Endangered and Threatened Wildlife and Plants; Revised 12-Month Finding to List the Upper Missouri River Distinct Population Segment of Arctic Grayling as Endangered or Threatened

DEPARTMENT OF THE INTERIOR

Fish and Wildlife Service

50 CFR Part 17

[Docket No. FWS-R6-ES-2009-0065]

[MO 92210-0-0008-B2]

Endangered and Threatened Wildlife and Plants; Revised 12-Month Finding to List the Upper Missouri River Distinct Population Segment of Arctic Grayling as Endangered or Threatened

AGENCY:

Fish and Wildlife Service, Interior.

ACTION:

Notice of revised 12-month finding.

SUMMARY:

We, the U.S. Fish and Wildlife Service (Service/USFWS), announce a revised 12-month finding on a petition to list the upper Missouri River Distinct Population Segment (Missouri River DPS) of Arctic grayling (

Thymallus arcticus

) as endangered or threatened under the Endangered Species Act of 1973, as amended. After review of all available scientific and commercial information, we find that listing the upper Missouri River DPS of Arctic grayling as endangered or threatened is warranted. However, listing the upper Missouri River DPS of Arctic grayling is currently precluded by higher priority actions to amend the Lists of Endangered and Threatened Wildlife and Plants. Upon publication of this 12-month finding, we will add the upper Missouri River DPS of Arctic grayling to our candidate species list. We will develop a proposed rule to list this DPS as our priorities allow. We will make any determination on critical habitat during development of the proposed listing rule. In the interim, we will address the status of this DPS through our annual Candidate Notice of Review (CNOR).

DATES:

The finding announced in this document was made on September 8, 2010.

ADDRESSES:

This finding is available on the Internet at

http://www.regulations.gov

at Docket Number

FWS-R6-ES-2009-0065

. Supporting documentation we used in preparing this finding is available for public inspection, by appointment, during normal business hours at the U.S. Fish and Wildlife Service, Montana Field Office, 585 Shepard Way, Helena, MT 59601. Please submit any new information, materials, comments, or questions concerning this finding to the above street address (Attention: Arctic grayling).

FOR FURTHER INFORMATION CONTACT:

Mark Wilson, Field Supervisor, Montana Field Office (see

ADDRESSES

); by telephone at 406-449-5225; or by facsimile at 406-449-5339. Persons who use a telecommunications device for the deaf (TDD) may call the Federal Information Relay Service (FIRS) at 800-877-8339.

SUPPLEMENTARY INFORMATION:

Background

Section 4(b)(3)(B) of the Endangered Species Act of 1973, as amended (ESA) (16 U.S.C. 1531

et seq.

), requires that, for any petition containing substantial scientific or commercial information indicating that listing the species may be warranted, we make a finding within 12 months of the date of receipt of the petition. In this finding, we determine that the petitioned action is: (a) Not warranted, (b) warranted, or (c) warranted, but immediate proposal of a regulation implementing the petitioned action is precluded by other pending proposals to determine whether species are endangered or threatened, and expeditious progress is being made to add or remove qualified species from the Federal Lists of Endangered and Threatened Wildlife and Plants. Section 4(b)(3)(C) of the ESA requires that we treat a petition for which the requested action is found to be warranted but precluded as though resubmitted on the date of such finding, that is, requiring a subsequent finding to be made within 12 months. We must publish these 12-month findings in the

Federal Register

.

Previous Federal Actions

We have published a number of documents on Arctic grayling and have been involved in litigation over previous findings. We describe our actions relevant to this notice below.

We initiated a status review for the Montana Arctic grayling (

Thymallus arcticus montanus

) in a

Federal Register

notice on December 30, 1982 (47 FR 58454). In that notice, we designated the purported subspecies, Montana Arctic grayling, as a Category 2 species. At that time, we designated a species as Category 2 if a listing as endangered or threatened was possibly appropriate, but we did not have sufficient data to support a proposed rule to list the species.

On October 9, 1991, the Biodiversity Legal Foundation and George Wuerthner petitioned us to list the fluvial (riverine populations) of Arctic grayling in the upper Missouri River basin as an endangered species throughout its historical range in the coterminous United States. We published a notice of a 90-day finding in the January 19, 1993,

Federal Register

(58 FR 4975), concluding the petitioners presented substantial information indicating that listing the fluvial Arctic grayling of the upper Missouri River in Montana and northwestern Wyoming may be warranted. This finding noted that taxonomic recognition of the Montana Arctic grayling (

Thymallus arcticus montanus

) as a subspecies (previously designated as a category 2 species) was not widely accepted, and that the scientific community generally considered this population a geographically isolated member of the wider species (

T. arcticus

).

On July 25, 1994, we published a notice of a 12-month finding in the

Federal Register

(59 FR 37738), concluding that listing the DPS of fluvial Arctic grayling in the upper Missouri River was warranted but precluded by other higher priority listing actions. This DPS determination predated our DPS policy (61 FR 4722, February 7, 1996), so the entity did not undergo a DPS analysis as described in the policy. The 1994 finding placed fluvial Arctic grayling of the upper Missouri River on the candidate list and assigned it a listing priority of 9. On May 4, 2004, we elevated the listing priority number of the fluvial Arctic grayling to 3 (69 FR 24881).

On May 31, 2003, the Center for Biological Diversity and Western Watersheds Project (Plaintiffs) filed a complaint in U.S. District Court in Washington, D.C., challenging our “warranted but precluded” determination for Montana fluvial Arctic grayling. On July 22, 2004, the Plaintiffs amended their complaint to challenge our failure to emergency list this population. We settled with the Plaintiffs in August 2005, and we agreed to submit a final determination on whether this population warranted listing as endangered or threatened to the

Federal Register

on or before April 16, 2007.

On April 24, 2007, we published a revised 12-month finding on the petition to list the upper Missouri River DPS of fluvial Arctic grayling (72 FR 20305) (“2007 finding”). In this finding, we determined that fluvial Arctic grayling of the upper Missouri River did not constitute a species, subspecies, or DPS under the ESA. Therefore, we found that the upper Missouri River population of fluvial Arctic grayling was not a listable entity under the ESA, and as a result, listing was not warranted. With that notice, we withdrew the fluvial Arctic grayling from the candidate list.

On November 15, 2007, the Center for Biological Diversity, Federation of Fly Fishers, Western Watersheds Project, George Wuerthner, and Pat Munday filed a complaint (CV-07-152, in the District Court of Montana) to challenge our 2007 finding. We settled this litigation on October 5, 2009. In the stipulated settlement, we agreed to: (a) Publish, on or before December 31, 2009, a notice in the

Federal Register

soliciting information on the status of the upper Missouri River Arctic grayling; and (b) submit, on or before August 30, 2010, a new 12-month finding for the upper Missouri River Arctic grayling to the

Federal Register

.

On October 28, 2009, we published a notice of intent to conduct a status review of Arctic grayling (

Thymallus arcticus

) in the upper Missouri River system (74 FR 55524). To ensure the status review was based on the best available scientific and commercial data, we requested information on the taxonomy, biology, ecology, genetics, and population status of the Arctic grayling of the upper Missouri River system; information relevant to consideration of the potential DPS status of Arctic grayling of the upper Missouri River system; threats to the species; and conservation actions being implemented to reduce those threats in the upper Missouri River system. The notice further specified that the status review may consider various DPS designations that include different life histories of Arctic grayling in the upper Missouri River system. Specifically, we may consider DPS configurations that include: Fluvial, adfluvial (lake populations), or all life histories of Arctic grayling in the upper Missouri River system.

This notice constitutes the revised 12-month finding (“2010 finding”) on whether to list the upper Missouri River DPS of Arctic grayling (

Thymallus arcticus

) as endangered or threatened.

Taxonomy and Species Description

The Arctic grayling (

Thymallus arcticus

) belongs to the family Salmonidae (salmon, trout, charr, whitefishes), subfamily

Thymallinae

(graylings), and it is represented by a single genus,

Thymallus

. Scott and Crossman (1998, p. 301) recognize four species within the genus:

T. articus

(Arctic grayling),

T. thymallus

(European grayling),

T. brevirostris

(Mongolian grayling), and

T. nigrescens

(Lake Kosgol, Mongolia). Recent research focusing on Eurasian

Thymallus

(Koskinen

et al.

2002, entire; Froufe

et al.

2003, entire; Froufe

et al.

2005, entire; Weiss

et al.

2006, entire) indicates that the systematic diversity of the genus is greater than previously thought, or at least needs better description (Knizhin

et al.

2008, pp. 725-726, 729; Knizhin and Weiss 2009, pp. 1, 7-8; Weiss

et al.

2007, p. 384).

Arctic grayling have elongate, laterally compressed, trout-like bodies with deeply forked tails, and adults typically average 300-380 millimeters (mm) (12-15 inches (in.)) in length. Coloration can be striking, and varies from silvery or iridescent blue and lavender, to dark blue (Behnke 2002, pp. 327-328). The sides are marked with a varying number of V-shaped or diamond-shaped spots (Scott and Crossman 1998, p. 301). During the spawning period, the colors darken and the males become more brilliantly colored than the females. A prominent morphological feature of Arctic grayling is the sail-like dorsal fin, which is large and vividly colored with rows of orange to bright green spots, and often has an orange border (Behnke 2002, pp. 327-328).

Distribution

Arctic grayling are native to Arctic Ocean drainages of Alaska and northwestern Canada, as far east as Hudson's Bay, and westward across northern Eurasia to the Ural Mountains (Scott and Crossman 1998, pp. 301-302; Froufe

et al.

2005, pp. 106-107; Weiss

et al.

2006, pp. 511-512; see Figure 1 below). In North America, they are native to northern Pacific Ocean drainages as far south as the Stikine River in British Columbia (Nelson and Paetz 1991, pp. 253-256; Behnke 2002, pp. 327-331).

BILLING CODE 4310-55-S

EP08SE10.000

FIGURE 1.

Approximate world-wide distribution of Arctic grayling (

Thymallus arcticus

) at the end of the most recent glacial cycle. The Missouri River distribution is based on Kaya (1992, pp. 47-51). The distribution of the extinct Michigan population is based on Vincent (1962, p. 12) and the University of Michigan (2010). The North American distribution in Canada and Alaska is based on Behnke (2002, p. 330) and Scott and Crossman (1998, pp. 301-302). The Eurasian distribution is based on Knizhin (2009, p. 32) and Knizhin (2010, pers. comm.).

Arctic grayling remains widely distributed across its native range, but within North America, the species has experienced range decline or contraction at the southern limits of its distribution. In British Columbia, Canada, populations in the Williston River watershed are designated as a provincial “red list” species, meaning the population is a candidate for further evaluation to determine if it should be granted endangered (facing imminent extirpation or extinction) or threatened status (likely to become endangered) (British Columbia Conservation Data Centre 2010). In Alberta, Canada, Arctic grayling are native to the Athabasca, Peace, and Hay River drainages. In Alberta, the species has undergone a range contraction of about 40 percent, and half of the province's subpopulations have declined in abundance by more than 90 percent (Alberta Sustainable Resource Development (ASRD) 2005, p. iv).

Distribution in the Conterminous United States

Two disjunct groups of Arctic grayling were native to the conterminous United States: One in the upper Missouri River basin in Montana and Wyoming (extant in Montana, see Figure 2), and another in Michigan that was extirpated in the late 1930s (Hubbs and Lagler 1949, p. 44). Michigan grayling formerly occurred in the Otter River of the Lake Superior drainage in northern Michigan and in streams of the lower peninsula of Michigan in both the Lake Michigan and Lake Huron drainages including the Au Sable, Cheboygan, Jordan, Pigeon, and Rifle Rivers (Vincent 1962, p. 12).

Introduced Lake Dwelling Arctic Grayling in the Upper Missouri River

System and western U.S. populations of Arctic grayling have been established in lakes outside their native range in Arizona, Colorado, Idaho, Montana, New Mexico, Utah, Washington, and Wyoming (Vincent 1962, p. 15; Montana Fisheries Information System (MFISH) 2009; NatureServe 2010). Stocking of hatchery grayling in Montana has been particularly extensive, and there are thought to be up to 78 introduced lacustrine (lake-dwelling) populations resulting from these introductions (see Table 1 below). Over three-quarters of these introductions (79.5 percent) were established outside the native geographic range of upper Missouri River grayling, while only 16 (20.5 percent) were established within the watershed boundary of the upper Missouri River system.

EP08SE10.001

BILLING CODE 4310-55-C

FIGURE 2.

Historical (dark grey lines) and current distribution (stars and circled portion of Big Hole River) of native Arctic grayling in the upper Missouri River basin. White bars denote mainstem river dams that are total barriers to upstream passage by fish.

TABLE 1. Introduced Lake-dwelling Populations of Arctic Grayling in Montana. The primary data source for these designations is MFISH (2009).

River Basin

Number of Introduced (Exotic) Populations

a

Outside Native Geographic Range In Montana

Columbia River

23

Middle Missouri River

2

Saskatchewan River

1

Yellowstone River

36

b

Within Watershed Boundary Of Native Geographic Range In Montana

Upper Missouri River

16

Total Exotic Populations

78

a

List of populations does not include lake populations derived from attempts to re-establish fluvial populations in Montana, native adfluvial populations, or genetic reserves of Big Hole River grayling.

b

Many of these populations may not reproduce naturally and are only sustained through repeated stocking (Montana Fish, Wildlife and Parks 2009, entire).

For the purposes of this finding, we are analyzing a petitioned entity that includes, at its maximum extent, populations of Arctic grayling considered native to the upper Missouri River. Introduced populations present in Montana (e.g., Table 1) or elsewhere are not considered as part of the listable entity because we do not consider them to be native populations. Neither the Act nor our implementing regulations expressly address whether introduced populations should be considered part of an entity being evaluated for listing, and no Service policy addresses the issue. Consequently, in our evaluation of whether or not to include introduced populations in the potential listable entity we considered the following: (1) Our interpretation of the intent of the Act with respect to the disposition of native populations, (2) a policy used by the National Marine Fishery Service (NMFS) to evaluate whether hatchery-origin populations warrant inclusion in the listable entity, and (3) a set of guidelines from another organization (International Union for Conservation of Nature and Natural Resources (IUCN)) with specific criteria for evaluating the conservation contribution of introduced populations.

Intent of the Endangered Species Act

The primary purpose of the Act is to provide a means whereby the ecosystems upon which endangered species and threatened species depend may be conserved. The Service has interpreted the Act to provide a statutory directive to conserve species in their native ecosystems (49 FR 33890, August 27, 1984) and to conserve genetic resources and biodiversity over a representative portion of a taxon's historical occurrence (61 FR 4723, February 7, 1996). This priority on natural populations is evident in the Service's DPS policy within the third significance criteria. In that, a discrete population segment may be significant if it represents the only surviving natural occurrence of the taxon that may be more abundant elsewhere as an introduced population outside of its historical range.

National Marine Fishery Service Hatchery Policy

In 2005, the NMFS published a final policy on the consideration of hatchery-origin fish in Endangered Species Act listing determinations for Pacific salmon and steelhead (anadromous

Oncorhynchus

spp.) (NMFS 2005, entire). A central tenet of this policy is the primacy of the conservation of naturally spawning salmon populations and the ecosystems on which they depend, consistent with the intent of the Act (NMFS 2005, pp. 37211, 37214). The policy recognizes that properly managed hatchery programs may provide some conservation benefit to the evolutionary significant unit (ESU, which is analogous to a DPS but applied to Pacific salmon) (NMFS 2005, p. 37211), and that hatchery stocks that contribute to survival and recovery of an ESU are considered during a listing decision (NMFS 2005, p. 37209). The policy states that since hatchery stocks are established and maintained with the intent of furthering the viability of wild populations in the ESU, that those hatchery populations have an explicit conservation value. Genetic divergence is the preferred metric to determine if hatchery fish should be included in the ESU, but NMFS recognizes that these data may be lacking in most cases (NMFS 2005, p. 37209). Thus, proxies for genetic divergence can be used, such as the length of time a stock has been isolated from its source population, the degree to which natural broodstock has been regularly incorporated into the hatchery population, the history of non-ESU fish or eggs in the hatchery population, and the attention given to genetic considerations in selecting and mating broodstocks (NMFS 2005, p. 37209).

The NMFS policy applies to artificially propagated (hatchery) populations. In this finding, however, the Service is deciding whether self-sustaining populations introduced outside its natural range should be included in the listable entity. Thus, the NMFS policy is not directly applicable. Nonetheless, if the NFMS policy's criteria are applied to the introduced lake-dwelling populations of Arctic grayling in Montana and elsewhere, these populations do not appear to warrant inclusion in the entity being evaluated for listing. First, there does not appear to be any formally recognized conservation value for the

introduced populations of Arctic grayling, and they are not being used in restoration programs. Recent genetic analysis indicates that many of the introduced Arctic grayling populations in Montana are derived, in part, from stocks in the Red Rock Lakes system (Peterson and Ardren 2009, p. 1767). Nonetheless, there have been concerns that introduced, lake-dwelling populations could pose genetic risks to the native fluvial population (Arctic Grayling Workgroup (AGW) 1995, p. 15), and in practice, these introduced populations have not been used for any conservation purpose. In fact, efforts are currently underway to establish a genetically pure brood reserve population of Red Rock Lakes grayling to be used for conservation purposes (Jordan 2010, pers. comm.), analogous to the brood reserves maintained for Arctic grayling from the Big Hole River (Rens and Magee 2007, pp. 22-24).

Second, introduced populations in lakes have apparently been isolated from their original source stock for decades without any supplementation from the wild. These populations were apparently established without any formal genetic consideration to selecting and mating broodstock, the source populations were not well documented (Peterson and Ardren 2009, p. 1767), and the primary intent of culturing and introducing these grayling appears to have been to provide recreational fishing opportunities in high mountain lakes.

Guidelines Used in Other Evaluation Systems

The IUCN uses its Red List system to evaluate the conservation status and relative risk of extinction for species, and to catalogue and highlight plant and animal species that are facing a higher risk of global extinction (

http://www.iucnredlist.org

). IUCN does not use the term “listable entity” as the Service does; however, IUCN does clarify that their conservation ranking criteria apply to any taxonomic group at the species level or below (IUCN 2001, p.4). Further, the IUCN guidelines for species status and scope of the categorization process focus on wild populations inside their natural range (IUCN 2001, p. 4; 2003, p. 10) or so-called “benign” or “conservation introductions,” which are defined as attempts to establish a species, for the purpose of conservation, outside its recorded distribution, when suitable habitat is lacking within the historical range (IUCN 1998, p. 6; 2003, pp. 6, 10). Guidelines for evaluating conservation status under the IUCN exclude introduced populations located outside the recorded distribution of the species if such populations were established for commercial or sporting purposes (IUCN 1998, p. 5; 2003, p. 24). In effect, the IUCN delineates between introduced and native populations in that non-benign introductions do not qualify for evaluation under the IUCN Red List system. Naturalized populations of Arctic grayling in lakes thus do not meet the IUCN criterion for a wild population that should be considered when evaluating the species status for two reasons. First, there remains ‘suitable habitat' for Arctic grayling in its native range, as evidenced by extant native populations in the Big Hole River, Madison River, Miner Lake, Mussigbrod Lake, and Red Rock Lakes. Second, the naturalized populations derived from widespread stocking were apparently aimed at establishing recreational fisheries.

Our interpretation is that the ESA is intended to preserve native populations in their ecosystems. While hatchery or introduced populations of fishes may have some conservation value, this does not appear to be the case with introduced populations of Arctic grayling in the conterminous United States. These populations were apparently established to support recreational fisheries, and without any formal genetic consideration to selecting and mating broodstock, and are not part of any conservation program to benefit the native populations. Consequently, we do not consider the introduced populations of Arctic grayling in Montana and elsewhere in the conterminous United States, including those in lakes and in an irrigation canal (Sun River Slope Canal), to be part of the listable entity.

Native Distribution in the Upper Missouri River System

The first Euro-American “discovery” of Arctic grayling in North America is attributed to members of the Lewis and Clark Expedition, who encountered the species in the Beaverhead River in August 1805 (Nell and Taylor 1996, p. 133). Vincent (1962, p. 11) and Kaya (1992, pp. 47-51) synthesized accounts of Arctic grayling occurrence and abundance from historical surveys and contemporary monitoring to determine the historical distribution of the species in the upper Missouri River system (Figure 2). We base our conclusions on the historical distribution of Arctic grayling in the upper Missouri River basin on these two reviews. Arctic grayling were widely but irregularly distributed in the upper Missouri River system above the Great Falls in Montana and in northwest Wyoming within the present-day location of Yellowstone National Park (Vincent 1962, p. 11). They were estimated to inhabit up to 2,000 kilometers (km) (1,250 miles (mi)) of stream habitat until the early 20th century (Kaya 1992, pp. 47-51). Arctic grayling were reported in the mainstem Missouri River, as well as in the Smith, Sun, Jefferson, Madison, Gallatin, Big Hole, Beaverhead, and Red Rock Rivers (Vincent 1962, p. 11; Kaya 1992, pp. 47-51; USFWS 2007; 72 FR 20307, April 24, 2007). “Old-timer” accounts report that the species may have been present in the Ruby River, at least seasonally (Magee 2005, pers. comm.), and were observed as recently as the early 1970s (Holton, undated).

Fluvial Arctic grayling were historically widely distributed in the upper Missouri River basin, but a few adfluvial populations also were native to the basin. For example, Arctic grayling are native to Red Rock Lakes, in the headwaters of the Beaverhead River (Vincent 1962, pp. 112-121; Kaya 1992, p. 47). Vincent (1962, p. 120) stated that Red Rock Lakes were the only natural lakes in the upper Missouri River basin accessible to colonization by Arctic grayling, and concluded that grayling there were the only native adfluvial population in the basin. However, it appears that Arctic grayling also were native to Elk Lake (in the Red Rocks drainage; Kaya 1990, p. 44) and a few small lakes in the upper Big Hole River drainage (Peterson and Ardren 2009, p. 1768).

The distribution of native Arctic grayling in the upper Missouri River went through a dramatic reduction in the first 50 years of the 20th century, especially in riverine habitats (Vincent 1962, pp. 86-90, 97-122, 127-129; Kaya 1992, pp. 47-53). The native populations that formerly resided in the Smith, Sun, Jefferson, Beaverhead, Gallatin, and mainstem Missouri Rivers are considered extirpated, and the only remaining indigenous fluvial population is found in the Big Hole River and some if its tributaries (Kaya 1992, pp. 51-53). The fluvial form currently occupies only 4 to 5 percent of its historic range in the Missouri River system (Kaya 1992, p. 51). Other remaining native populations in the upper Missouri River occur in two small, headwater lakes in the upper Big Hole River system (Miner and Mussigbrod Lakes); the Madison River upstream from Ennis Reservoir; and the Red Rock Lakes in the headwaters of the Beaverhead River system (Everett 1986, p. 7; Kaya 1992, p. 53; Peterson and Ardren 2009, pp. 1762, 1768; Figure 1 above, and Table 2 below).

TABLE 2. Extant Native Arctic Grayling Populations in the Upper Missouri River Basin.

Big Hole River Drainage

a

1. Big Hole River

2. Miner Lake

3. Mussigbrod Lake

Madison River Drainage

4. Madison River-Ennis Reservoir

Beaverhead River Drainage

5. Red Rock Lakes

a

Arctic grayling also occur in Pintler Lake in the Big Hole River drainage, but this population has not been evaluated with genetic markers to determine whether it constitutes a native remnant population.

Origins, Biogeography, and Genetics of Arctic Grayling in North America

North American Arctic grayling are most likely descended from Eurasian

Thymallus

that crossed the Bering land bridge during or before the Pleistocene glacial period (Stamford and Taylor 2004, pp. 1533, 1546). A Eurasian origin is suggested by the substantial taxonomic diversity found in the genus in that region. There were multiple opportunities for freshwater faunal exchange between North America and Asia during the Pleistocene, but genetic divergence between North American and Eurasian Arctic grayling suggests that the species could have colonized North America as early as the mid-late Pliocene (more than 3 million years ago) (Stamford and Taylor 2004, p. 1546).

The North American distribution of Arctic grayling was strongly influenced by patterns of glaciation. Genetic studies of grayling using mitochondrial DNA (mtDNA, maternally-inherited DNA located in cellular organelles called mitochondria) and microsatellite DNA (repeating sequences of nuclear DNA) have shown that North American Arctic grayling consist of at least three major lineages that originated in distinct Pleistocene glacial refugia (Stamford and Taylor 2004, p. 1533). These three groups include a South Beringia lineage found in western Alaska to northern British Columbia, Canada; a North Beringia lineage found on the North Slope of Alaska, the lower Mackenzie River, and to eastern Saskatchewan; and a Nahanni lineage found in the lower Liard River and the upper Mackenzie River drainage (Stamford and Taylor 2004, pp. 1533, 1540). The Nahanni lineage is the most genetically distinct group (Stamford and Taylor 2004, pp. 1541-1543). Arctic grayling from the upper Missouri River basin were tentatively placed in the North Beringia lineage because a small sample (three individuals) of Montana grayling shared a mtDNA haplotype (form of the mtDNA) with populations in Saskatchewan and the lower Peace River, British Columbia (Stamford and Taylor 2004, p. 1538).

The existing mtDNA data suggest that Missouri River Arctic grayling share a common ancestry with the North Beringia lineage, but other genetic markers and biogeographic history indicate that Missouri River grayling have been physically and reproductively isolated from northern populations for millennia. The most recent ancestors of Missouri River Arctic grayling likely spent the last glacial cycle in an ice-free refuge south of the Laurentide and Cordilleran ice sheets. Pre-glacial colonization of the Missouri River basin by Arctic grayling was possible because the river flowed to the north and drained into the Arctic-Hudson Bay prior to the last glacial cycle (Cross

et al.

1986, pp. 374-375; Pielou 1991, pp. 194-195). Low mtDNA diversity observed in a small number of Montana grayling samples and a shared ancestry with Arctic grayling from the north Beringia lineage suggest a more recent, post-glacial colonization of the upper Missouri River basin. In contrast, microsatellite DNA show substantial divergence between Montana and Saskatchewan (i.e., same putative mtDNA lineage) (Peterson and Ardren 2009, entire). Differences in the frequency and size distribution of microsatellite alleles between Montana populations and two Saskatchewan populations indicate that Montana grayling have been isolated long enough for mutations (i.e., evolution) to be responsible for the observed genetic differences.

Additional comparison of 21 Arctic grayling populations from Alaska, Canada, and the Missouri River basin using 9 of the same microsatellite loci as Peterson and Ardren (2009, entire) further supports the distinction of Missouri River Arctic grayling relative to populations elsewhere in North America (USFWS, unpublished data). Analyses of these data using two different methods clearly separates sample fish from 21 populations into two clusters: one cluster representing populations from the upper Missouri River basin, and another cluster representing populations from Canada and Alaska (USFWS, unpublished data). These new data, although not yet peer reviewed, support the interpretation that the previous analyses of Stamford and Taylor (2004, entire) underestimated the distinctiveness of Missouri River Arctic grayling relative to other sample populations, likely because of the combined effect of small sample sizes and the lack of variation observed in the Missouri River for the markers used in that study (Stamford and Taylor 2004, pp. 1537-1538). Thus, these recent microsatellite DNA data suggest that Arctic grayling may have colonized the Missouri River before the onset of Wisconsin glaciation (more than 80,000 years ago).

Genetic relationships among native and introduced populations of Arctic grayling in Montana have recently been investigated (Peterson and Ardren 2009, entire). Introduced, lake-dwelling populations of Arctic grayling trace much of their original ancestry to Red Rock Lakes (Peterson and Ardren 2009, p. 1767), and stocking of hatchery grayling did not appear to have a large effect on the genetic composition of the extant native populations (Peterson and Ardren 2009, p. 1768). Differences between native populations of the two grayling ecotypes (adfluvial, fluvial) do not appear to be as large as differences resulting from geography (i.e., drainage of origin).

Habitat

Arctic grayling generally require clear, cold water. Selong

et al.

(2001, p. 1032) characterized Arctic grayling as belonging to a “coldwater” group of salmonids, which also includes bull trout (

Salvelinus confluentus

) and Arctic char (

Salvelinus alpinus

). Hubert

et al.

(1985, p. 24) developed a habitat suitability index study for Arctic grayling and concluded that thermal habitat was optimal between 7 to 17 °C (45 to 63 °F), but became unsuitable above 20°C (68°F). Arctic grayling fry may be more tolerant of high water temperature than adults (LaPerriere and Carlson 1973, p. 30; Feldmeth and Eriksen 1978, p. 2041).

Having a broad, nearly-circumpolar distribution, Arctic grayling occupy a variety of habitats including small streams, large rivers, lakes, and even bogs (Northcote 1995, pp. 152-153; Scott and Crossman 1998, p. 303). They may even enter brackish water (less than or equal to 4 parts per thousand) when migrating between adjacent river systems (West

et al.

1992, pp. 713-714). Native populations are found at elevations ranging from near sea level, such as in Bristol Bay, Alaska, to high-elevation montane valleys (more than 1,830 meters (m) or 6,000 feet (ft)), such as the Big Hole River and Centennial Valley in southwestern Montana. Despite this broad distribution, Arctic grayling have specific habitat requirements that can constrain their local distributions, especially water temperature and channel gradient. At the local scale, Arctic grayling prefer cold water and are often associated with spring-fed habitats in regions with warmer climates (Vincent 1962, p. 33). Arctic grayling are generally not found in swift, high-gradient streams, and Vincent (1962, p. 36-37, 41-43) characterized typical Arctic grayling habitat in Montana (and Michigan) as low-to-moderate gradient (less than 4 percent) streams and rivers with low-to-moderate water velocities (less than 60 centimeters/sec). Juvenile and adult Arctic grayling in streams and rivers spend much of their time in pool habitat (Kaya 1990 and references therein, p. 20; Lamothe and Magee 2003, pp. 13-14).

Breeding

Arctic grayling typically spawn in the spring or early summer, depending on latitude and elevation (Northcote 1995, p. 149). In Montana, Arctic grayling generally spawn from late April to mid-May by depositing adhesive eggs over gravel substrate without excavating a nest (Kaya 1990, p. 13; Northcote 1995, p. 151). In general, the reproductive ecology of Arctic grayling differs from other salmonid species (trout and salmon) in that Arctic grayling eggs tend to be comparatively small; thus, they have higher relative fecundity (females have more eggs per unit body size). Males establish and defend spawning territories rather than defending access to females (Northcote 1995, pp. 146, 150-151). The time required for development of eggs from embryo until they emerge from stream gravel and become swim-up fry depends on water temperature (Northcote 1995, p. 151). In the upper Missouri River basin, development from embryo to fry averages about 3 weeks (Kaya 1990, pp. 16-17). Small, weakly swimming fry (typically 1-1.5 centimeters (cm) (0.4-0.6 in.) at emergence) prefer low-velocity stream habitats (Armstrong 1986, p. 6; Kaya 1990, pp. 23-24; Northcote 1995, p. 151).

Arctic grayling of all ages feed primarily on aquatic and terrestrial invertebrates captured on or near the water surface, but also will feed opportunistically on fish and fish eggs (Northcote 1995, pp. 153-154; Behnke 2002, p. 328). Feeding locations for individual fish are typically established and maintained through size-mediated dominance hierarchies where larger individuals defend favorable feeding positions (Hughes 1992, p. 1996).

Life History Diversity

Migratory behavior is a common life-history trait in salmonid fishes such as Arctic grayling (Armstrong 1986, pp. 7-8; Northcote 1995, pp. 156-158; 1997, pp. 1029, 1031-1032, 1034). In general, migratory behavior in Arctic grayling and other salmonids results in cyclic patterns of movement between refuge, rearing-feeding, and spawning habitats (Northcote 1997, p. 1029).

Arctic grayling may move to refuge habitat as part of a regular seasonal migration (e.g., in winter), or in response to episodic environmental stressors (e.g., high summer water temperatures). In Alaska, Arctic grayling in rivers typically migrate downstream in the fall, moving into larger streams or mainstem rivers that do not completely freeze (Armstrong 1986, p. 7). In Arctic rivers, fish often seek overwintering habitat influenced by groundwater (Armstrong 1986, p. 7). In some drainages, individual fish may migrate considerable distances (greater than 150 km or 90 mi) to overwintering habitats (Armstrong 1986, p. 7). In the Big Hole River, Montana, similar downstream and long-distance movement to overwintering habitat has been observed in Arctic grayling (Shepard and Oswald 1989, pp. 18-21, 27). In addition, Arctic grayling in the Big Hole River may move downstream in proximity to colder tributary streams in summer when thermal conditions in the mainstem river become stressful (Lamothe and Magee 2003, p. 17).

In spring, mature Arctic grayling leave overwintering areas and migrate to suitable spawning sites. In river systems, this typically involves an upstream migration to tributary streams or shallow riffles within the mainstem (Armstrong 1986, p. 8). Arctic grayling in lakes typically migrate to either the inlet or outlet to spawn (Armstrong 1986, p. 8; Northcote 1997, p. 148). In either situation, Arctic grayling typically exhibit natal homing, whereby individuals spawn in or near the location where they were born (Northcote 1997, pp. 157-160).

Fry from river populations typically seek feeding and rearing habitats in the vicinity where they were spawned (Armstrong 1986, pp. 6-7; Northcote 1995, p. 156), while those from lake populations migrate downstream (inlet spawners) or upstream (outlet spawners) to the adjacent lake. Following spawning, adults move to appropriate feeding areas if they are not adjacent to spawning habitat (Armstrong 1986, pp. 7-8). Juvenile Arctic grayling may undertake seasonal migrations between feeding and overwintering habitats until they reach maturity and add the spawning migration to this cycle (Northcote 1995, pp. 156-157).

Life History Diversity in Arctic Grayling in the Upper Missouri River

Two general life-history forms or ecotypes of native Arctic grayling occur in the upper Missouri River Arctic: Fluvial and adfluvial. Fluvial fish use river or stream (lotic) habitat for all of their life cycles and may undergo extensive migrations within river habitat. Adfluvial fish live in lakes and migrate to tributary streams to spawn. These same life-history forms also are expressed by Arctic grayling elsewhere in North America (Northcote 1997, p. 1030). Historically, the fluvial life-history form predominated in the Missouri River basin above the Great Falls, perhaps because there were only a few lakes accessible to natural colonization of Arctic grayling that would permit expression of the adfluvial ecotype (Kaya 1992, p. 47). The fluvial and adfluvial life-history forms of Arctic grayling in the upper Missouri River do not appear to represent distinct evolutionary lineages. Instead, they appear to represent an example of adaptive radiation (Schluter 2000, p. 1), whereby the forms

differentiated from a common ancestor developed traits that allowed them to exploit different habitats. The primary evidence for this conclusion is genetic data that indicate that within the Missouri River basin the two ecotypes are more closely related to each other than they are to the same ecotype elsewhere in North America (Redenbach and Taylor 1999, pp. 27-28; Stamford and Taylor 2004, p. 1538; Peterson and Ardren 2009, p. 1766). Historically, there may have been some genetic exchange between the two life-history forms as individuals strayed or dispersed into different populations (Peterson and Ardren 2009, p. 1770), but the genetic structure of current populations in the upper Missouri River basin is consistent with reproductive isolation.

The fluvial and adfluvial forms of Arctic grayling appear to differ in their genetic characteristics, but there appears to be some plasticity in behavior where individuals from a population can exhibit a range of behaviors. Arctic grayling fry in Montana can exhibit heritable, genetically-based differences in swimming behavior between fluvial and adfluvial ecotypes (Kaya 1991, pp. 53, 56-58; Kaya and Jeanes 1995, pp. 454, 456). Progeny of Arctic grayling from the fluvial ecotype exhibited a greater tendency to hold their position in flowing water relative to progeny from adfluvial ecotypes (Kaya 1991, pp. 53, 56-58; Kaya and Jeanes 1995, pp. 454, 456). Similarly, young grayling from inlet and outlet spawning adfluvial ecotypes exhibited an innate tendency to move downstream and upstream, respectively (Kaya 1989, pp. 478-480). All three studies (Kaya 1989, entire; 1991, entire; Kaya and Jeanes 1995, entire) demonstrate that the response of fry to flowing water depended strongly on the life-history form (ecotype) of the source population, and that this behavior has a genetic basis. However, behavioral responses also were mediated by environmental conditions (light—Kaya 1991, pp. 56-57; light and water temperature—Kaya 1989, pp. 477-479), and some progeny of each ecotype exhibited behavior characteristic of the other; for example some individuals from the fluvial ecotype moved downstream rather than holding position, and some individuals from an inlet-spawning adfluvial ecotype held position or moved upstream (Kaya 1991, p. 58). These observations indicate that some plasticity for behavior exists, at least for very young Arctic grayling.

However, the ability of one ecotype of Arctic grayling to give rise to a functional population of the other ecotype within a few decades is much less certain, and may parallel the differences in plasticity that have evolved between river- and lake-type European grayling (Salonen 2005, entire). Circumstantial support for reduced plasticity in adfluvial Arctic grayling comes from observations that adfluvial fish stocked in river habitats almost never establish populations (Kaya 1990, pp. 31-34). In contrast, a population of Arctic grayling in the Madison River that would have presumably expressed a fluvial ecotype under historical conditions has apparently adapted to an adfluvial life-history after construction of an impassible dam, which impounded Ennis Reservoir (Kaya 1992, p. 53; Jeanes 1996, pp. 54). We note that adfluvial Arctic grayling retain some life-history flexibility—at least in lake environments—as naturalized populations derived from inlet-spawning stocks have established outlet-spawning demes (a deme is a local populations that shares a distinct gene pool) in Montana and in Yellowstone National Park (Kruse 1959, p. 318; Kaya 1989, p. 480). While in some cases Arctic grayling may be able to adapt or adjust rapidly to a new environment, the frequent failure of introductions of Arctic grayling suggest a cautionary approach to the loss of particular life-history forms is warranted. Healey and Prince (1995, entire) reviewed patterns of genotypic and phenotypic variation in Pacific salmon and warn that recovery of lost life-history forms may not follow directly from conservation of the genotype (p. 181), and reason that the critical conservation unit is the population within its habitat (p. 181).

Age and Growth

Age at maturity and longevity in Arctic grayling varies regionally and is probably related to growth rate, with populations in colder, northern latitudes maturing at later ages and having a greater lifespan (Kruse 1959, pp. 340-341; Northcote 1995 and references therein, pp. 155-157). Arctic grayling in the upper Missouri River typically mature at age 2 (males) or age 3 (females), and individuals greater than age 6 are rare (Kaya 1990, p. 18; Magee and Lamothe 2003, pp. 16-17). Similarly, Nelson (1954, pp. 333-334) observed that the majority of the Arctic grayling spawning in two tributaries in the Red Rock Lakes system, Montana, were age 3, and the oldest individuals aged from a larger sample were age 6. Mogen (1996, pp. 32-34) found that Arctic grayling spawning in Red Rock Creek were mostly ages 2 to 5, but he did encounter some individuals age 7.

Generally, growth rates of Arctic grayling are greatest during the first years of life then slow dramatically after maturity. Within that general pattern, there is substantial variation among populations from different regions. Arctic grayling populations in Montana (Big Hole River and Red Rock Lakes) appear to have very high growth rates relative to those from British Columbia, Asia, and the interior and North Slope of Alaska (Carl

et al.

1992, p. 240; Northcote 1995, pp. 155-157; Neyme 2005, p. 28). Growth rates of Arctic grayling from different management areas in Alberta are nearly as high as those observed in Montana grayling (ASRD 2005, p. 4).

Distinct Population Segment

In its stipulated settlement with Plaintiffs, the Service agreed to consider the appropriateness of DPS designations for Arctic grayling populations in the upper Missouri River basin that included: (a) All life ecotypes or histories, (b) the fluvial ecotype, and (c) the adfluvial ecotype. The fluvial ecotype has been the primary focus of past Service action and litigation, but the Service also has alluded to the possibility of alternative DPS designations in previous candidate species assessments (USFWS 2005, p. 11). Since the 2007 finding (72 FR 20305), additional research has been conducted and new information on the genetics of Arctic grayling is available. This finding contains a more comprehensive and robust distinct population segment analysis than the 2007 finding.

Distinct Population Segment Analysis for Native Arctic Graying in the Upper Missouri River

Discreteness

The discreteness standard under the Service's and National Oceanic and Atmospheric Administration's (NOAA) joint Policy Regarding the Recognition of Distinct Vertebrate Population Segments Under the Endangered Species Act (61 FR 4722) requires an entity to be adequately defined and described in some way that distinguishes it from other representatives of its species. A segment is discrete if it is: (1) Markedly separated from other populations of the same taxon as consequence of physical, physiological, ecological, or behavioral factors (quantitative measures of genetic or morphological discontinuity may provide evidence of this separation); or (2) delimited by international

governmental boundaries within which differences in control of exploitation, management of habitat, conservation status, or regulatory mechanisms exist that are significant in light of section 4(a)(1)(D) of the ESA.

Arctic grayling native to the upper Missouri River are isolated from populations of the species inhabiting the Arctic Ocean, Hudson Bay, and north Pacific Ocean drainages in Asia and North America (see Figure 1). Arctic grayling native to the upper Missouri River occur as a disjunct group of populations approximately 800 km (500 mi) to the south of the next-nearest Arctic grayling population in central Alberta, Canada. Missouri River Arctic grayling have been isolated from other populations for at least 10,000 years based on historical reconstruction of river flows at or near the end of the Pleistocene (Cross

et al.

1986, p. 375; Pileou 1991, pp. 10-11;). Genetic data confirm Arctic grayling in the Missouri River basin have been reproductively isolated from populations to the north for millennia (Everett 1986, pp. 79-80; Redenbach and Taylor 1999, p. 23; Stamford and Taylor 2004, p. 1538; Peterson and Ardren 2009, pp. 1764-1766; USFWS, unpublished data). Consequently, we conclude that Arctic grayling native to the upper Missouri River are markedly separated from other native populations of the taxon as a result of physical factors (isolation), and therefore meet the first criterion of discreteness under the DPS policy. As a result, Arctic grayling native to the upper Missouri River are considered a discrete population according to the DPS policy. Because the entity meets the first criterion (markedly separated), an evaluation with respect to the second criterion (international boundaries) is not needed.

Significance

If we determine that a population meets the DPS discreteness element, we then consider whether it also meets the DPS significance element. The DPS policy states that, if a population segment is considered discrete under one or more of the discreteness criteria, its biological and ecological significance will be considered in light of congressional guidance that the authority to list DPSs be used “sparingly” while encouraging the conservation of genetic diversity (see U.S. Congress 1979, Senate Report 151, 96

th

Congress, 1st Session). In making this determination, we consider available scientific evidence of the discrete population's importance to the taxon to which it belongs. Since precise circumstances are likely to vary considerably from case to case, the DPS policy does not describe all the classes of information that might be used in determining the biological and ecological importance of a discrete population. However, the DPS policy does provide four possible reasons why a discrete population may be significant. As specified in the DPS policy, this consideration of significance may include, but is not limited to, the following: (1) Persistence of the discrete population segment in a unique or unusual ecological setting; (2) evidence that loss of the discrete segment would result in a significant gap in the range of the taxon; (3) evidence that the discrete population segment represents the only surviving natural occurrence of the taxon that may be more abundant elsewhere as an introduced population outside of its historic range; or (4) evidence that the discrete population segment differs markedly from other populations of the species in its genetic characteristics.

Unique Ecological Setting

Water temperature is a key factor influencing the ecology and physiology of ectothermic (body temperature regulated by ambient environmental conditions) salmonid fishes, and can dictate reproductive timing, growth and development, and life-history strategies. Groundwater temperatures can be related to air temperatures (Meisner 1990, p. 282), and thus reflect the regional climatic conditions. Warmer groundwater influences ecological factors such as food availability, the efficiency with which food is converted into energy for growth and reproduction, and ultimately growth rates of aquatic organisms (Allan 1995, pp. 73-79). Aquifer structure and groundwater temperature is important to salmonid fishes because groundwater can strongly influence stream temperature, and consequently egg incubation and fry growth rates, which are strongly temperature-dependent (Coutant 1999, pp. 32-52; Quinn 2005, pp. 143-150).

Missouri River Arctic grayling occur within the 4 to 7 °C (39 to 45 °F) ground water isotherm (see Heath 1983, p. 71; an isotherm is a line connecting bands of similar temperatures on the earth's surface), whereas most other North American grayling are found in isotherms less than 4 °C, and much of the species' range is found in areas with discontinuous or continuous permafrost (Meisner

et al.

1988, p. 5). Much of the historical range of Arctic grayling in the upper Missouri River is encompassed by mean annual air temperature isotherms of 5 to 10 °C (41 to 50 °F) (USGS 2009), with the colder areas being in the headwaters of the Madison River in Yellowstone National Park. In contrast, Arctic grayling in Canada, Alaska, and Asia are located in regions encompassed by air temperature isotherms 5 °C and colder (41 °F and colder), with much of the species distributed within the 0 to -10 °C isolines (32 to 14 °F). This difference is significant because Arctic grayling in the Missouri River basin have evolved in isolation for millennia in a generally warmer climate than other populations. The potential for thermal adaptations makes Missouri River Arctic grayling a significant biological resource for the species under expected climate change scenarios.

TABLE 3. Differences Between the Ecological Setting of the Upper Missouri River and Elsewhere in the Species' Range of Arctic Grayling.

Ecological Setting Variable

Missouri River

Rest of Taxon

Ocean watershed

Gulf of Mexico-Atlantic Ocean

Hudson Bay, Arctic Ocean, or north Pacific

Bailey's Ecoregion

Dry Domain: Temperate Steppe

Polar Domain: Tundra & Subarctic Humid Temperate: Marine,

Prairie, Warm Continental

Mountains

Air temperature (isotherm)

5 to 10 °C

(41 to 50 °F)

-15 to 5 °C

(5 to 41 °F)

Groundwater temperature (isotherm)

4 to 7°C

(39 to 45 °F)

Less than 4 °C

(less than 39 °F)

Native occurrence of large-bodied fish predators on salmonids

None, in most of the range

a

Bull trout, lake trout, northern pike, taimen

a

Lake trout are native to two small lakes in the upper Missouri River basin (Twin Lakes and Elk Lake), where their distributions presumably overlapped with the native range of Arctic grayling, so they would not have interacted with most Arctic grayling populations in the basin that were found in rivers.

Arctic grayling in the upper Missouri River basin occur in a temperate ecoregion distinct from all other Arctic grayling populations worldwide, which occur in Arctic or sub-Arctic ecoregions dominated by Arctic flora and fauna. An ecoregion is a continuous geographic area within which there are associations of interacting biotic and abiotic features (Bailey 2005, pp. S14, S23). These ecoregions delimit large areas within which local ecosystems recur more or less in a predictable fashion on similar sites (Bailey 2005, p. S14). Ecoregional classification is hierarchical, and based on the study of spatial coincidences, patterning, and relationships of climate, vegetation, soil, and landform (Bailey 2005, p. S23). The largest ecoregion categories are domains, which represent subcontinental areas of similar climate (e.g., polar, humid temperate, dry, and humid tropical) (Bailey 1994; 2005, p. S17). Domains are divided into divisions that contain areas of similar vegetation and regional climates. Arctic grayling in the upper Missouri River basin are the only example of the species naturally occurring in a dry domain (temperate steppe division; see Table 3 above). The vast majority of the species' range is found in the polar domain (all of Asia, most of North America), with small portions of the range occurring in the humid temperate domain (northern British Columbia and southeast Alaska). Occupancy of Missouri River Arctic grayling in a temperate ecoregion is significant for two primary reasons. First, an ecoregion represents a suite of factors (climate, vegetation, landform) influencing, or potentially influencing, the evolution of species within that ecoregion. Since Missouri River Arctic grayling have existed for thousands of years in an ecoregion quite different from the majority of the taxon, they have likely developed adaptations during these evolutionary timescales that distinguish them from the rest of the taxon, even if we have yet to conduct the proper studies to measure these adaptations. Second, the occurrence of Missouri River Arctic grayling in a unique ecoregion helps reduce the risk of species-level extinction, as the different regions may respond differently to environmental change.

Arctic grayling in the upper Missouri River basin have existed for at least 10,000 years in an ecological setting quite different from that experienced by Arctic grayling elsewhere in the species' range. The most salient aspects of this different setting relate to temperature and climate, which can strongly and directly influence the biology of ectothermic species (like Arctic grayling). Arctic grayling in the upper Missouri River have experienced warmer temperatures than most other populations. Physiological and life-history adaptation to local temperature regimes are regularly documented in salmonid fishes (Taylor 1991, pp. 191-193), but experimental evidence for adaptations to temperature, such as unusually high temperature tolerance or lower tolerance to colder temperatures, is lacking for Missouri River Arctic grayling because the appropriate studies have not been conducted. Lohr

et al.

(1996, p. 934) studied the upper thermal tolerances of Arctic grayling from the Big Hole River, but their research design did not include other populations from different thermal regimes, so it was not possible to make between-population contrasts under a common set of conditions. Arctic grayling from the upper Missouri River demonstrate very high growth rates relative to other populations (Northcote 1995, p. 157). Experimental evidence obtained by growing fish from populations under similar conditions would be needed to measure the relative influence of genetics (local adaptation) versus environment.

An apex fish predator that preys successfully on salmonids has been largely absent from most of the upper Missouri River basin over evolutionary time scales (tens of thousands of years). This suggests that Arctic grayling in the upper Missouri River basin have faced a different selective pressure than Arctic grayling in many other areas of the species' range, at least with respect to predation by fishes. Predators can exert a strong selective pressure on populations. One noteworthy aspect of the aquatic biota experienced by Arctic grayling in the upper Missouri River is the apparent absence of a large-bodied fish that would be an effective predator on juvenile and adult salmonids. In contrast, one or more species of large predatory fishes like northern pike (

Esox lucius

), bull trout, taimen (

Hucho taimen

), and lake trout (

Salvelinus namaycush

) are broadly distributed across much of the range of Arctic grayling in Canada and Asia (Northern pike—Scott and Crossman 1998, pp. 302, 358; taimen—VanderZanden

et al.

2007, pp. 2281-2282; Esteve

et al.

2009, p. 185; bull trout—Behnke 2002, pp. 296, 330; lake trout —Behnke 2002, pp. 296, 330). The only exceptions to this general pattern are where Arctic grayling formerly coexisted with lake trout native to Twin Lakes and Elk Lake (Beaverhead County) (Vincent 1963, pp. 188-189), but both of these Arctic grayling populations are thought to be extirpated (Oswald 2000, pp. 10, 16; Oswald 2006, pers. comm.). The burbot (

Lota lota

) is a freshwater fish belonging to the cod family and is native to the Missouri, Big Hole, Beaverhead, Ruby, and Madison Rivers in Montana (MFISH 2010); thus its distribution significantly overlapped the historical and current ranges of Arctic grayling in the upper Missouri River system. Burbot are voracious predators, but tend to be benthic (bottom-oriented) and apparently prefer the deeper portions of larger rivers and lakes. A few studies have investigated the diet of burbot where they overlap with native Arctic grayling in Montana, but did not detect any predation on Arctic grayling (Streu 1990, pp. 16-20; Katzman 1998, pp. 98-100). Burbot apparently do not consume salmonids in significant amounts, even when they are very abundant (Katzman 1998 and references therein, p. 106). The response of Arctic grayling in the Missouri River basin to introduced,

nonnative trout suggests they were not generally pre-adapted to cope with the presence of a large-bodied salmonid predator. Missouri River Arctic grayling lack a co-evolutionary history with brown trout, and there are repeated observations that the two species tend not to coexist and that brown trout displace Arctic grayling (Kaya 1992, p. 56; 2000, pp. 14-15). We caution that competition with and predation by brown trout has not been directly studied with Arctic grayling, but at least some circumstantial evidence indicates that Missouri River Arctic grayling may not coexist well with brown trout.

We conclude that the occurrence of Arctic grayling in the upper Missouri River is biogeographically important to the species, that grayling there have occupied a distinctly different ecological setting relative to the rest of the species (see Table 3 above), and that they have been on a different evolutionary trajectory for at least 10,000 years. Consequently, we believe that Arctic grayling in the upper Missouri River occupy a unique ecological setting. The role that this unique setting plays in influencing adaptations or determining unique traits is unclear, and therefore a determination of the significance of this ecological setting to the taxon is unknown.

Gap in the Range

Arctic grayling in the upper Missouri River basin occur in an ocean drainage basin that is distinct from all other Arctic grayling populations worldwide. All other Arctic grayling occur in drainages of Hudson Bay, the Arctic Ocean, or the north Pacific Ocean; the Missouri River is part of the Gulf of Mexico-Atlantic Ocean drainage. The significance of occupancy of this drainage basin is that the upper Missouri River basin represents an important part of the species' range from a biogeographic perspective. The only other population of Arctic grayling to live in a non-Arctic environment was the Michigan-Great Lakes population that was extirpated in the 1930s.

Arctic grayling in Montana (southern extent is approximately 44°36′23″ N latitude) represent the southern-most extant population of the species' distribution since the Pleistocene glaciation (Figure 1). The next-closest native Arctic grayling population outside the Missouri River basin is found in the Pembina River (approximately 52°55′6.77″ N latitude) in central Alberta, Canada, west of Edmonton (Blackburn and Johnson 2004, pp. ii, 17; ASRD 2005, p. 6). Loss of the native Arctic grayling of the upper Missouri River would shift the southern distribution of Arctic grayling by more than 8° latitude. Such a dramatic range constriction would constitute a significant geographic gap in the species' range, and eliminate a genetically distinct group of Arctic grayling, which may limit the species' ability to cope with future environmental change.

Marginal populations, defined as those on the periphery of the species' range, are believed to have high conservation significance (see reviews by Scudder 1989, entire; Lesica and Allendorf 1995, entire; Fraser 2000, entire). Peripheral populations may occur in suboptimal habitats and thus be subjected to very strong selective pressures (Fraser 2000, p. 50). Consequently, individuals from these populations may contain adaptations that may be important to the taxon in the future. Lomolino and Channell (1998, p. 482) hypothesize that because peripheral populations should be adapted to a greater variety of environmental conditions, then they may be better suited to deal with anthropogenic (human-caused) disturbances than populations in the central part of a species' range. Arctic grayling in the upper Missouri River have, for millennia, existed in a climate warmer than that experienced by the rest of the taxon. If this selective pressure has resulted in adaptations to cope with increased water temperatures, then the population segment may contain genetic resources important to the taxon. For example, if northern populations of Arctic grayling are less suited to cope with increased water temperatures expected under climate warming, then Missouri River Arctic grayling might represent an important population for reintroduction in those northern regions. We believe that Arctic grayling from the upper Missouri River's occurrence at the southernmost extreme of the range contributes to its significance that may increased adaptability and contribute to the resilience of the overall taxon.

Only Surviving Natural Occurrence of the Taxon that May be More Abundant Elsewhere as an Introduced Population Outside of its Historical Range

This criterion does not directly apply to the Arctic grayling in the upper Missouri River because it is not the only surviving natural occurrence of the taxon; there are native Arctic grayling populations in Canada, Alaska, and Asia. That said, there are introduced Lake Dwelling Arctic Grayling within the native range in the Upper Missouri River System and Arctic grayling have been established in lakes outside their native range in Arizona, Colorado, Idaho, Montana, New Mexico, Utah, Washington, and Wyoming (Vincent 1962, p. 15; Montana Fisheries Information System (MFISH) 2009; NatureServe 2010).

Differs Markedly in Its Genetic Characteristics

Differences in genetic characteristics can be measured at the molecular genetic or phenotypic level. Three different types of molecular markers (allozymes, mtDNA, and microsatellites) demonstrate that Arctic grayling from the upper Missouri River are genetically different from those in Canada, Alaska, and Asia (Everett 1986, pp. 79-80; Redenbach and Taylor 1999, p. 23; Stamford and Taylor 2004, p. 1538; Peterson and Ardren 2009, pp. 1764-1766; USFWS, unpublished data). These data confirm the reproductive isolation among populations that establishes the discreteness of Missouri River Arctic grayling under the DPS policy. Here, we speak to whether these data also establish significance.

Allozymes

Using allozyme electrophoretic data, Everett (1986, entire) found marked genetic differences among Arctic grayling collected from the Chena River in Alaska, those descended from fish native to the Athabasca River drainage in the Northwest Territories, Canada, and native upper Missouri River drainage populations or populations descended from them (see Leary 2005, pp. 1-2). The Canadian population had a high frequency of a unique isocitrate dehydrogenase allele (form of a gene) and a unique malate dehydrogenase allele, which strongly differentiated them from all the other samples (Everett 1986, p. 44). With the exception one introduced population in Montana that is believed to have experienced extreme genetic bottlenecks, the Chena River (Alaskan) fish were highly divergent from all the other samples as they possessed an unusually low frequency of superoxide dismutase (Everett 1986, p. 60; Leary 2005, p. 1), and contained a unique variant of the malate dehydrogenase (Leary 2005, p. 1). Overall, each of the four native Missouri River populations examined (Big Hole, Miner, Mussigbrod, and Red Rock) exhibited statistically significant differences in allele frequencies relative to both the Chena River (Alaska) and Athabasca River (Canada) populations (Everett 1986, pp. 15, 67).

Combining the data of Everett (1986, entire), Hop and Gharrett (1989, entire), and Leary (1990, entire) results in

information from 21 allozyme loci (genes) from the five native upper Missouri River drainage populations, five native populations in the Yukon River drainage in Alaska, and the one population descended from the Athabasca River drainage in Canada (Leary 2005, pp. 1-2). Examination of the genetic variation in these samples indicated that most of the genetic divergence is due to differences among drainages (29 percent) and comparatively little (5 percent) results from differences among populations within a drainage (Leary 2005, p. 1).

Mitochondrial DNA

Analysis using mtDNA suggest that Arctic grayling in North America represent at least three evolutionary lineages that are associated with distinct glacial refugia (Redenbach and Taylor 1999, entire; Stamford and Taylor 2004, entire). Arctic grayling in the Missouri River basin belong to the so-called North Beringia lineage (Redenbach and Taylor 1999, pp. 27-28; Samford and Taylor 2004, pp. 1538-1540). Analysis of Arctic grayling using restriction enzymes and DNA sequencing indicated that the fish from the upper Missouri River drainage possessed, in terms of North American fish, an ancestral form of the molecule (different forms of mtDNA molecules are referred to as haplotypes) that was generally absent from populations collected from other locations within the species' range in North America (Redenbach and Taylor 1999, pp. 27-28; Stamford and Taylor 2004, p. 1538). The notable exceptions were that some fish from the lower Peace River drainage in British Columbia, Canada (2 of 24 individuals in the population), and all sampled individuals from the Saskatchewan River drainage Saskatchewan, Canada (a total of 30 individuals from 2 populations), also possessed this haplotype (Stamford and Taylor 2004, p. 1538).

Variation in mtDNA haplotypes based on sequencing a portion of the ‘control region' of the mtDNA molecule of Arctic grayling from 26 different populations seems to support the groupings proposed by Stamford and Taylor (2004, entire) (USFWS unpublished data). Two haplotypes were common in the five native Missouri River populations (Big Hole, Red Rock, Madison, Miner, and Mussigbrod - total sample size 143 individuals; USFWS unpublished data). Fish from three populations in Saskatchewan or near Hudson's Bay also had one of these Missouri River haplotypes at very high frequency (50 of 51 individuals sequenced had the same haplotype; USFWS unpublished data). The two “common” Missouri River haplotypes also occurred at low frequency in handful of other populations elsewhere in Canada and Alaska. For example, there a total of five such populations where a few individuals contained had one or the other of the two common Missouri River haplotypes (25 of 107 individuals sequenced; USFWS unpublished data). Also similar to the earlier study by Stamford and Taylor (2004, entire), a few individuals (9 of 40 individuals) from two populations from the Lower Peace River and the Upper Yukon River also had one or the other of the two common Missouri River haplotypes (USFWS unpublished data).

The distribution of the common Missouri River haplotype compared to others suggested that Arctic grayling native to the upper Missouri River drainage probably originated from a glacial refuge in the drainage and subsequently migrated northwards when the Missouri River temporarily flowed into the Saskatchewan River and was linked to an Arctic drainage (Cross

et al.

1986, pp. 374-375; Pielou 1991, p. 195). When the Missouri River began to flow southwards because of the advance of the Laurentide ice sheet (Cross

et al.

1986, p. 375; Pileou 1991, p. 10), the Arctic grayling in the drainage became physically and reproductively isolated from the rest of the species' range (Leary 2005, p. 2; Campton 2006, p. 6), which would have included those populations in Saskatchewan. Alternatively, the Missouri River Arctic grayling could have potentially colonized Saskatchewan or the Lower Peace River (in British Columbia) or both post-glacially (Stamford 2001, p. 49) via a gap in the Cordilleran and Laurentide ice sheets (Pielou 1991, pp. 10-11), which also might explain the low frequency of one or the other of the ‘Missouri River' haplotypes in grayling in the Lower Peace River and Upper Yukon River.

We do not interpret the observation that Arctic grayling in Montana and Saskatchewan, and to lesser extent those from the Lower Peace and Upper Yukon River systems, share a mtDNA haplotype to mean that these groups of fish are genetically identical. Rather, we interpret it to mean that these fish shared a common ancestor tens to hundreds of thousands of years ago.

Microsatellite DNA

Recent analysis of microsatellite DNA (highly variable portions of nuclear DNA that exhibit tandem repeats of DNA base pairs) that included samples from five native Missouri River populations and two from Saskatchewan showed substantial divergence between these groups (Peterson and Ardren 2009, entire). Genetic differentiation between sample populations can be compared in terms of the genetic variation within relative to among populations, measured in terms of allele frequencies, a metric called

F

st

(Allendorf and Luikart 2007, pp. 52-54, 198-199). An analogous metric, named

R

st

, also measures genetic differentiation between populations based on microsatellite DNA, but differs from

F

st

in that it also considers the size differences between alleles (Hardy

et al.

2003, p. 1468). An

F

st

or

R

st

of 0 indicates that populations are the same genetically (all genetic diversity within a species is shared by all populations), whereas a value of 1 indicates the populations are completely different (all the genetic diversity within a species is found as fixed differences among populations).

F

st

values ranged from 0.13 to 0.31 (average 0.18) between Missouri River and Saskatchewan populations (Peterson and Ardren 2009, pp. 1758, 1764-1765), whereas

R

st

values ranged from 0.47 to 0.71 (average 0.54) for the same comparisons (Peterson and Ardren 2009, pp. 1758, 1764-1765). This indicates that the two groups (Missouri vs. Saskatchewan populations) differ significantly in allele frequency and also in the size differences, and therefore divergence, among those alleles. This indicates that the observed genetic differences are not simply due to random loss of genetic variation because the populations are isolated (genetic drift), but they also are due to mutational differences, which suggests the groups may have been separated for millennia (Peterson and Ardren 2009, pp. 1767-1768).

Comparison of 435 individuals from 21 Arctic grayling populations from Alaska, Canada, and the Missouri River basin using nine of the same microsatellite loci as Peterson and Ardren (2009, entire) further supports the distinction of Missouri River Arctic grayling relative to populations elsewhere in North America (USFWS, unpublished data). A statistical analysis that determines the likelihood that an individual fish belongs to a particular group (e.g., STRUCTURE) (Pritchard

et al.

2000, entire), clearly separated the sample fish from 21 populations into two clusters: one cluster representing populations from the upper Missouri River basin, and another cluster representing populations from across Canada and Alaska (USFWS, unpublished data). Factorial correspondence analysis (FCA) plots of individual fish also separated the fish

into two groups, or clouds of data points when visualized in a three-dimensional space (USFWS, unpublished data). The FCA is a multivariate data analysis technique used to simplify presentation of complex data and to identify systematic relations between variables, in this case the multi-locus genotypes of Arctic grayling. As with the other analysis, the FCA plots clearly distinguished Missouri River Arctic grayling from those native to Canada and Alaska (USFWS, unpublished data). Divergence in size among these alleles further supports the distinction between Missouri River grayling from those in Canada and Alaska (USFWS, unpublished data). The interpretation of these data is that the Missouri River populations and the Canada/Alaska populations are most genetically distinct at the microsatellite loci considered.

Phenotypic Characteristics Influenced by Genetics—Meristics

Phenotypic variation can be evaluated by counts of body parts (i.e., meristic counts of the number of gill rakers, fin rays, and vertebrae characteristics of a population) that can vary within and among species. These meristic traits are influenced by both genetics and the environment (Allendorf and Luikart 2007, pp. 258-259). When the traits are controlled primarily by genetic factors, then meristic characteristics can indicate significant genetic differences among groups. Arctic grayling north of the Brooks Range in Alaska and in northern Canada had lower lateral line scale counts than those in southern Alaska and Canada (McCart and Pepper 1971, entire). These two scale-size phenotypes are thought to correspond to fish from the North and South Beringia glacial refuges, respectively (Stamford and Taylor 2004, p. 1545). Arctic grayling from the Red Rock Lakes drainage had a phenotype intermediate to the large- and small-scale types (McCart and Pepper 1971, pp. 749, 754). Arctic grayling populations from the Missouri River (and one each from Canada and Alaska) could be correctly assigned to their group 60 percent of the time using a suite of seven meristic traits (Everett 1986, pp. 32-35). Those native Missouri River populations that had high genetic similarity also tended to have similar meristic characteristics (Everett 1986, pp. 80, 83).

Arctic grayling from the Big Hole River showed marked differences in meristic characteristics relative to two populations from Siberia, and were correctly assigned to their population of origin 100 percent of the time (Weiss

et al.

2006, pp. 512, 515-516, 518). The populations that were significantly different in terms of their meristic characteristics also exhibited differences in molecular genetic markers (Weiss

et al.

2006, p. 518).

Inference Concerning Genetic Differences in Arctic Grayling of the Missouri River Relative to Other Examples of the Taxon

We believe the differences between Arctic grayling in the Missouri River and sample populations from Alaska and Canada measured using microsatellite DNA markers (Peterson and Ardren 2009, pp. 1764-1766; USFWS, unpublished data) represent “marked genetic differences” in terms of the extent of differentiation (e.g.,

F

st

,

R

st

) and the importance of that genetic legacy to the rest of the taxon. The presence of morphological characteristics separating Missouri River Arctic grayling from other populations also likely indicates genetic differences, although this conclusion is based on a limited number of populations (Everett 1986, pp. 32-35; Weiss

et al.

2006, entire), and we cannot entirely rule out the influence of environmental variation.

The intent of the DPS policy and the ESA is to preserve important elements of biological and genetic diversity, not necessarily to preserve the occurrence of unique alleles in particular populations. In Arctic grayling of the Missouri River, the microsatellite DNA data indicate that the group is evolving independently from the rest of the species. The extirpation of this group would mean the loss of the genetic variation in one of the two most distinct groups identified in the microsatellite DNA analysis, and the loss of the future evolutionary potential that goes with it. Thus, the genetic data support the conclusion that Arctic grayling of the upper Missouri River represent a unique and irreplaceable biological resource of the type the ESA was intended to preserve. Thus, we conclude that Missouri River Arctic grayling differ markedly in their genetic characteristics relative to the rest of the taxon.

Conclusion

We find that a population segment that includes all native ecotypes of Arctic grayling in the upper Missouri River basin satisfies the discreteness standard of the DPS policy. The segment is physically isolated, and genetic data indicates that Arctic grayling in the Missouri River basin have been separated from other populations for thousands of years. The population segment occurs in an ocean drainage different from all other Arctic grayling populations worldwide, and we find that loss of this population segment would create a significant gap in the species' range. Molecular genetic data clearly differentiate Missouri River Arctic grayling from other Arctic grayling populations, including those in Canada and Alaska. We conclude that because Arctic grayling of the upper Missouri River basin satisfy the criteria for being discrete and significant under our DPS policy, we determined that this population constitutes a DPS under our policy and the Act.

In our stipulated settlement agreement, we also agreed to consider the appropriateness of distinct population segments based on the two different ecotypes (fluvial and adfluvial) expressed by native Arctic grayling of the upper Missouri River. We acknowledge there are cases where the Service has designated distinct population segments primarily on life-history even when they co-occur with another ecotype that can be part of the same gene pool (e.g., anadromous steelhead and resident rainbow trout,

Oncorhynchus mykiss

(71 FR 838, January 5, 2006). However, we conclude that designation of a single population segment for Arctic grayling in the upper Missouri River is more appropriate than designating two separate distinct population segments delineated by life-history type. In the Missouri River basin, the two ecotypes share a common evolutionary history, and do not cluster genetically based strictly on ecotype. As we discussed above, the fluvial and adfluvial life-history forms of Arctic grayling in the upper Missouri River do not appear to represent distinct evolutionary lineages. There appears to be some plasticity in behavior where individuals from a population can exhibit a range of behaviors. From a practical standpoint, we observe that only five native Arctic grayling populations remain in the Missouri River basin, and we believe that both fluvial and adfluvial native ecotypes have a role in the conservation of the larger population segment. We believe that the intent of the ESA and the DPS policy, and our obligation to assess the appropriateness of alternate DPS designations in the settlement agreement are best served by designating a single distinct population segment, rather than multiple population segments.

As we described above, we are not including introduced populations that occur in lakes in the Upper Missouri River basin in the DPS. The Service has interpreted the Act to provide a statutory directive to conserve species in their native ecosystems (49 FR 33890,

August 27, 1984) and to conserve genetic resources and biodiversity over a representative portion of a taxon's historical occurrence (61 FR 4723, February 7, 1996). The introduced Arctic grayling occur in lakes apart from native fluvial environments and from lakes where native adfluvial grayling occur. These introduced populations have not been used for any conservation purpose and could pose genetic risks to the native Arctic grayling population.

We find that the Arctic grayling of the upper Missouri River basin constitute a distinct population segment. We define the historical range of this population segment to include the major streams, lakes, and tributary streams of the upper Missouri River (mainstem Missouri, Smith, Sun, Beaverhead, Jefferson, Big Hole, and Madison Rivers, as well as their key tributaries, as well as a few small lakes where Arctic grayling are or were believed to be native (Elk Lake, Red Rock Lakes, Miner Lake, and Mussigbrod Lake, all in Beaverhead County, Montana). We define the current range of the DPS to consist of extant native populations in the Big Hole River, Miner Lake, Mussigbrod Lake, Madison River-Ennis Reservoir, and Red Rock Lakes. We refer to this DPS as the native Arctic grayling of the upper Missouri River. The remainder of this finding will thus focus on the population status of and threats to this entity.

Population Status and Trends for Native Arctic Grayling in the Upper Missouri River

We identified a DPS for Arctic grayling in the upper Missouri River basin that includes five extant populations: (1) Big Hole River, (2) Miner Lake, (3) Mussigbrod Lake, (4) Madison River-Ennis Reservoir, and (5) Red Rock Lakes. In general, we summarize what is known about the historical distribution and abundance of each of these populations, describe their current distributional extent, summarize any available population monitoring data, identify the best available information that we use to infer the current population status, and summarize the current population status and trends.

TABLE 4. Extent and current estimated effective population sizes (

N

e

) of native Arctic grayling populations in the Missouri River basin. Values in parentheses represent 95 percent confidence intervals.

Population Name

Population Extent

a

N

e

b

Biological Date of Population Size

c

Estimated Adult Population Size Assuming:

N

e

/N ratio 0.25

d

N

e

/N ratio 0.14

e

Big Hole River

158 mi

208 (176 to 251)

2000-2003

828 (704 to 1,004)

1,486 (1,257 to 1,793)

Miner Lakes

26.9 ha

286 (143 to 4,692)

2001-2003

1,144 (572 to 18,768)

2,043 (1,021 to 33,514)

Mussigbrod Lake

42.5 ha

1,497 (262 to ∞)

2001-2003

5,988 (1,048 to ∞)

10,693 (1,871 to ∞)

Madison River-Ennis Reservoir

1,469 ha

162 (76 to ∞)

1991-1993

648 (304 to ∞)

1,157 (543 to ∞)

Red Rock Lakes

890 ha

228 (141 to 547)

2000-2002

912 (564 to 2,188)

1,629 (1,007 to 3,907)

a

Approximate maximum spatial extent over which Arctic grayling are encountered in a given water.

b

Effective population size estimates from Peterson and Ardren (2009, p.1767). Confidence intervals that include infinity (∞) can result from statistical artifacts of the linkage disequilibrium method (Waples and Do 2007, p. 10; Russell and Fewster 2009, pp. 309-310). The usual interpretation is that there is no evidence for any disequilibrium caused by genetic drift due to a finite number of parents—it can all be explained by sampling error (Waples and Do 2007, p. 10). Thus, the effective size is infinitely large. Small sample sizes may influence estimates in some cases (e.g., Madison River-Ennis Reservoir).

c

Approximate date to which the

N

e

estimate refers. For example,

N

e

for the Big Hole River based on genotyping a sample of fish from 2005-2006, but the interpretation of

N

e

is the number of breeding adults that produced the fish in the observed sample. Thus the true biological date of the

N

e

estimate is one generation before 2005-2006, or approximately 2000-2003.

d

Adult population size estimated from

N

e

assuming

N

e

/N = 0.25. This value was the midpoint of a range of values (0.2-0.3) commonly cited for

N

e

/N ratios in salmonid fishes (Allendorf

et al.

1997, p. 143; McElhahey

et al.

2000, p. 63; Rieman and Allendorf 2001, p. 762; Palm

et al.

2003, p. 260).

e

Adult population size estimated from

N

e

assuming

N

e

/N = 0.14. This value was the median

N

e

/N ratio based on a meta analysis of 83 studies for 65 different species (Palstra and Ruzzante 2008, p. 3428).

Big Hole River

Historically, Arctic grayling presumably had access to and were distributed throughout much of the Big Hole River, including the lower reaches of many tributary streams, such as Big Lake, Deep, Doolittle, Fishtrap, Francis, Governor, Johnson, LaMarche, Miner, Mussigbrod, Odell, Pintlar, Rock, Sand Hollow, Swamp, Seymour, Steel, Swamp, and Wyman Creeks, as well as the Wise River (Liknes 1981, p. 11; Liknes and Gould 1987, p. 124; Kaya 1990, pp. 36-40). Presently, Arctic grayling are found primarily in the mainstem Big Hole River between the towns of Glen and Jackson, Montana, a distance of approximately 181 river km (113 mi), and in 11 tributaries, totaling an additional 72 river km (45 mi) (Magee 2010a, pers. comm.; see Table 4 above). The total current maximum extent of Arctic grayling occurrence in the Big Hole River is approximately 250 river km (156 mi). However, the fish are not continuously distributed across this distance, and instead tend to be concentrated in discrete patches (Magee

et al.

2006, pp. 27-28; Rens and Magee 2007, p. 15) typically associated with spawning and rearing habitats or cold-water sites that provide a thermal refuge from high summer water temperatures.

Kaya (1992, pp. 50-52) noted the general lack of monitoring data for the Big Hole River fluvial Arctic grayling population prior to the late 1970s, but data collected since that time indicate the overall range has contracted over the last 2 decades. During 1978 and 1979 Arctic grayling were observed in Governor Creek (in the headwaters of the Big Hole River) and downstream in the Big Hole River near Melrose, Montana (Liknes 1981, p. 11). Arctic grayling have not recently been encountered in Governor Creek (Rens and Magee 2007, p. 15; Montana Fish, Wildife and Parks (MFWP), unpublished data), but are occasionally

encountered in the Big Hole River downstream of Divide, Montana, at very low densities and as far downstream as Melrose or Glen, Montana (Oswald 2005a, pers. comm.). More recently, Arctic grayling have become less abundant in historical spawning and rearing locations in the upper watershed near Wisdom, Montana, and also in downstream river segments with deep pool habitats considered important for overwintering (Magee and Lamothe 2003, pp. 18-21; MFWP unpublished data). Comparatively, greater numbers of Arctic grayling are encountered in the lower reaches of tributaries to the upper Big Hole River, including LaMarche, Fishtrap, Steel, and Swamp Creeks (Rens and Magee 2007, p. 13).

Based on the best available data, the adult population declined by one half between the early 1990s and the early 2000s (see Figure 3, USFWS unpublished data), which is equivalent to a decline of 7 percent per year, on average. Monitoring data collected by MFWP also support the conclusion that the Arctic grayling population in the Big Hole River declined during this time period (Byorth 1994a, p. 11; Rens and Magee 2007, entire; MFPW, unpublished data).

EP08SE10.002

FIGURE 3

. Effective population size (

N

e

) of Big Hole River Arctic grayling based on microsatellite DNA genotypes from fish collected in three time periods (USFWS, unpublished data). The

N

e

are estimated using the linkage disequilibrium method of Waples and Do (2008, entire), and error bars represent 95% confidence intervals estimated by the jackknife method.

Miner Lakes

The Miner Lakes are a complex of small lakes in the upper Big Hole River drainage. Lower Miner Lakes are two small lakes in the middle of the Miner Creek drainage connected by a narrow section approximately 100 m (330 ft) in length, functionally representing a single lake for fish populations. Arctic grayling occur in Lower Miner Lakes (hereafter Miner Lakes population), which has a total surface area of 26.7 hectares (ha) or 0.267 km

2

(66 acres (ac)). Arctic grayling primarily reside in the lake, and presumably move into the inlet or outlet tributary to spawn. Surveys conducted upstream and downstream of the Lower Miner Lakes in 1992 and 1994, respectively, captured no Arctic grayling (Downing 2006, pers. comm.). Apparently, adults do not remain in the stream long after spawning and young-of-the-year (YOY) move into Lower Miner Lakes.

The MFWP conducted limited surveys in Lower Miner Lakes, but the abundance of the population has not been estimated by traditional fishery methods. Arctic grayling are classified as “common” in Lower Miner Lakes (MFISH 2010). Introduced brook trout also are present.

The best available information on the abundance of Miner Lakes Arctic grayling comes from a genetic assessment of that population. Based on a sample of fish from 2006, Peterson and Ardren (2009, p. 1767) estimated an effective population size of 286. This estimate represents an approximation of abundance of breeding adults at a single point in time, and there are no data on which to base an assessment of the population trend.

Mussigbrod Lake

Mussigbrod Lake has a surface area of 42.5 ha (105 ac), and is found in the middle reaches of Mussigbrod Creek, a tributary to the North Fork Big Hole River. Arctic grayling primarily reside in the lake. We do not know whether Arctic grayling spawn in the inlet stream or within the lake (Magee and

Olsen 2010, pers. comm.). Arctic grayling occasionally pass over a diversion structure downstream at the outlet of Mussigbrod Lake, and become trapped in a pool that is isolated because of stream dewatering. The MFWP periodically capture grayling in this pool and return them to the lake.

Data for the Mussigbrod Lake population of Arctic grayling is minimal. The MFWP has conducted very limited surveys and the abundance of the population has not been estimated by traditional fishery methods. Genetic data indicate that Arctic grayling are comparatively abundant (see Table 4 above). Based on a sample from 2006, Peterson and Ardren (2009, p. 1767) estimated an effective size of 1,497. The best available data indicate that the Mussigbrod Lake population is comparatively large, but we have no data about the population trend.

Madison River - Ennis Reservoir

Historically, Arctic grayling were reported to be abundant in the middle and upper Madison River, but have undergone a dramatic decline in the past 100 years with the species becoming rare by the 1930s (Vincent 1962, pp. 11, 85-87). Native Arctic grayling are thought be extirpated from the upper Madison River. A major impact to fish in that area was the construction of Hebgen Dam, which flooded Horsethief Springs, a small tributary that was reportedly one of the most important streams for Arctic grayling (Vincent 1962, pp. 40-41, 128). In the middle Madison River, Arctic grayling were apparently common to plentiful in the mainstem River near Ennis, Montana, and some associated tributaries (Jack, Meadow, and O'Dell Creeks) (Vincent 1962, p. 128). In 1906, construction of Ennis Dam blocked all upstream movement of fishes, and apparently had a large negative effect on Arctic grayling. Vincent (1962) noted that “early settlers reported scooping up boxes full of grayling at the base of Ennis Dam the year after it was constructed” (p. 128), and that the species apparently became quite rare by the late 1930s (Vincent 1962, p. 85).

The current distribution of Arctic grayling in the Madison River is primarily restricted to the Ennis Reservoir and upstream into the river approximately 6.5 km (approximately 4 mi) to the Valley Garden Fishing Access Site (Byorth and Shepard 1990, p. 21). Arctic grayling are occasionally encountered in the Madison River downstream and upstream from Ennis Reservoir (Byorth and Shepard 1990, p. 25; Clancey 2004, p. 22; 2008, p. 21). Arctic grayling migrate from the reservoir into the river to spawn, then return to the reservoir (Byorth and Shepard 1990, pp. 21-22; Rens and Magee 2007, pp. 20-21). The YOY Arctic grayling spawned in the Madison River migrate downstream into Ennis Reservoir about 1 month after emergence, but while they are in the river, they are typically encountered in backwater or slackwater habitat (Jeanes 1996, pp. 31-34).

The MFWP has sporadically monitored Arctic grayling in the Madison River near Ennis Reservoir since about 1990. Despite sparse data, declining catches for both spawning adults and YOY indicate the population is less abundant now compared to the early 1990s. The highest numbers of YOY Arctic grayling were encountered in the early 1990s, and no more than two have been captured in any given year since that time. Our interpretation of this information is that Arctic grayling in the Madison River-Ennis Reservoir population have declined during the past 20 years and are presently at very low abundance.

Abundance of the Madison River-Ennis Reservoir Arctic grayling has been estimated twice. In 1990, the adult population was estimated to be 545, but the authors cautioned that the accuracy of the estimate was questionable as it was based on recapturing only. From a sample of fish collected mostly in 1996, the effective size of the population (breeding adults) was estimated as 162 (Peterson and Ardren 2009, p. 1767). The average number of Arctic grayling captured per unit effort (CPUE) declined by approximately a factor of 10 between the early 1990s and recent samples (Clancey 1998, p. 10; Clancey 2007, p.16; Clancey 2008, pp. ii, 21, A2-2; Clancey and Lohrenz 2009, pp. 30, B2; Clancey 2010a, pers. comm.; Clancey 2010b, pers. comm.). Adult Arctic grayling may currently exist at only 10 to 20 percent of the abundance observed in the early 1990s. Based on the best available data, we conclude that this Arctic grayling population has been in a decline during the past 20 years and may only consist of a few hundred adults.

Red Rocks Lakes

Arctic grayling are native to waters of the upper Beaverhead River system, including the Red Rock River drainage. During the past 50 to 100 years, both the distribution and abundance of Arctic grayling in the Centennial Valley, Beaverhead County, Montana (which contains the Red Rock River), has severely declined (Vincent 1962, pp. 115-121; Unthank 1989, pp. 13-17; Mogen 1996, pp. 2-5, 75-84). As of about 50 years ago, Arctic grayling spawned in at least 12 streams in the Centennial Valley (Mogen 1996, p. 17), but they appear to have been extirpated from all but 2 streams (Boltz 2006, p. 6). Presently, Arctic grayling spawn in two locations within the Red Rock River drainage: Odell Creek, a tributary to Lower Red Rock Lake; and Red Rock Creek, the primary tributary to Upper Red Rock Lake (Mogen 1996, pp. 47-48; Boltz 2006, p. 1). Lower and Upper Red Rock Lakes are connected by a short segment of river, and both lakes are contained within the boundaries of the Red Rock Lakes National Wildlife Refuge (NWR). The upper lake appears to be the primary rearing and overwintering habitat for Arctic grayling. Red Rock Creek is the only stream where Arctic grayling spawn in appreciable numbers (Mogen 1996, pp. 45-48). Collectively, we refer to this population as the Red Rocks Lakes Arctic grayling, and characterize it as having the adfluvial ecotype.

Arctic grayling in the Red Rock Lakes have been monitored intermittently since the 1970s. Most of that effort focused on Red Rock Creek, but periodic sampling also occurred in Odell Creek. The MFWP and the Service occasionally sampled for Arctic grayling in Odell Creek, where grayling abundance declined over the past few decades. On average, the minimum sizes of the spawning runs in Red Rock Creek since 1994 are about half of those recorded 4 decades ago (i.e., 623 vs. 308 per year) (data summarized from Mogen 1996, p. 70 and Boltz 2006, p. 7). The spawning runs into Red Rock Creek fluctuated during the 1990s and early 2000s, but about 450 or fewer adult Arctic grayling have been captured in 6 of 7 years in which weirs traps were operated. Electrofishing surveys conducted in Red Rock Creek by MFWP seem to corroborate a decline in the spawning population, as total catches decreased even as sampling effort increased (Rens and Magee 2007, pp. 16-18).

Based on a sample of fish from Red Rock Creek in 2005, Peterson and Ardren (2009, pp. 1761, 1767) estimated an effective size of 228, which is interpreted as the number of breeding adults that produced the fish sampled in 2005. The best available data indicate that the Red Rock Lakes Arctic grayling population has declined over the past 2 decades.

Population viability analysis (PVA) of native Missouri River Arctic grayling

To gauge the probability that the different native populations of Arctic grayling in the upper Missouri River

basin will go extinct from unpredictable events in the foreseeable future, we conducted a simple population viability analysis (PVA) (see Dennis

et al.

(1991, entire) in Morris and Doak 2002, pp. 85-87 for details on the PVA model and the software code to run the model). We assumed that a population with 50 or fewer adults is likely influenced by demographic stochasticity (chance variation in the fates of individuals within a given year) and genetic stochasticity (random changes in a population's genetic makeup), and would not be expected to persist long as a viable population. For the different PVA scenarios, we assume either the population has stabilized, or the estimated decline will continue at a constant rate.

We considered the probability of extinction individually by population, as populations appear to be reproductively isolated. The relative risk of extinction in the foreseeable future (30 years based on the observation that the variability in predictions for extinction risk from the PVA model increases substantially after 30 years) varies among the different populations, with the largest population, Mussigbrod Lake, having a very low probability of extinction (less than 1 percent) in the foreseeable future, even given a population decline. The other four populations have comparatively greater probabilities of extinction in the foreseeable future, with all being roughly similar in magnitude (13-55 percent across populations) when considering only stochastic (random or chance) processes. The Madison River has the greatest probability of extinction by stochastic processes (36-55 percent), followed by Big Hole (33-42 percent), Red Rocks (31-40 percent), and Miner (13-37 percent).

Overall, the PVA analyses indicate that four populations (Madison, Big Hole, Red Rocks, and Miner) appear to be at risk from chance environmental variation because of low population abundance. This is a general conclusion, and the actual risk may vary substantially among populations (USFWS unpublished data). For example, Arctic grayling in the Big Hole River population spawn in different locations, which would reduce the risk that an environmental catastrophe would simultaneously kill all breeding adults, relative to a situation where adults appear to be primarily in a single location or reach of river (e.g., Red Rocks and Madison populations).

Arctic Grayling Conservation Efforts

Native Arctic Grayling Genetic Reserves and Translocation

Given concern over the status of native Arctic grayling, the Montana Arctic Grayling Recovery Program (AGRP) was formed in 1987, to address conservation concerns for primarily the fluvial ecotype in Big Hole River, and to a lesser extent the native adflvuial population in Red Rock Lakes (Memorandum of Understanding (MOU) 2007, p. 2). The AGW was established as an ad hoc technical workgroup of the AGRP. In 1995, the AGW finalized a restoration plan that outlined an agenda of restoration tasks and research, including management actions to secure the Big Hole River population, brood stock development, and a program to re-establish four additional fluvial populations (AGW 1995, pp. 7-17).

Consequently, the State of Montana established genetic reserves of Big Hole River grayling (Leary 1991, entire), and has used the progeny from those reserves in efforts to re-establish additional fluvial populations within the historical native range in the Missouri River basin (Rens and Magee 2007, pp. 21-38). Currently, brood (genetic) reserves of Big Hole River grayling are held in two closed-basin lakes in south-central Montana (Rens and Magee 2007, p. 22). These fish are manually spawned to provide gametes for translocation efforts in Montana (Rens and Magee 2007, p. 22). Functionally, these brood reserves are hatchery populations maintained in a natural setting, and we do not consider them wild populations for the purposes of evaluating the status of native Arctic grayling in the Missouri River basin. However, they are important to recovery efforts.

For more than 13 years, MFWP has attempted to re-establish populations of fluvial Arctic grayling in various locations in the Missouri River basin, including the Ruby, Sun, Beaverhead, Missouri, Madison, Gallatin, and Jefferson Rivers (Lamothe and Magee 2004a, pp. 2, 28). A self-sustaining population has not yet been established from these reintroductions (Lamothe and Magee 2004a, p. 28; Rens and Magee 2007, pp. 35-36, 38). Recent efforts have focused more intensively on the Ruby and Sun Rivers, and have used methods that should improve reintroduction success (Rens and Magee 2007, pp. 24-36). Encouragingly, natural reproduction by Arctic grayling in the Ruby River was confirmed during fall 2009 (Magee 2010b, pp. 6-7, 22). Monitoring will continue in subsequent years to determine whether the population has become a stable and viable population, as defined by the guidance and implementation documents of the translocation programs (AGW 1995, p. 1; Memorandum of Agreement (MOA) 1996, p. 2). Consequently, we do not consider the Ruby River to represent a self-sustaining population for the purposes of evaluating the population status of Missouri River grayling in this finding. Arctic grayling presumably from previous translocations are occasionally encountered near translocation sites in other waters (Rens and Magee 2007, pp. 35-38; MFWP, unpublished data). There is no evidence that these individuals represent progeny from a re-established population, so we cannot consider them elements of a stable and viable population for the purposes of evaluating the population status of Missouri River Arctic grayling in this finding.

Big Hole River Candidate Conservation Agreement with Assurances

On August 1, 2006, the Service issued ESA section 10(a)(1)(A) enhancement of survival permit (TE-104415-0) to Montana Fish, Wildlife and Parks (MFWP) to implement a Candidate Conservation Agreement with Assurances for Arctic grayling in the upper Big Hole River (Big Hole Grayling CCAA) (MFWP

et al.

2006, entire). This permit is valid through August 1, 2026. The goal of the Big Hole Grayling CCAA is to secure and enhance a population of fluvial Arctic grayling within the upper reaches of their historic range in the Big Hole River drainage by working with non-Federal property owners to implement conservation measures on their lands. The guidelines of this CCAA will be met by implementing conservation measures that improve stream flows, protect and restore riparian habitats, identify and reduce or eliminate entrainment (inadvertent capture) of grayling in irrigation ditches, and remove human-made barriers to grayling migration (MFWP

et al.

2006, p. 3). Currently, 32 landowners representing 64,822 ha (160,178 ac) in the upper Big Hole River drainage are participating in the CCAA (Lamothe 2009, p. 5). The MFWP leads the Big Hole Grayling CCAA implementation effort, and is supported by Montana Department of Natural Resources and Conservation (MDNRC), USDA Natural Resources Conservation Service (NRCS), and the Service. Other groups helping implement the CCAA include the Big Hole Watershed Committee, the Big Hole River Foundation, Montana Trout Unlimited, the Western Water Project (affiliated with Trout Unlimited), and

The Nature Conservancy (Lamothe 2008, p. 23). Detailed information on conservation actions and restoration projects implemented under the plan are available in various reports (AGW 2010, p. 4; Everett 2010, entire; Lamothe

et al.

2007, pp. 6-35; Lamothe 2008, pp. 7-21; Lamothe 2009, entire; Lamothe 2010, entire; Magee 2010b, entire; Roberts 2010, entire).

Biological Effectiveness of the Ongoing Conservation Programs

The current and anticipated effects of the aforementioned conservation programs on the biological status and threats to Arctic grayling of the upper Missouri River are discussed elsewhere in the document (see

Summary of Information Pertaining to the Five Factors

and

Finding

sections, below). We continue to encourage and promote collaborative efforts to secure existing populations, and to increase the distribution of the Arctic grayling within its historical range in the upper Missouri River basin.

Summary of Information Pertaining to the Five Factors

Section 4 of the ESA (16 U.S.C. 1533) and implementing regulations (50 CFR 424) set forth procedures for adding species to the Federal Lists of Endangered and Threatened Wildlife and Plants. Under section 4(a)(1) of the ESA, a species may be determined to be endangered or threatened based on any of the following five factors: (A) The present or threatened destruction, modification, or curtailment of its habitat or range; (B) overutilization for commercial, recreational, scientific, or educational purposes; (C) disease or predation; (D) the inadequacy of existing regulatory mechanisms; or (E) other natural or manmade factors affecting its continued existence. In making this finding, information pertaining to the Missouri River DPS of Arctic grayling in relation to the five factors provided in section 4(a)(1) of the Act is discussed below.

In considering what factors might constitute threats to a species, we must look beyond the exposure of the species to a factor to evaluate whether the species may respond to the factor in a way that causes actual impacts to the species. If there is exposure to a factor and the species responds negatively, the factor may be a threat and we attempt to determine how significant a threat it is. The threat is significant if it drives, or contributes to, the risk of extinction of the species such that the species warrants listing as endangered or threatened as those terms are defined in the Act.

A. The Present or Threatened Destruction, Modification, or Curtailment of Its Habitat or Range

Curtailment of Range and Distribution

The number of river kilometers (miles) occupied by the fluvial ecotype of Arctic grayling in the Missouri River has been reduced by approximately 95 percent during the past 100 to 150 years (Kaya 1992, p. 51). The fluvial life history is only expressed in the population residing in the Big Hole River; the remnant population in the Madison River near Ennis Reservoir has apparently diverged toward an adfluvial life history. Arctic grayling distribution within the Centennial Valley in the upper Beaverhead River also has been severely curtailed during the last 50 to 100 years, such that the only remaining example of the species in that drainage is an adfluvial population associated with the Red Rock Lakes. Indigenous populations in the Big Hole River, Madison River, and Red Rock Lakes all exist at reduced densities on both contemporary and historical timescales. The Miner Lakes and Mussigbrod Lake populations appear to have been reproductively isolated for hundreds of years (USFWS, unpublished data), so a restricted distribution may represent the natural historical condition for these populations. The curtailment of range and distribution is a current threat, because the probability of extirpation of the DPS is related to the number of populations and their resilience. Since the DPS currently exists as a set of generally small, isolated populations that cannot naturally re-found or ‘rescue' another population. Thus, the curtailment of range and distribution will remain a threat in the foreseeable future, absent the reestablishment of additional populations within the DPS' historical range. Reintroduction attempted under the auspices of the 1995 Restoration Plan (AGW 1995, entire) have been underway since 1997, but have not yet resulted in re-establishment of populations or the expansion of the DPS' current range.

Dams on Mainstem Rivers

The majority of the historical range of the Upper Missouri River DPS of Arctic grayling has been altered by the construction of dams and reservoirs that created barriers obstructing migrations to spawning, wintering, or feeding areas; inundated grayling habitat; and impacted the historical hydrology of river systems (Kaya 1990, pp. 51-52; Kaya 1992, p. 57). The construction of large dams on mainstem river habitats throughout the upper Missouri River system fragmented river corridors necessary for the expression of migratory life histories. Construction of dams that obstructed fish passage on the mainstem Missouri River (Hauser, Holter, Canyon Ferry, and Toston), Madison River (Madison-Ennis, Hebgen), Beaverhead River and its tributary Red Rock River (Clark Canyon, Lima), Ruby River (Ruby), and Sun River (Gibson) all contributed to the rangewide decline of this DPS (Vincent 1962, pp. 127-128; Kaya 1992, p. 57; see Figure 2).

Dams also may continue to impact the extant population in the Madison River. The Madison Dam (also known as Ennis Dam), as with the aforementioned dams, is a migration barrier with no fish passage facilities. Anglers have reported encountering Arctic grayling in pools below the dam, implying that fish occasionally pass (downstream) over or through the dam. These fish would be “lost” to the population residing above the dam because they cannot return upstream, but have apparently not established populations downstream. Operational practices of the Madison Dam also have been shown to affect the resident fishes. A population decline of Arctic grayling coincided with a reservoir drawdown in winter 1982-1983 that was intended to reduce the effects of aquatic vegetation on the hydroelectric operations at the dam (Byorth and Shepard 1990, pp. 52-53). This drawdown likely affected the forage base, rearing habitat, and spawning cycle of Arctic grayling in the reservoir.

The presence of mainstem dams is a historical, current, and future threat to the DPS. Lack of fish passage at these dams contributed to the extirpation of Arctic grayling from some waters by blocking migratory corridors (Vincent 1962, p. 128), curtailing access to important spawning and rearing habitats, and impounding water over former spawning locations (Vincent 1962, p. 128). These dams are an impediment to fish migration and limit the ability of fish to disperse between existing populations or recolonize habitat fragments, and will continue to act in this manner for the foreseeable future. We believe the presence of a mainstem dam is an immediate and imminent threat to the Madison River population, as the remaining grayling habitat is adjacent to Ennis Dam (see Figure 2). We not aware of any plans to retrofit the Ennis Dam or any other mainstem dam to provide upstream fish passage, so we expect the current situation to continue. The Federal Energy Regulatory Commission (FERC) license for hydroelectric generation at Ennis Dam will not expire until the year

2040 (FERC 2010, entire). The upper Missouri River basin dam having the FERC license with the latest expiration date is Clark Canyon Dam, which will not expire until 2059 (FERC 2010, entire). Thus, mainstem dams will remain a threat in the foreseeable future, which is 30 to 50 years based on the duration of existing FERC licenses in the upper basin.

Agriculture and Ranching

The predominant use of private lands in the upper Missouri River basin is irrigated agriculture and ranching, and these activities had and continue to have significant effects on aquatic habitats. In general, these effects relate to changes in water availability and alteration to the structure and function of aquatic habitats. The specific activities and their impacts are discussed below.

Smaller Dams and Fish Passage Barriers

Smaller dams or diversions associated with irrigation structures within specific watersheds continue to pose problems to Arctic grayling migratory behavior, especially in the Big Hole River drainage. In the Big Hole River, numerous diversion structures have been identified as putative fish migration barriers (Petersen and Lamothe 2006, pp. 8, 12-13, 29) that may limit the ability of Arctic grayling to migrate to spawning, rearing, or sheltering habitats under certain conditions. The Divide Dam on the Big Hole River near the town of Divide, Montana, has existed for nearly 80 years and is believed to be at least a partial barrier to upstream movement by fishes (Kaya 1992, p. 58). As with the larger dams, these smaller fish passage barriers can reduce reproduction (access to spawning habitat is blocked), reduce growth (access to feeding habitat is blocked), and increase mortality (access to refuge habitat is blocked). A number of planned or ongoing conservation actions to address connectivity issues on the Big Hole River and its tributaries may reduce the threat posed by movement barriers for Arctic grayling in that habitat. The Divide Dam is being replaced with a new structure that provides fish passage, and construction began in July 2010 (Nicolai 2010, pers. comm.). At least 17 fish ladders have been installed at diversion structures in the Big Hole River since 2006 as part of the Big Hole Grayling CCAA (AGW 2010, p. 4), and a culvert barrier at a road crossing on Governor Creek (headwaters of Big Hole River) was replaced with a bridge that is expected to provide upstream passage for aquatic organisms under all flow conditions (Everett 2010, pp. 2-6). Non-Federal landowners who control approximately 50 to 70 percent of the points of irrigation diversion in the upper Big Hole River are enrolled in the CCAA (Roberts and Lamothe 2010, pers. comm.), so the threats posed by fish passage barriers should be substantially reduced in the Big Hole River during the next 10 to 20 years (foreseeable future) based on the minimum duration of site-specific plans for landowners enrolled in the CCAA and the duration of the ESA section 10(a)(1)(A) enhancement of survival permit (TE 104415-0) associated with the CCAA (MFWP

et al.

2006, p. 75).

Fish passage barriers also have been noted in the Red Rock Lakes system (Unthank 1989, p. 9). Henshall (1907, p. 5) noted that spawning Arctic grayling migrated from the Jefferson River system, through the Beaverhead River and Red Rock River through the Red Rock Lakes and into the upper drainage, and then returned downstream after spawning. The construction of a water control structure (sill) at the outlet of Lower Red Rock Lake in 1930 (and reconstructed in 1957 (USFWS 2009, p. 74)) created an upstream migration barrier that blocked these migrations (Unthank 1989, p. 10; Gillin 2001, p. 4-4). This structure, along with mainstem dams at Lima and Clark Canyon, extirpated spawning runs of Arctic grayling that historically migrated through the Beaverhead and Red Rock Rivers (see Figure 2; USFWS 2009, p. 72). All of these structures preclude upstream movement by fishes, and continue to prohibit immigration of Arctic grayling from the Big Hole River (see Figure 2). Because recovery of Arctic grayling will necessitate expansion into unoccupied habitat, and the Big Hole River includes some of the best remaining habitat for the species, these dams constitute a threat to Arctic grayling now and in the foreseeable future, which is 30 to 50 years based on the duration of existing FERC licenses in the upper basin.

In Mussigbrod Lake, Arctic grayling occasionally pass downstream over a diversion structure at the lake outlet, and become trapped in a pool that is isolated because of stream dewatering (Magee and Olsen 2010, pers. comm.). However, the potential for mortality in these fish is partially mitigated by MFWP, which periodically captures Arctic grayling in this pool and returns them to the lake.

In the Red Rock Lakes system, the presence of fish passage barriers represents a past and present threat. The magnitude of the threat may be reduced in the next 15 years as a result of implementation of the Red Rock Lakes NWR Comprehensive Conservation Plan (CCP) (USFWS 2009, entire — see Factor D discussion below), but we conclude that not all barriers that potentially affect the population will addressed during this time (e.g., Lower Red Rock Lake Water Control Structure) (USFWS 2009, p. 43). Thus, fish passage barriers will remain a threat to the Red Rock Lakes grayling in the foreseeable future.

In the Big Hole River, fish passage barriers represent a past and present threat. The magnitude of the threat in the Big Hole River should decrease appreciably during the next 10 to 20 years, which represents the foreseeable future in terms of the potential for the Big Hole Grayling CCAA to address the threat. Additional projects, such as the replacement of the Divide Dam, also should reduce the threat in the foreseeable future.

Dewatering From Irrigation and Consequent Increased Water Temperatures

Demand for irrigation water in the semi-arid upper Missouri River basin has dewatered many rivers formerly or currently occupied by Arctic grayling. The primary effects of this dewatering are: 1) Increased water temperatures, and 2) reduced habitat capacity. In ectothermic species like salmonid fishes, water temperature sets basic constraints on species distribution and physiological performance, such as activity and growth (Coutant 1999, pp. 32-52). Increased water temperatures can reduce the growth and survival of Arctic grayling (physiological stressor). Reduced habitat capacity can concentrate fishes and thereby increase competition and predation (ecological stressor).

In the Big Hole River system, surface-water (flood) irrigation has substantially altered the natural hydrologic function of the river and has led to acute and chronic stream dewatering (Shepard and Oswald 1989, p. 29; Byorth 1993, p. 14; 1995, pp. 8-10; Magee

et al.

2005, pp. 13-15). Most of the Big Hole River mainstem exceeds water quality standards under the Clean Water Act (33. U.S.C. 1251

et seq.

; see discussion under Factor D, below) because of high summer water temperatures (Flynn

et al.

2008, p. 2). Stream water temperature is affected by flow volume, stream morphology, and riparian shading, along with other factors, but an inverse relationship between flow volume and water temperature is apparent in the Big Hole River (Flynn

et al.

2008, pp. 18-19). Summer water temperatures exceeding 21 °C (70 °F) are

considered to be physiologically stressful for cold-water fish species, such as Arctic grayling (Hubert

et al.

1985, pp. 7, 9). Summer water temperatures consistently exceed 21 °C (70 °F) in the mainstem of Big Hole River (Magee and Lamothe 2003, pp. 13-14; Magee

et al.

2005, p. 15; Rens and Magee 2007, p. 11). Recently, summer water temperatures have consistently exceeded the upper incipient lethal temperature (UILT) for Arctic grayling (e.g., 25 °C or 77 °F) (Lohr

et al.

1996) at a number of monitoring stations throughout the Big Hole River (Magee and Lamothe 2003, pp. 13-14; Magee

et al.

2005, p. 15; Rens and Magee 2007, p. 11). The UILT is the temperature that is survivable indefinitely (for periods longer than 1 week) by 50 percent of the “test population” in an experimental setting. Fish kills are a clear result of high water temperature and have been documented in the Big Hole River (Lohr

et al.

1996, p. 934). Consequently, water temperatures that are high enough to cause mortality of fish in the Big Hole River represent a clear threat to Arctic grayling because of the potential to directly and quickly reduce the size of the population.

Water temperatures below that which can lead to instant mortality also can affect individual fish. At water temperatures between 21 °C (70 °F) and 25 °C (77 °F), Arctic grayling can survive but experience chronic stress that can impair feeding and growth, reduce physiological performance, and ultimately reduce survival and reproduction. As described above, the Big Hole River periodically experiences summer water temperatures high enough to cause morality and chronic stress to Arctic grayling. Increased water temperature also appears to be a threat to Arctic grayling in the Madison River and Red Rock watershed. Mean and maximum summer water temperatures can exceed 21 °C (70 °F) in the Madison River below Ennis Reservoir (U.S. Geological Survey (USGS) 2010), and have exceeded 22 °C (72 °F) in the reservoir, and 24 °C (75 °F) in the reservoir inlet (Clancey and Lohrenz 2005, p. 34). Similar or higher temperatures have been noted at these same locations in recent years (Clancey 2002, p. 17; 2003, p. 25; 2004, pp. 29-30). Surface water temperatures in Upper Red Rock Lake as high as 24 °C (75 °F) have been recorded (Gillin 2001, p. 4-6), and presence of Arctic grayling in the lower 100 m (328 ft) of East Shambow Creek in 1994 was attributed to fish seeking refuge from high water temperatures in the lake (Mogen 1996, p. 44). Mean summer water temperatures in Red Rock Creek can occasionally exceed 20°C or 68°F during drought conditions (Mogen 1996, pp. 19, 45). Arctic grayling can survive but experience chronic stress that can impair feeding and growth, reduce physiological performance, and ultimately reduce survival and reproduction.

Experimental data specifically linking hydrologic alteration and dewatering to individual and population-level effects for Arctic grayling is generally lacking (Kaya 1992, p. 54), but we can infer effects from observations that the abundance and distribution of Arctic grayling has declined concurrent with reduced streamflows (MFWP

et al.

2006, pp. 39-40) and increased water temperatures associated with low streamflows.

In the Big Hole River system, early-season (April through May) irrigation withdrawals may dewater grayling spawning sites (Byorth 1993, p. 22), preventing spawning or causing egg mortality; can prevent juvenile grayling from accessing cover in the vegetation along the shoreline; and may reduce connectivity between necessary spawning, rearing, and refuge habitats. Severe dewatering reduces habitat volume and may concentrate fish, increasing the probability of competition and predation among and between species. Nonnative trout species presently dominate the salmonid community in the Big Hole River, so dewatering would tend to concentrate Arctic grayling in habitats where interactions with these nonnative trout would be likely.

Especially in the Big Hole River, dewatering from irrigation represents a past and present threat to Arctic grayling. Thermal loading has apparently been a more frequent occurrence in the Big Hole River than in other locations containing native Arctic grayling (e.g., Red Rock Creek and Madison River-Ennis Reservoir). Implementation of the Big Hole Grayling CCAA during the next 20 years, which requires conservation measures to increase stream flows and restore riparian habitats (MFWP 2006, pp. 22-48), should significantly reduce the threat of thermal loading for Big Hole River grayling in the foreseeable future. While we expect agricultural and ranching-related use of water to continue, we expect that the threat will be reduced, but not eliminated, in the foreseeable future in the Big Hole River as a consequence of the CCAA. The ability of the Big Hole Grayling CCAA to augment streamflows should be substantial, as non-Federal landowners who control approximately 50 to 70 percent of the points of irrigation diversion in the upper Big Hole River are enrolled in the CCAA (Roberts and Lamothe 2010, pers. comm.). However, the Big Hole River constitutes one population in the DPS and high water temperatures are likely to continue to affect grayling in the Madison River and Red Rock Lakes. Thus, stream dewatering and high water temperatures are expected to remain a threat to the DPS in the foreseeable future.

Entrainment

Entrainment can permanently remove individuals from the natural population and strand them in a habitat that lacks the required characteristics for reproduction and survival. Irrigation ditches may dry completely when irrigation headgates are closed, resulting in mortality of entrained grayling. Entrainment of individual Arctic grayling in irrigation ditches occurs in the Big Hole River (Skarr 1989, p. 19; Streu 1990, pp. 24-25; MFWP

et al.

2006, p. 49; Lamothe 2008, p. 22). Over 1,000 unscreened diversion structures occur in the upper Big Hole River watershed, and more than 300 of these are located in or near occupied grayling habitat (MFWP

et al.

2006, pp. 48-49).

The magnitude of entrainment at unscreened diversions can depend on a variety of physical and biological factors, including the volume of water diverted (Kennedy 2009, p. iv, 36-38; but see Post

et al.

2007, p. 885), species-specific differences in the timing of migratory behavior relative to when water is being diverted (Carlson and Rahel 2007, pp. 1340-1341), and differences in vulnerability among body size or life-stage (Gale 2005, pp. 30-47; Post

et al.

2006, p. 975; Carlson and Rahel 2007 pp. 1340-1341). Studies of other salmonid species in a river basin in southwestern Wyoming determined that ditches typically entrain a small proportion (less than 4 percent) of the total estimated trout in the basin (Carlson and Rahel 2007, p. 1335) and that this represented a very small percentage of the total mortality for those populations (Post

et al.

2006, pp. 875, 884; Carlson and Rahel 2007, pp. 1335, 1339). Whether or not this amount of mortality can cause population instability is unclear (Post

et al.

2006, p. 886; Carlson and Rahel 2007, pp. 1340-1341). However, in some cases, even small vital rate changes in a trout population can theoretically cause population declines (Hilderbrand 2003, pp. 260-261).

The overall magnitude and population-level effect of entrainment on Arctic grayling in the Big Hole River

is unknown but possibly significant given the large number of unscreened surface-water diversions in the system and the large volumes of water diverted for irrigation. Given the low abundance of the species, even a small amount of entrainment may be biologically significant and is unlikely to be offset by compensatory effects (i.e., higher survival in Arctic grayling that are not entrained).

Entrainment also may be a problem for Arctic grayling at some locations within the Red Rock Lakes system (Unthank 1989, p. 10; Gillin 2001, pp. 2-4, 3-18, 3-25), particularly outside of the Red Rock Lakes NWR (Boltz 2010, pers. comm.).

Entrainment has been a past threat to Arctic grayling in the Big Hole River and the Red Rock Lakes system. It remains a current threat as most, if not all, irrigation diversions located in occupied habitat do not have any devices to exclude fish (i.e., fish screens). Entrainment will remain a threat in the foreseeable future unless diversion structures are modified to exclude fish. The Big Hole Grayling CCAA has provisions to reduce entrainment at diversions operated by enrolled landowners (MFWP

et al.

2006, pp. 50-52). Non-Federal landowners enrolled in the CCAA control approximately 50 to 70 percent of the points of irrigation diversion in the upper Big Hole River (Roberts and Lamothe 2010, pers. comm.), so the threat of entrainment in the Big Hole River should be significantly reduced in the foreseeable future. We consider the foreseeable future to represent approximately 20 years based on the duration of the Big Hole Grayling CCAA. Under the auspices of the Red Rock Lakes NWR CCP, a fish screen is planned to be installed on at least one diversion on the Red Rock Creek (USFWS 2009, p. 72), which is the primary spawning tributary for Arctic grayling in the Red Rock Lakes system. Overall, we anticipate it may take years to design and install fish screens on all the diversions that can entrain grayling in the Big Hole River and Red Rock Lakes systems; thus we conclude that entrainment remains a current threat that will continue to exist, but will decline in magnitude during the foreseeable future (next 10 to 20 years) because of implementation of the CCAA and CCP.

Degradation of Riparian Habitat

Riparian corridors are important for maintaining habitat for Arctic grayling in the upper Missouri River basin, and in general are critical for the ecological function of aquatic systems (Gregory

et al.

1991, entire). These riparian zones are important for Arctic grayling because of their effect on water quality and role in creating and maintaining physical habitat features (pools) used by the species.

Removal of willows and riparian clearing concurrent with livestock and water management along the Big Hole River has apparently accelerated in recent decades, and, in conjunction with streamside cattle grazing, has led to localized bank erosion, channel instability, and channel widening (Confluence Consulting

et al.

2003, pp. 24-26; Petersen and Lamothe 2006, pp. 16-17; Bureau of Land Management (BLM) 2009a, pp. 14-21). Arctic grayling abundance in the upper Big Hole River is positively related to the presence of overhanging vegetation, primarily willows, which are associated with pool habitat (Lamothe and Magee 2004b, pp. 21-22). Degradation of riparian habitat in the upper Big Hole River has led to a shift in channel form (from multiple threads to a single wide channel), increased erosion rates, reduced cover, increased water temperatures, and reduced recruitment of large wood into the active stream channel (Confluence Consulting

et al.

2003, pp. 24-26). All of these combine to reduce the suitability of the habitat for species like Arctic grayling, and likely reduce grayling growth, survival, and reproduction.

Livestock grazing both within the Red Rock Lakes NWR and on adjacent private lands has negatively affected the condition of riparian habitats on tributaries to the Red Rock Lakes (Mogen 1996, pp. 75-77; Gillin 2001, pp. 3-12, 3-14). In general, degraded riparian habitat limits the creation and maintenance of aquatic habitats, especially pools, that are preferred habitats for adult Arctic grayling (Lamothe and Magee 2004b, pp. 21-22; Hughes 1992, entire). Loss of pools likely reduces growth and survival of adult grayling. Loss of riparian vegetation increases bank erosion, which can lead to siltation of spawning gravels, which may in turn harm grayling by reducing the extent of suitable spawning habitat and reducing survival of Arctic grayling embryos already present in the stream gravels. The condition of riparian habitats upstream from the Upper and Lower Red Rock Lakes may have improved during the 1990s (Mogen 1996, p. 77), and ongoing efforts to improve grazing management and restore riparian habitats are ongoing both inside the Red Rock Lakes NWR (USFWS 2009, pp. 67, 75) and upstream (AGW 2010, p. 7; Korb 2010, pers. comm.). However, the existing condition of riparian habitats continues to constitute a threat to Arctic grayling because the loss of pool habitat and the deposition of fine sediments may take some time to be reversed after the recovery of riparian vegetation.

Much of the degradation of riparian habitats in the Big Hole River and Red Rock Lakes systems has occurred within the past 50 to 100 years, but the influence of these past actions continues to affect the structure and function of aquatic habitats in these systems. Thus, while the actual loss of riparian vegetation has presumably slowed during the past 10 years, the effect of reduced riparian vegetation continues to promote channel widening and sedimentation, and limits the creation and maintenance of pool habitats. Thus, degradation of riparian habitats is a current threat. Degradation of riparian habitats will remain a threat in the foreseeable future until riparian vegetation recovers naturally or through direct restoration, which may occur during the next 20 years in the Big Hole River and portions of the Red Rock Lakes system. Protection and direct restoration of riparian habitats in the Big Hole River is occurring on a fairly large scale under the provisions of the Big Hole Grayling CCAA (Lamothe

et al.

2007, pp. 13-26; Everett 2010, pp. 10-23), which should substantially reduce threats from riparian habitat degradation on private lands. Protection and restoration of riparian habitats implemented under the Red Rock Lakes NWR's CCP (see discussion under Factor D, below) should reduce threats from riparian habitat degradation within the NWR's boundary, but similar actions need to be taken on private lands adjacent to it (AGW 2010, p. 7; Korb 2010, pers. comm.) to appreciably reduce these threats in the foreseeable future and to expand the distribution of the species into formerly occupied habitat within that drainage.

Sedimentation

Sedimentation has been proposed as a mechanism behind the decline of Arctic grayling and its habitat in the Red Rock Lakes (Unthank 1989, p. 10; Mogen 1996, p. 76). Livestock grazing upstream has led to accelerated sediment transport in tributary streams, and deposition of silt in both stream and lakes has likely led to loss of fish habitat by filling in pools, covering spawning gravels, and reducing water depth in Odell and Red Rock Creeks, where Arctic grayling are still believed to spawn (MFWP 1981, p. 105; Mogen 1996, pp. 73-76).

Sedimentation in the Upper and Lower Red Rock Lakes is believed to

affect Arctic grayling by, in winter, reducing habitat volume (e.g., lakes freezing to the bottom) and promoting hypoxia (low oxygen), which generally concentrates fish in specific locations which have suitable depth, and thus increases the probability of competition and predation, and, in summer, causing thermal loading stress (see Dewatering From Irrigation and Consequent Increased Water Temperatures discussion, above). Depths in the Red Rock Lakes have decreased significantly, with a decline in maximum depth from 7.6 to 5.0 m (25 to 16.4 ft) to less than 2 m (6.5 ft) noted in Upper Red Rock Lake over the past century (Mogen 1996, p. 76). Lower Red Rock Lake has a maximum depth of approximately 0.5 m (1.6 ft), and freezes within a few inches of the bottom or freezes solid (Unthank 1989, p. 10). Consequently, the Lower Red Rock Lake does not appear to provide suitable overwintering habitat for adfluvial Arctic grayling and may be devoid of grayling except for the few individuals that may migrate between Odell Creek and Upper Red Rock Lake (Mogen, 1996, p. 47).

Dissolved oxygen levels in Upper Red Rock Lake during winter 1994-1995 dropped as low as 0.5 to 0.15 parts per million (ppm; Gangloff 1996, pp. 41-42, 72), well below the critical minimum of 1.3 to 1.7 ppm measured for adult Arctic grayling acclimated to water temperatures less than or equal to 8 °C (46 °F) (Feldmeth and Eriksen 1978, pp. 2042-2043). Thus, lethally low oxygen levels can occur during winter in Upper Red Rock Lake, the primary overwintering area for adfluvial Arctic grayling in the system. Winter kill of invertebrates and fishes (e.g., suckers

Catostomus

spp.) has been recorded in Upper Red Rock Lake (Gangloff 1996, pp. 39-40). Gangloff (1996, pp. 71, 79) hypothesized that Arctic grayling in Upper Red Rock Lake exhibit behavioral mechanisms or physiological adaptations that permit them to survive otherwise lethally low oxygen levels. Oxygen conditions in the lake during winter are related to the effect of snowpack and ice cover on light penetration and the density of macrophytes (rooted aquatic plants) during the preceding growing season (Gangloff 1996, pp. 72-74). Arctic grayling under winter ice seek areas of higher oxygen concentration (oxygen refugia) within the lake or near inlet streams of Upper Red Rock Lake (Gangloff 1996, pp. 78-79). Consequently, we expect factors leading to reduced lake depth due to upstream erosion and sedimentation within the lake, or factors that promote eutrophication due to macrophyte growth, to lead to more frequent winter hypoxia (low dissolved oxygen concentrations detrimental to aquatic organsims) in Upper Red Rock Lake, which is the most important overwintering habitat for adfluvial Arctic grayling in the system.

The effects of erosion and sedimentation on spawning gravels and reduction of habitat volume in Upper and Lower Red Rock Lakes are past and current threats. Improved land use may be reducing the rates of erosion in tributary streams (USFWS 2009, pp. 75-76; Korb 2010, pers. comm.). However, sedimentation of the lakes will likely remain a threat (because of reduced overwintering habitat, and high water temperatures in summer) in the foreseeable future unless some event mobilizes these sediments and transports them out of the lakes.

Protection and restoration of riparian habitats implemented under the Red Rock Lakes NWR's CCP (see discussion under Factor D, below) should reduce the magnitude of sedimentation within the NWR's boundaries, but similar actions need to be taken on private lands adjacent to it (AGW 2010, p. 7; Korb 2010, pers. comm.) to appreciably reduce threats in the foreseeable future.

Summary of Factor A

Based on the best available information, we find that the historical range of the Missouri River DPS of Arctic grayling has been greatly reduced, and the remaining native populations continue to face significant threats to their habitat. Large-scale habitat fragmentation by dams was likely a significant historical factor causing the range-wide decline of the DPS. The most significant current threats to the DPS are from land and water use activities that have affected the structure and function of aquatic systems, namely stream dewatering from irrigation withdrawals, which reduces habitat volume and increases summer water temperatures; potential loss of individuals in irrigation ditches (entrainment); degraded riparian habitats promoting erosion, sedimentation, increased water temperatures, and loss of pool habitat; and migration barriers that restrict movement to and from spawning, feeding, and sheltering habitats. These are among the significant current threats to Arctic grayling populations in the Big Hole River, Madison River-Ennis Reservoir, and Red Rock Lakes system. The habitat-related threats to the Big Hole River population should be reduced in the foreseeable future by implementation of the Big Hole Grayling CCAA, a formalized conservation plan with 32 private landowners currently enrolled. The Big Hole Grayling CCAA is expected to reduce threats from dewatering, high water temperatures, barriers to fish passage, and entrainment in irrigation ditches that are associated with land and water use in the upper Big Hole River watershed during the foreseeable future (next 20 years based on the duration of the CCAA). Non-Federal landowners enrolled in the Big Hole Grayling CCAA control or own approximately 50 to 70 percent of the points of irrigation diversion in the upper Big Hole River, so these landowners should have the ability to reduce habitat-related threats to Arctic grayling in the Big Hole River by a corresponding amount. However, the present or threatened destruction, modification, or curtailment of habitat remains a threat to the DPS overall. This factor is expected to continue to be a threat to the species in the foreseeable future because it is not comprehensively addressed for other populations, especially those in the Madison River and Red Rock Lakes systems where ongoing habitat-related threats (described above) may be making unoccupied habitat unsuitable for Arctic grayling, and may thus limit the recovery potential of the DPS.

B. Overutilization for Commercial, Recreational, Scientific, or Educational Purposes

Arctic grayling of the upper Missouri River are handled for recreational angling; and for scientific, population monitoring, and restoration purposes.

Recreational Angling

Arctic grayling are highly susceptible to capture by angling (ASRD 2005, pp. 19-20), and intense angling pressure can reduce densities and influence the demography of exploited populations (Northcote 1995, pp. 171-172). Overfishing likely contributed to the rangewide decline of the DPS in the upper Missouri River system (Vincent 1962, pp. 49-52, 55; Kaya 1992, pp. 54-55). In 1994, concern over the effects of angling on fluvial Arctic grayling led the State of Montana to implement catch-and-release regulations for Arctic grayling captured in streams and rivers within its native range, and those regulations remain in effect (MFWP 2010, p. 52). Catch-and-release regulations for Arctic grayling in the Big Hole River have been in effect since 1988 (Byorth 1993, p. 8). Catch-and-release regulations also are in effect for Ennis Reservoir on the Madison River (MFWP 2010, p. 61). Angling is not

permitted in either of the Red Rock Lakes to protect breeding waterfowl and trumpeter swans (

Cygnus buccinator

) (USFWS 2009, p. 147), and catch-and-release regulations remain in effect for any Arctic grayling captured in streams (e.g., Odell Creek or Red Rock Creek) in the Red Rock Lakes system (MFWP 2010, p. 56).

In Miner and Mussigbrod Lakes, anglers can keep up to 5 Arctic grayling per day and have up to 10 in possession, in accordance with standard daily and possession limits for that angling management district (MFWP 2010, p. 52). The current abundance of Arctic grayling in Mussigbrod Lake (see Table 4 above) suggests that present angling exploitation rates are not a threat to that population. Miner Lakes grayling are less abundant compared to Mussigbrod Lake, but we are not sure whether angling exploitation constitutes a threat to Miner Lakes grayling.

Repeated catch-and-release angling may harm individual fish, causing physiological stress and injury (i.e., hooking wounds). Catch-and-release angling also can result in mortality at a rate dependent on hooking location, hooking duration, fish size, water quality, and water temperature (Faragher

et al.

2004, entire; Bartholomew and Bohnsack 2005, p. 140). Repeated hooking (up to five times) of Arctic grayling in Alaska did not result in significant additional mortality (rates 0 to 1.4 percent; Clark 1991, pp. 1, 25-26). In Michigan, hooking mortality of Arctic grayling in lakes averaged 1.7 percent per capture event based on 355 individuals captured with artificial flies and lures (Nuhfer 1992, pp. 11, 29). Higher mortality rates (5 percent) have been reported for Arctic grayling populations in the Great Slave Lake area, Canada (Falk and Gillman 1975, cited in Casselman 2005, p. 23). Comparatively high catch rates for Arctic grayling have been observed in the Big Hole River, Montana (Byorth 1993, pp. 26-27, 36), and average hooking wound rates ranged from 15 to 30 percent among study sections (Byorth 1993, p. 28). However, overall hooking mortality from single capture events was low (1.4 percent), which led Byorth to conclude that the Big Hole River population was not limited by angling (Byorth 1994b, entire).

Compared to the average catch-and-release mortality rates of 4.2 to 4.5 percent in salmonids as reported by Schill and Scarpella (1997, p. 873), and the mean and median catch-and-release mortality rates of 18 percent and 11 percent from a meta-analysis of 274 studies (Bartholomew and Bohnsack 2005, pp. 136-137), the catch-and-release mortality rates for Arctic grayling are comparatively low (Clark 1991, pp. 1, 25-26; Nuhfer 1992, pp. 11, 29; Byorth 1994b, entire). We are uncertain whether these lower observed rates reflect an innate resistance to effects of catch-and-release angling in Arctic grayling or whether they reflect differences among particular populations or study designs used to estimate mortality. Even if catch-and-release angling mortality is low (e.g., 1.4 percent as reported in Byorth 1994b, entire), the high catchability of Arctic grayling (ASRD 2005, pp. 19-20) raises some concern about the cumulative mortality of repeated catch-and-release captures. For example, based on the Arctic grayling catch rates and angler pressure reported by Byorth (1993, pp. 25-26) and the population estimate for the Big Hole River reported in Byorth (1994a, p. ii), a simple calculation suggests that age 1 and older grayling susceptible to recreational angling may be captured and released 3 to 6 times per year.

The MFWP closes recreational angling in specific reaches of the Big Hole River when environmental conditions are considered stressful. Specific streamflow and temperature thresholds initiate mandatory closure of the fishery (Big Hole Watershed Committee 1997, entire). Such closures have been implemented in recent years. For example, the upper segment of the Big Hole River between Rock Creek Road to the confluence of the North Fork Big Hole River has been closed to angling at various times during 2004 (Magee

et al.

2005, p. 7), 2005 (Magee

et al.

2006, p. 20), and 2006 (Rens and Magee 2007, p. 8).

In conclusion, angling harvest may have significantly reduced the abundance and distribution of the upper Missouri River DPS of Arctic grayling during the past 50 to 100 years, but current catch-and-release fishing regulations (or angling closures) in most waters occupied by extant populations have likely ameliorated the past threat of overharvest. Although we have some concerns about the potential for cumulative mortality caused by repeated catch-and-release of individual Arctic grayling in the Big Hole River, we have no strong evidence indicating that repeated capture of Arctic grayling under catch-and-release regulations is currently limiting that population or the DPS. Moreover, fishing is restricted in the Big Hole River, an important recreational fishing destination in southwestern Montana, when streamflow and temperature conditions are likely to increase stress to captured grayling. Anglers can still capture and keep Arctic grayling in Miner and Mussigbrod Lakes in accordance with State fishing regulations, but we have no evidence that current levels of angling are affecting these populations. We thus have no evidence that recreational angling represents a current threat to the DPS. If we assume that future fishing regulations would be at least as conservative as current regulations, and that the current levels of angling pressure will continue, then recreational angling does not represent a threat in the foreseeable future.

Monitoring and Scientific Study

The MFWP consistently monitors the Arctic grayling population in the Big Hole River and its tributaries, and to a lesser extent those populations in the Madison River and Red Rock Lakes system (Rens and Magee 20007, entire). Electrofishing (use of electrical current to temporarily and non-lethally immobilize a fish for capture) is a primary sampling method to monitor Arctic grayling in the Big Hole River, Madison River, and Red Rock Lakes (Rens and Magee 2007, pp. 13, 17, 20). A number of studies have investigated the effects of electrofishing on various life stages of Arctic grayling. Dwyer and White (1997, p. 174) found that electrofishing reduced the growth of juvenile Arctic grayling and concluded that long-term, sublethal effects of electrofishing were possible. Hughes (1998, pp. 1072, 1074-1075) found evidence that electrofishing and tagging affected the growth rate and movement behavior of Arctic grayling in the Chena River, Alaska. Roach (1999, p. 923) studied the effects of electrofishing on fertilized Arctic grayling eggs and found that while electrofishing could result in egg mortality, the population-level effects of such mortality were not likely to be significant. Lamothe and Magee (2003, pp. 16, 18-19) noted mortality of Arctic grayling in the Big Hole River during a radio-telemetry study, and concluded that handling stress or predation were possible causes of mortality. Population monitoring activities in the Big Hole River are curtailed when environmental conditions become unsuitable (Big Hole Watershed Committee 1997, entire), and recent monitoring reports (Magee and Lamothe 2004, entire; Magee

et al.

2005, entire; Rens and Magee 2007, entire) provide no evidence that electrofishing is harming the Arctic grayling population in the Big Hole River.

A study in the Big Hole River is investigating the availability and use of coldwater thermal refugia for Arctic grayling and other resident fishes (Vatland and Gressewell 2009, entire).

The study uses fish tagged with passive integrated transponder (PIT) tag technology to record movement past receiving antennas. The PIT tags are small (23 mm or less than 1 in. long) and implanted into the body cavity of the fish during a quick surgical procedure. During 2007-2008, a total of 81 Arctic grayling from the Big Hole River and its tributaries were implanted with these PIT tags (Vatland and Gressewell 2009, p. 12). A short-term study on the potential effects of PIT tag implantation on Arctic grayling found 100 percent retention of tags and 100 percent survival of tagged individuals during a 4-day trial (Montana State University 2008, p. 7). Based on the results of the controlled trials, we have no evidence to indicate that PIT tagging the wild Arctic grayling in the Big Hole River constitutes a significant threat to the population.

Traps, electrofishing, and radio telemetry have been used to monitor and study Arctic graying in the Red Rock Lakes system (Gangloff 1996, pp. 13-14; Mogen 1996, pp. 10-13, 15; Kaeding and Boltz 1999, p. 4; Rens and Magee 2007, p. 17); however, there is no data to indicate these monitoring activities reduce the growth and survival of individual Arctic grayling or otherwise constitute a current or future threat to the population.

The Arctic grayling population in the Madison River-Ennis Reservoir is not monitored as intensively as the Big Hole River population (Rens and Magee 2007, pp. 20-21). When electrofishing surveys targeting Arctic grayling in the Madison River do occur, they are conducted during the spawning run for that population (Clancey 1996, p. 6). Capture and handling during spawning migrations or during actual spawning could affect the reproductive success of individual Arctic grayling. However, under recent monitoring frequencies, any population-level effect of these activities is likely negligible, and we have no data to indicate these monitoring activities reduce the growth and survival of individual Arctic grayling or otherwise constitute a current or future threat to the Madison River population.

The Miner Lakes and Mussigbrod Lake populations of Arctic grayling are infrequently monitored (Olsen 2010, pers. comm.). Since monitoring of these populations has been minimal, we do not believe that monitoring or scientific study constitutes a current or foreseeable threat to these particular populations.

The intensity of monitoring and scientific investigation varies among the different populations in the DPS, but we have no evidence suggesting that monitoring or scientific study has influenced the decline of Arctic grayling in the Missouri River basin. We also have no evidence indicating these activities constitute a current threat to the DPS that would result in measurable, population-level effects. We expect similar levels of population monitoring and scientific study in the future, and we have no basis to conclude that these activities represent a threat in the foreseeable future.

Reintroduction Efforts

Attempts to restore or re-establish native populations of both fluvial and adfluvial Arctic grayling may result in the mortality of embryos and young fish. The MFWP attempted to restore fluvial Arctic graying to historic waters in the upper Missouri River using a combination of stocking and embryo incubating devices (remote site incubators) placed in target streams (Rens and Magee 2007, pp. 24-38). Currently, gametes (eggs and sperm) used to re-establish the fluvial ecotype come from captive brood reserves of Big Hole River grayling maintained in Axolotl and Green Hollow II Lakes (Rens and Magee 2007, pp. 22-24). Removal of gametes from the wild Big Hole River population was necessary to establish this brood reserve (Leary 1991, entire). The previous removal of gametes for conservation purposes may have reduced temporarily the abundance of the wild population if the population was unable to compensate for this effective mortality by increased survival of remaining individuals. However, the establishment of a brood reserve provides a conservation benefit from the standpoint that gametes from the reserve can be harvested to use for translocation efforts to benefit the species. Unfortunately, these translocations have not yet resulted in establishment of any fluvial populations. Ultimately, we do not have any data to indicate that past gamete collection from the Big Hole River population harmed the wild population. Consequently, we have no basis to conclude that gamete collection from the wild Big Hole River Arctic grayling population constitutes a current or future threat to the population.

Efforts to re-establish native, genetically pure populations of adfluvial Arctic grayling in the Red Rock Lakes system and to maintain a brood reserve for that population have resulted in the direct collection of eggs from Arctic grayling spawning runs in Red Rock Creek. During 2000-2002, an estimated 315,000 Arctic grayling eggs were collected from females captured in Red Rock Creek (Boltz and Kaeding 2002, pp. v, 8). The Service placed over 180,000 of these eggs in remote site incubators in streams within the Red Rock Lakes NWR that historically supported Arctic grayling spawning runs (Boltz and Kaeding 2002, pp. v, 10). Despite preliminary observations of grayling spawning in historically occupied waters within the Red Rock Lakes NWR following the use of remote site incubators (Kaeding and Boltz 2004, pp. 1036), spawning runs at these locations have apparently not become established (Boltz 2006, pers. comm.). Attempts to establish a brood reserve of adfluvial Arctic grayling within the NWR's boundaries (MacDonald Pond) were not successful (Boltz and Kaeding 2002, pp. 21-22). Red Rock Lakes NWR plans to re-establish Arctic grayling in Elk Springs and Picnic Creeks and establish a brood stock in Widgeon Pond as part of its CCP (USFWS 2009, pp. 72, 75). The MFWP and the Service are currently collaborating on an effort to re-establish an Arctic grayling spawning run in Elk Springs Creek and to establish a genetically pure brood reserve of Red Rock Lakes grayling in Elk Lake as no such population exists for use in conservation and recovery (Jordan 2010, pers. comm.). These actions will require the collection of gametes (approximately 360,000 eggs) from Arctic grayling captured in Red Rock Creek (Jordan 2010, pers. comm.). Approximately 10 percent of these eggs will be returned to Red Rock Creek and incubated in that stream (using a remote site incubation method that results in high survivorship of embryos) (Kaeding and Boltz 2004, entire) to mitigate for collection of gametes from the wild spawning population (Jordan 2010, pers. comm.). We presume these ongoing actions may necessitate the collection of gametes from wild Arctic grayling in Red Rock Creek, so the potential effect of such collections on the extant wild population should be evaluated and mitigation for the use of these gametes (e.g., using remote site incubators at the collection source or another method) should continue.

Overall, we have no evidence to indicate that collection of gametes from the wild populations in the Big Hole River and Red Rock Lakes systems have contributed to population-level declines in those populations, or that the p

This text is long and has been trimmed here. Open the source document for the complete record.