. 2012). Fish in the outdoor tanks were raised as described by Hutson et al. (2013). These tanks were “seeded” with water from the refugium and fertilized twice a week to mimic water conditions in the refugium. Fish in outdoor tanks were fed with RGSM flake feed to apparent satiation twice daily for 117 d (hereafter referred to as the “Supplemented” group). Nine fish from each treatment (Refugium, Hatchery, and Supplemented) were sampled on October 8 for whole-body proximate analysis and fatty acid profiles. Because of their small size, three fish each from the same treatment were pooled. Consequently, there were three pooled samples of three fish per treatment.

Wild fish (n = 9) were collected from the Angostura Reach of the Middle Rio Grande on November 26 (hereafter referred to as the “Wild” group). The Angostura Reach is the uppermost of three reaches of the Middle Rio Grande where RGSM occur (U.S. Fish and Wildlife Service 2010). Fish were sampled from this reach for two reasons. First, it was the only reach that did not experience significant drying in 2012, making it less likely that stressful environmental conditions would alter fatty acid profiles. Second, it had been last stocked in 2007 and, presumably, all of the fish collected in this reach would be wild-spawned fish since few wild RGSM live more than 3 y (Horwitz et al. 2011). Nine fish were captured from two sample areas along the west channel of the Middle Rio Grande near the La Orilla Drain in Albuquerque with a beach seine (3.1 × 1.8 m; 3-mm mesh). Three fish were collected from the southern area and six fish were collected from the northern area. Coordinates (GPS) for the two sample areas were: northern area 35°09′30.8988″N, 106°40′13.2518″W; southern area 35°09′21.9780″N, 106°40′16.5836″W.32.

All fish were immediately euthanized with MS-222 (250 mg/L), measured (total length) to the nearest millimeter, weighed to the nearest 0.01 g, put in plastic freezer bags (three fish per bag) placed on ice, and transported to an ultralow freezer (−70°C). A sample of the RGSM flake feed was also stored at −70°C. All samples were shipped on dry ice overnight to the University of Missouri-Columbia Agricultural Experiment Station Chemical Laboratories, Columbia, Missouri for whole-body lipid determination using the Folch extraction method (Bligh and Dyer 1959). Fatty acids were analyzed using an Agilent Technologies 7890A gas chromatograph (Agilent Technologies, Santa Clara, CA) and Association of Official Analytical Chemists official methods (996.06, 965.49; AOAC 1995) and American Oil Chemists Society official method Ce 2-66 (AOCS 2010). Separated fatty acids were identified by comparing retention times with those of standards (Supelco, Bellefonte, PA) and expressed on a relative percentage basis. Moisture in samples was measured using the official method Ca 2c-25 in Official Methods and Recommended Practices of the AOCS (AOCS 2010).

Fatty acid percentages were compared across treatments using the general linearized model procedure PROC GLM in SAS, ver. 9.3 (Cary, NC) after assessing normality using Shapiro–Wilk tests and homogeneity of variances using Bartlett's test. Tukey's honestly significant difference tests were used post hoc for each comparison to determine significance between treatments. Data are expressed as means ± SEM. A significance criterion of P ≤ 0.05 was used for all comparisons and tests.

Food quality has long been known to be a significant factor in fish growth and health (Halver and Hardy 2002). Although comparison of growth among groups was not an objective of this study, RGSM sampled from the Hatchery group (indoor tanks) had higher average weight (P = 0.046) than RGSM from the Supplemented group (outdoor tanks), even though they started from a “common garden” as 1-mo-old fry and had access to the same hatchery diet. Mean length and weight of fish in the four groups sampled for proximate analysis were: Refugium (natural food organisms only), 75.6 mm and 3.97 g; Hatchery (artificial feed only), 75.2 mm and 4.28 g; Supplemented (both natural food organisms and artificial feed), 74.0 mm and 3.77 g; Wild fish, 75.6 mm and 3.89 g. It should be noted that similar sizes of the fish in the four treatments was an unplanned coincidence.

Considerable work has been done on essential fatty acid requirements in various fish species (reviewed by Glencross 2009; Tocher 2003; 2010; Tocher and Glencross 2015) and artificial diets for cultured fish typically contain more than sufficient amounts of essential fatty acids to avoid any potential deficiencies. The RGSM diet formulation (Table 1) as described by Caldwell et al. (2010) was analyzed for fatty acid composition (Table 2). By way of example, the amount of linoleic acid present in the RGSM hatchery diet was 1.83% (14.82% of total lipid), which is well above what is believed to be required for freshwater fish at 1% of diet dry weight (NRC 2011; Tocher and Glencross 2015). This diet, previously optimized by Caldwell et al. (2010), contains more than sufficient quantities of essential fatty acids and any signs of deficiencies in RGSM consuming this diet have not been reported.

Percent whole-body lipid did not differ between Wild RGSM and Refugium RGSM (7.8% ± 0.2% and 8.3% ± 0.06%, respectively; Figure 1). RGSM fed the hatchery diet with access to natural food sources (Supplemented) had significantly higher percent whole-body lipid than Wild or Refugium fish (10.6% ± 0.1%). RGSM only fed the hatchery diet (Hatchery) had significantly higher percent whole-body lipid (15.5% ± 0.5%) and also significantly lower moisture content (65.7% ± 0.95%) than other treatment groups. Moisture content did not differ among Wild, Refugium, or Supplemented treatments (71.1% ± 0.55%, 71.5% ± 0.12%, 69.4% ± 0.27%, respectively). Data for crude fat and moisture analyses of samples and individual fatty acid percentages within treatments are provided in a supplemental table in Data S1.

Although overall length was similar among groups, condition factor (Fulton's K factor: K = 100[W/L³]) of fish fed the hatchery diet was also higher, K = 1.00 and 0.93 for the Hatchery and Supplemented groups, respectively, compared with the Refugium group with 0.91. In comparison, condition factor of the Wild RGSM sample was significantly lower at 0.90 than the Hatchery group (P = 0.030). With respect to percent whole-body fat and condition factor, Refugium fish appear more similar to Wild fish than Supplemented or Hatchery RGSM in the two tank systems. These data highlight several important points regarding intensive, semi-intensive, and extensive propagation of RGSM. The Supplemented and Hatchery groups with access to the hatchery diet likely had greater fat content due to differences between minimum required dietary fatty acid levels reported for various fish species and optimal levels, which result in increased growth (Tocher 2003, 2010; Glencross 2009). The hatchery diet contained marine fish oil, which is not part of the dietary regime of RGSM in the wild or refugium. The hatchery diet also contained high percentages of animal feedstuffs (fish meal, krill meal, liver meal, and egg solids) not available as natural food items. Whether or not higher whole-body fat content confers a survival advantage to captively raised RGSM when released into the Rio Grande is unknown. Information on RGSM dietary fatty acid requirements and optimal dietary levels of fatty acids that enhance growth and survival is lacking.

Comparison of fatty acid profiles among RGSM nutritional groups showed marked differences in fatty acid percentages between Wild fish, Refugium fish, and Hatchery fish (those fed the hatchery diet) (Table 3). Total omega-3 fatty acids were highest in Wild fish but similar among other groups. Within omega-3 fatty acids, Wild fish contained higher amounts of EPA and clupanodonic acid (22:5n3). Levels of α-linolenic acid and DHA were similar in Wild fish and the Hatchery fish exclusively fed the hatchery diet. Omega-3 fatty acids play key roles in cell membrane fluidity and are known to change with temperature (Wodtke and Cossins 1991). Fish eggs are rich in DHA, as are neural and retinal tissues (Ahlgren et al. 2009).

Clupanodonic acid is a precursor to both EPA and DHA, with interconversion to EPA being a reversible reaction involving the enzymes elongase 2 and elongase 5 (Kaur et al. 2011). Clupanodonic acid is typically found in fish oil. The RGSM hatchery diet had a fatty acid profile indicating 0.97% clupanodonic acid, 10.17% EPA, and 6.38% DHA (Table 2). Clupanodonic acid percentage was significantly higher in RGSM samples from the Wild group than other groups (Table 3). Since clupanodonic acid is not prevalent in natural food items, its presence in Wild RGSM tissues likely reflects a metabolic intermediate between α-linolenic acid being converted to DHA or EPA and indirectly supports the higher percentage of DHA and EPA in Wild RGSM.

The high levels of EPA observed in Wild RGSM may also be explained, in part, by high EPA levels in their natural food sources in addition to active conversion of α-linolenic acid. Freshwater diatoms, an important food source for RGSM, contain a high percentage of EPA, 16.9% ± 8.2% (Ahlgren et al. 2009). However, the EPA percentage observed in Refugium fish that also consume diatoms (Bixby and Burdett 2014) was lower than the percentage of EPA in Wild RGSM. This difference is likely a reflection of dissimilarities between food assemblages of the refugium and the Rio Grande, even though the refugium is geographically very close to the Rio Grande. Changes in diatom assemblages have been previously shown to affect overall lipid content in benthic algae grazers (Steinman 1996). Comparisons between Refugium fish and the Supplemented group (access to both natural and artificial feed) show similar levels in every omega-3 fatty acid except for α-linolenic acid, which was significantly higher in the Refugium group (Table 3). This indicates that intake of natural food present in the outdoor tanks had an impact on fatty acid profiles of the Supplemented group apart from what was provided by the available hatchery diet.

Total omega-6 fatty acid profiles showed a significantly lower overall percentage in the Wild group than in other groups, with omega-6 fatty acid levels nearly five times higher in Hatchery group (Table 3). Although consumption of the hatchery diet can explain the levels of linoleic acid among the Hatchery and Supplemented groups, it cannot explain the higher levels of linoleic acid in the Refugium group. The significantly higher linoleic acid percentage in the Refugium group has to be associated with the particular assemblage of food items available in the refugium. Levels of arachidonic acid in fish from the Hatchery group were similar to percentages in the RGSM hatchery diet. Thus, the trend for increasing linoleic acid percentages in Wild, Refugium, Supplemented, and Hatchery groups respectively was not mirrored in arachidonic acid percentages and may be a consequence of different rates of conversion of linoleic acid as well as different dietary percentages of arachidonic acid. How differing levels of linoleic acid and arachidonic acid or rates of conversion of arachidonic acid manifest in RGSM health and growth is presently unknown.

The differing percentages of linoleic and arachidonic acids observed among treatments in this study serve to underscore the complex milieu of food item availability, selective foraging, supplementation with an artificial diet, and conversion of essential fatty acids within the tissues. More important, specific fatty acids such as arachidonic acid are becoming increasingly recognized as significant dietary nutrients for juvenile fish because of arachidonic acid's role as a precursor to eicosanoids, key elements of the immune system (Bell and Sargent 2003; Bransden et al. 2004). Additionally, arachidonic acid has been shown to be important in fish such as turbot and Japanese flounder for somatic growth, survival, and pigmentation (Estevez et al. 1997, 1999; Bell and Sargent 2003). Despite differences among treatments, all cultured RGSM in this study had levels of essential linoleic and arachidonic fatty acids at or above what was observed in Wild RGSM.

Comparisons of n-3:n-6 ratios showed Wild RGSM to be higher (9.09 ± 0.28) than other treatment groups, with a much higher proportion of omega-3 fatty acids and significantly less omega-6 fatty acids as noted above (Figure 2A, Table 3, and Data S1). No significant differences were observed in n-3:n-6 ratios among cultured RGSM, which were 1.75 ± 0.06, 1.42 ± 0.03, and 1.34 ± 0.01 in the Refugium, Supplemented, and Hatchery groups, respectively. The n-3:n-6 ratio of the RGSM diet (1.34 ± 0.01) was similar to that observed in the Hatchery group. Likewise, Hatchery RGSM also had significantly higher DHA:arachidonic acid ratios (5.83 ± 0.26) than all other groups (Figure 2B), which is likely a reflection of the even higher hatchery diet DHA:arachidonic acid ratio (8.51 ± 0.07). The DHA:arachidonic acid ratio of the Wild group (4.25 ± 0.27) was significantly higher than Supplemented (1.73 ± 0.06) and Refugium (0.85 ± 0.13) RGSM, which did not differ. Clupanodonic acid and DHA ratios (clupanodonic acid:DHA) were similar between Wild RGSM and Refugium RGSM (0.54 ± 0.01 and 0.52 ± 0.02, respectively). These ratios were significantly different (P = 0.024) from the Supplemented and Hatchery RGSM treatments (0.48 ± 0.01 and 0.27 ± 0.03, respectively). In this instance, clupanodonic acid was significantly higher in Wild RGSM than observed in cultured RGSM, but DHA percentages were similar between Wild and Hatchery RGSM. Within this study, differences in the ratios of biologically active fatty acids among treatments also underscore the complexity of dietary availability and fatty acid conversion.

Numerous studies have shown differences in fatty acid composition between cultured and wild fish of various species, particularly when n-3:n-6 or DHA:arachidonic acid ratios are compared (Otwell and Richards 1982; Van Vliet and Katan 1990; Yang and Dick 1994; Pickova et al. 1999). Typically, n-3:n-6 ratios are two to three times higher in intensively cultured fish than in wild fish and DHA:arachidonic acid ratios are three to five times higher in cultured fish than in wild fish. This is again mostly a reflection of artificial diets and fish oil used in the captive culture of the various species. In this study, however, we observed the opposite (Figure 2A), with the n-3:n-6 ratio in Wild RGSM (9.09 ± 0.28) more than five times higher than the next highest ratio observed in fish from the Refugium group (1.75 ± 0.06) and nearly seven times higher than Hatchery RGSM (1.34 ± 0.01). As mentioned, the high n-3:n-6 ratio observed in Wild RGSM may be a consequence of their diet containing a high percentage of freshwater diatoms, which have reported mean n-3:n-6 ratios of 7.6 ± 4.8 (Muller-Navarra 1995, 2006; Desvilettes et al. 1997; Gatenby et al. 2003; Caramujo et al. 2008). The DHA:arachidonic acid ratio in Wild RGSM (4.25 ± 0.27) was also five times higher than the ratio observed in the Refugium group (0.85 ± 0.13) but, in contrast, slightly less than the ratio observed in the Hatchery RGSM (5.83 ± 0.26; Figure 2B). The DHA:arachidonic acid ratio was also higher in the hatchery diet (8.47 ± 0.07) than observed in RGSM from the Supplemented and Hatchery groups.

All groups showed significant differences in lipid level, suggesting that RGSM selectively foraged on various food items available in each of their respective habitats and have plasticity in a range food item selection. Given that fish from the Los Lunas Silvery Minnow Refugium and both tank systems were from a genetic common garden, it follows that their rates of incorporation, turnover, and conversion of α-linolenic acid into DHA and EPA and conversion of linoleic acid into arachidonic acid would be similar given comparable diets. Regarding preference for natural food items or the hatchery diet, RGSM in the Supplemented group apparently consumed both. The hatchery diet provides essential fatty acids at levels above what are considered to be minimum required levels and RGSM readily consume the diet even when natural food items are available. Omega-3 fatty acid percentages were similar between the Supplemented and Refugium RGSM and omega-6 fatty acid percentages were intermediate in these two groups along a decreasing trend from the highest percentage observed in Hatchery RGSM to a low percentage in Wild fish.

Diatoms, ostracods, and dipterans are important natural food resources in the refugium since these food items are preferentially consumed by RGSM (Bixby and Burdett 2014). Despite the refugium's proximity to the Rio Grande, where available natural forage would be postulated to be similar, RGSM from the refugium were different in their fatty acid composition compared with wild fish. This is not entirely surprising since the refugium receives inputs of organic and inorganic fertilizers to boost primary productivity (Hutson et al. 2012) that can alter the abundance of some food items relative to the nearby Rio Grande, which does not receive those inputs. Outside tanks (Supplemented group) were seeded with refugium water and similarly fertilized to mimic refugium natural production as much as possible. That said, fatty acid profiles of fish from the Supplemented group were intermediary between the Hatchery group and Refugium fish. This suggests that although Refugium fish are selectively foraging on somewhat different food items than in the Rio Grande, intake goals and quality of food are likely comparable given condition factor and overall fat content similarities to Wild RGSM. In this respect, similarities of Wild fish to Refugium RGSM further support and affirm the conservation goals of the refugium, which is to say the conservation aquaculture management used at the refugium produces an environment that is similar nutritionally to that in the Rio Grande. Furthermore, it lends support to the idea that conservation aquaculture management can be used to produce fish that are more similar to wild fish than fish that are produced using production-type aquaculture management.

Underlying physiological mechanisms that govern energy allocation and reserve partitioning in fish inhabiting stable (i.e., hatchery) vs. less-stable environments (i.e., natural environments) remain poorly understood. Although a variety of health effects of dietary lipid and fatty acids have been reported, data are often confounding between species and inconsistent between experimental trials (Tocher and Glencross 2015). In general, condition factors have often been used to quantify relative fish health, inferring that fish with higher indices have greater energy stores and thus greater ability to engage in energetically challenging activities such as reproduction or migration (Stevenson and Woods 2006). Fulton's condition factor has long been used as a metric for population studies because of its simplicity (Anderson and Neuman 1996). In this study, hatchery-fed fish had a higher Fulton's condition factor than wild RGSM. Although intuitively greater energy stores may seem advantageous, some salmonid studies have shown that excess body fat may compromise fitness (Pearsons et al. 2013). Likewise, Fulton's condition factor does not always correlate well with fat content, gender, and seasonal differences or nutritional status (Sutton et al. 2000). Whether a higher condition factor for hatchery fish imparts any poststocking fitness advantage in RGSM is unknown and relying on condition factor as a measure of fitness needs to be determined empirically to help with recovery efforts.

In summary, this study identified significant differences in lipid content and fatty acid composition between wild RGSM and cultured Silvery Minnows, reflecting their respective diets and culture regimes. Given similarities between fat content and condition factor between wild RGSM and fish raised in the refugium, we conclude that it is not necessary to supplement feeding with an artificial diet for this type of naturalized conservation management. Rio Grande Silvery Minnows readily forage on the variety of natural food items present and also artificial feed when available, indicating dietary plasticity, which is advantageous for fish culture and future recovery. The greatest value of naturalized rearing (i.e., the refugium) probably lies in other similarities it shares with the Rio Grande such as natural light, stream velocity, temporal flooding, and other habitat parameters (Hutson et al. 2012). Determining how well RGSM survive after stocking regardless of culture regime will require longitudinal studies with multiple time points and more samples.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Data S1. Spreadsheet comparing fatty acid profiles (expressed as % = g/100 g of lipid) among Rio Grande Silvery Minnows (Hybognathus amarus) either collected from the Rio Grande on November 26, 2013 (Wild) or raised from June 13 to October 8, 2013 in the Los Lunas Silvery Minnow Refugium (Los Lunas, New Mexico), an outdoor refugium (Refugium), in outdoor tanks with access to natural food and supplemented with a hatchery diet (Supplemented), or raised in an indoor recirculating aquaculture system and exclusively fed a hatchery diet (Hatchery).

Found at DOI: http://dx.doi.org/10.3996/072016-JFWM-055.S1 (54 KB PDF).

Reference S1. Bixby R, Burdett A. 2014. Resource utilization by the Rio Grande silvery minnow at the Los Lunas Silvery Minnow Refugium. Interstate Stream Commission, Santa Fe, NM.

Found at DOI: http://dx.doi.org/10.3996/072016-JFWM-055.S2 (873 KB PDF).

Reference S2. Flagg TA, Nash CE. 1999. A conceptual framework for conservation hatchery strategies for Pacific salmonids. U.S. Department of Commerce, NOAA Technical Memo NMFS-NWFCS-38.

Found at DOI: http://dx.doi.org/10.3996/072016-JFWM-055.S3 (694 KB PDF); also available at https://www.nwfsc.noaa.gov/publications/scipubs/techmemos/tm38/tm38.htm).

Reference S3. Tave D, Hutson AM. 2012. 2012 Annual Report. USFWS TE Permit 169770-5. Los Lunas Silvery Minnow Refugium, 1000 Main Street NW, Building H, Los Lunas, NM, USA.

Found at DOI: http://dx.doi.org/10.3996/072016-JFWM-055.S4 (1109 KB PDF).

Reference S4. U.S. Fish and Wildlife Service. 2010. Revised Rio Grande Silvery Minnow (Hybognathus amarus) Recovery Plan. Albuquerque, NM: U.S. Fish and Wildlife Service. Found at DOI: http://dx.doi.org/10.3996/072016-JFWM-055.S5; also available at http://ecos.fws.gov/docs/recovery_plan/022210_v2.pdf (1109 KB PDF).

We thank Eric Gonzales and Pauletta Dodge for assistance in capturing fish from the Angostura Reach and Ken Ferjancic (HDR, Inc.) and Grace Haggerty for reviewing the manuscript. We also thank the anonymous reviewers and Associate Editor for their constructive comments that greatly improved the manuscript. Fish were cultured under U.S. Fish and Wildlife Service Permit TE169770-5 and New Mexico Department of Game and Fish Permit 3417. Funds for the project were provided by U.S. Bureau of Reclamation contract RO9PC40009 and Grant Agreement No. 08–FG–40–2803 and by the New Mexico Interstate Stream Commission.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.