Endangered and Threatened Wildlife and Plants; Threatened Status for the Northern Mexican Gartersnake and Narrow-Headed Gartersnake

DEPARTMENT OF THE INTERIOR

Fish and Wildlife Service

50 CFR Part 17

[Docket No. FWS-R2-ES-2013-0071: 4500030113]

RIN 1018-AY23

Endangered and Threatened Wildlife and Plants; Threatened Status for the Northern Mexican Gartersnake and Narrow-Headed Gartersnake

AGENCY:

Fish and Wildlife Service, Interior.

ACTION:

Final rule.

SUMMARY:

We, the U.S. Fish and Wildlife Service (Service), determine threatened species status under the Endangered Species Act of 1973 (Act), as amended, for the northern Mexican gartersnake (

Thamnophis eques megalops

) and the narrow-headed gartersnake (

Thamnophis rufipunctatus

), native species from Arizona and New Mexico in the United States. We also finalize a rule under authority of section 4(d) of the Endangered Species Act of 1973, as amended (Act), that provides measures that are necessary and advisable to provide for the conservation of the northern Mexican gartersnake. Both species are listed as threatened throughout their range, which, for the northern Mexican gartersnake, also includes the Mexican states of Sonora, Chihuahua, Durango, Coahuila, Zacatecas, Guanajuato, Nayarit, Hidalgo, Jalisco, San Luis Potosí, Aguascalientes, Tlaxacala, Puebla, México, Veracruz, and Querétaro. The effect of this regulation will be to add these species to the lists of Endangered and Threatened Wildlife and Plants.

DATES:

This rule becomes effective August 7, 2014.

ADDRESSES:

This final rule is available on the internet at

http://www.regulations.gov

(Docket No. FWS-R2-ES-2013-0071) and

http://www.fws.gov/southwest/es/arizona.

Comments and materials we received, as well as supporting documentation we used in preparing this rule, are available for public inspection at

http://www.regulations.gov.

All of the comments, materials, and documentation that we considered in this rulemaking are available by appointment, during normal business hours at: U.S. Fish and Wildlife Service, Arizona Ecological Services Field Office, 2321 West Royal Palm Road, Suite 103, Phoenix, AZ 85021; telephone: 602-242-0210; facsimile: 602-242-2513.

FOR FURTHER INFORMATION CONTACT:

Steve Spangle, Field Supervisor, U.S. Fish and Wildlife Service, Arizona Ecological Services Field Office, 2321 West Royal Palm Road, Suite 103, Phoenix, AZ 85021; telephone: 602-242-0210; facsimile: 602-242-2513. Persons who use a telecommunications device for the deaf (TDD) may call the Federal Information Relay Service (FIRS) at 800-877-8339.

SUPPLEMENTARY INFORMATION:

Executive Summary

Why we need to publish a rule.

Under the Endangered Species Act, a species may warrant protection through listing if it is endangered or threatened throughout all or a significant portion of its range. Listing a species as an endangered or threatened species requires issuing a rule. This rule will finalize the listing of the northern Mexican gartersnake (

Thamnophis eques megalops

) and narrow-headed gartersnake (

Thamnophis rufipunctatus

) as threatened species, initiated with our proposed listing rule published on July 10, 2013 (78 FR 41500), and finalize a rule under authority of section 4(d) of the Act that provides measures that are necessary and advisable to provide for the conservation of the northern Mexican gartersnake.

The basis for our action.

Under the Endangered Species Act, we can determine that a species is an endangered or threatened species based on any of 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. We have determined that predation from and competition with nonnative species such as bass (

Micropterus

sp.), flathead catfish (

Pylodictis

sp.), channel catfish (

Ictalurus

sp.), Chihuahuan catfish (

Ictalurus chihuahua

), bullheads (

Ameiurus

sp.), sunfish (

Lepomis

sp.), and crappie (

Pomoxis

sp.), brown trout (

Salmo trutta

), American bullfrogs (

Lithobates catesbeiana

), and crayfish (northern (virile) crayfish (

Orconectes virilis

) and red swamp crayfish (

Procambarus clarkia

)) are the most significant threat affecting these gartersnakes across their range. Throughout the remainder of this final rule, the nonnative species identified immediately above will be referred to collectively as “harmful nonnative species.” Large-scale wildfires and land uses that divert, dry up, or significantly pollute aquatic habitat have also been found to be significant threats. Collectively, these threats have adversely affected gartersnake populations, and most of their native prey species, such that the gartersnakes' resiliency, redundancy, and representation across their ranges have been significantly compromised.

Peer review and public comment.

We sought comments from independent specialists to ensure that our designation is based on scientifically sound data, assumptions, and analyses. We invited these peer reviewers to comment on our listing proposal. We also considered all other comments and information received during the comment period on the proposed listing rule. All comments are available at

http://www.regulations.gov

(Docket No. FWS-R2-ES-2013-0071).

Previous Federal Action

Please refer to the proposed listing rule for the northern Mexican gartersnake and narrow-headed gartersnake (78 FR 41500; July 10, 2013) for a detailed description of previous Federal actions concerning this species.

We will also be finalizing the designation of critical habitat for the northern Mexican gartersnake and narrow-headed gartersnake in a separate rule in the future. Information regarding designation of critical habitat for these species is available at

http://www.regulations.gov

(Docket No. FWS-R2-ES-2013-0022).

Background

Northern Mexican Gartersnake

Subspecies Description

The northern Mexican gartersnake ranges in color from olive to olive-brown or olive-gray with three lighter-colored stripes that run the length of the body, the middle of which darkens toward the tail. This species may inhabit the same area as other native gartersnake species and can be difficult for people without specific expertise to identify. The snake may reach a maximum known length of 44 inches (in) (112 centimeters (cm)). The pale yellow to light-tan lateral (side of body) stripes distinguish the northern Mexican gartersnake from other sympatric (co-occurring) gartersnake species because a portion of the lateral stripe is found on the fourth scale row, while it is confined to lower scale rows for other species. Paired black spots extend along the olive dorsolateral fields (region adjacent to the top of the

snake's back) and the olive-gray ventrolateral fields (region adjacent to the area of the snake's body in contact with the ground). The scales are keeled (possessing a ridge down the center of each scale). A more detailed subspecies description can be found in our September 26, 2006 (71 FR 56227), or November 25, 2008 (73 FR 71788) 12-month findings for this subspecies, or by reviewing Rosen and Schwalbe (1988, p. 4), Rossman

et al.

(1996, pp. 171-172), Ernst and Ernst (2003, pp. 391-392), or Manjarrez and Garcia (1993, pp. 1-5).

Taxonomy

The northern Mexican gartersnake (

Thamnophis eques megalops

) is a member of the family Colubridae and subfamily Natricinae (harmless live-bearing snakes) (Lawson

et al.

2005, p. 596; Pyron

et al.

2013, p. 31). The taxonomy of the genus

Thamnophis

has a complex history, partly because many of the species are similar in appearance and arrangement of scales and many of the early museum specimens were in such poor and faded condition that it was difficult to study them (Conant 2003, p. 6).

Prior to 2003,

Thamnophis eques

was considered to have three subspecies,

T. e. eques, T. e. megalops,

and

T. e. virgatenuis

(Rossman

et al.

1996, p. 175). In 2003, an additional seven new subspecies were identified under

T. eques:

(1)

T. e. cuitzeoensis;

(2)

T. e. patzcuaroensis;

(3)

T. e. insperatus;

(4)

T. e. obscurus;

(5)

T. e. diluvialis;

(6)

T. e. carmenensis;

and (7)

T. e. scotti

(Conant 2003, p. 3). Common names were not provided, so in this final rule, we use the scientific name for all subspecies of Mexican gartersnake other than the northern Mexican gartersnake. These seven new subspecies were described based on morphological differences in coloration and pattern, have highly restricted distributions, and occur in isolated wetland habitats within the mountainous Transvolcanic Belt region of southern Mexico, which contains the highest elevations in the country (Conant 2003, pp. 7-8). Additional information regarding this subspecies' taxonomy can be found in de Queiroz

et al.

(2002, p. 323), de Queiroz and Lawson (1994, p. 217), Rossman

et al.

(1996, pp. xvii-xviii, 171-175), Rosen and Schwalbe (1988, pp. 2-3), Liner (1994, p. 107), and Crother

et al.

(2012, p. 70). A description of the taxonomy of the northern Mexican gartersnake is found in our September 26, 2006 (71 FR 56227) and November 25, 2008 (73 FR 71788) 12-month findings for this subspecies.

Habitat and Natural History

Throughout its rangewide distribution, the northern Mexican gartersnake occurs at elevations from 130 to 8,497 feet (ft) (40 to 2,590 meters (m)) (Rossman

et al.

1996, p. 172) and is considered a “terrestrial-aquatic generalist” (Drummond and Marcías-García 1983, pp. 24-26). The northern Mexican gartersnake is a riparian obligate (generally found in riparian areas when not engaged in dispersal, gestation, or hibernation behaviors) and occurs chiefly in the following general habitat types: (1) Small, often isolated wetlands (e.g., cienegas (mid-elevation wetlands with highly organic, reducing (basic or alkaline) soils), or stock tanks (small earthen impoundment)); (2) large-river riparian woodlands and forests; and (3) streamside gallery forests (as defined by well-developed broadleaf deciduous riparian forests with limited, if any, herbaceous ground cover or dense grass) (Hendrickson and Minckley 1984, p. 131; Rosen and Schwalbe 1988, pp. 14-16). Emmons and Nowak (2013, p. 14) found this subspecies most commonly in protected backwaters, braided side channels and beaver ponds, isolated pools near the river mainstem, and edges of dense emergent vegetation that offered cover and foraging opportunities when surveying in the upper and middle Verde River region. Additional information on the habitat requirements of the northern Mexican gartersnake within the United States and Mexico can be found in our 2006 (71 FR 56227) and 2008 (73 FR 71788) 12-month findings for this subspecies and in Rosen and Schwalbe (1988, pp. 14-16), Rossman

et al.

(1996, p. 176), McCranie and Wilson (1987, pp. 11-17), Ernst and Ernst (2003, p. 392), and Cirett-Galan (1996, p. 156).

The northern Mexican gartersnake is surface active at ambient (air) temperatures ranging from 71 degrees Fahrenheit (°F) to 91 °F (22 degrees Celsius (°C) to 33 °C) and forages along the banks of waterbodies (Rosen 1991, p. 305, Table 2). While conducting visual surveys, Rosen (1991, pp. 308-309) found that northern Mexican gartersnakes spent up to 60 percent of their time moving, 13 percent of their time basking on vegetation, 18 percent of their time basking on the ground, and 9 percent of their time under surface cover. However, preliminary telemetry data from a population of northern Mexican gartersnakes at the Bubbling Ponds State Fish Hatchery show individuals were surface active during 16 percent of telemetry observations, not surface active during 64 percent of telemetry observations, and surface activity was undetermined for 20 percent of the telemetry observations (Boyarsky 2013, pers. comm.); at Tavasci Marsh along the upper Verde River, they were inactive 60 percent of the time (Emmons 2013b, pers. comm.). In the northern-most part of its range, the northern Mexican gartersnake appears to be most active during July and August, followed by June and September (Emmons and Nowak 2013, p. 14). Northern Mexican gartersnakes may use different sites as hibernacula during a single cold-season and will bask occasionally (Emmons 2014, pers. comm.).

Although considered a highly aquatic species, the northern Mexican gartersnake uses terrestrial habitat for hibernation (Young and Boyarski 2012b, pp. 25-28), gestation, seeking mates, and dispersal. Along the middle Verde River preliminary telemetry data for the northern Mexican gartersnake found that the species may travel at least 528 feet (161 m) from the nearest water and as much as 0.4 mi (0.6 km) in a single day (total distance traveled) (Emmons 2014, pers. comm.). Terrestrial habitat use in open, grassland-dominated landscapes with scattered livestock tanks, such as in southern Arizona, may reflect that greater distances are traveled as suggested by the observation of a large female northern Mexican gartersnake observed in O'Donnell Canyon, which was far from source populations and may have been dispersing overland (Rosen and Schwalbe 1988, p. 14). Preliminary data from the population at Bubbling Ponds State Fish Hatchery show that home ranges vary from 1.7 acres (0.7 ha) to 10.4 acres (4.2 ha), with a mean home range size of 6.2 acres (2.51 ha) (Young and Boyarski 2012b, p. 23).

The northern Mexican gartersnake is an active predator and depends on smaller animals for its prey base (Rosen and Schwalbe 1988, pp. 18, 20). Northern Mexican gartersnakes forage along vegetated banklines, searching for prey in water and on land, using different strategies (Alfaro 2002, p. 209), or may forage along the edges of open water and thick stands of vegetation such as cattails. Generally, its diet consists of native amphibians and fishes, such as adult and larval (tadpoles) native leopard frogs (e.g., lowland leopard frog (

Lithobates yavapaiensis

) and Chiricahua leopard frog (

Lithobates chiricahuensis

)), as well as juvenile and adult native fish species (e.g., Gila topminnow (

Poeciliopsis occidentalis occidentalis

), desert pupfish (

Cyprinodon macularius

), Gila chub (

Gila intermedia

), and roundtail chub (

Gila robusta

)) (Rosen and

Schwalbe 1988, p. 18). Drummond and Marcías-García (1983, pp. 25, 30) found that as a subspecies, Mexican gartersnakes fed primarily on frogs. The northern Mexican gartersnake may congregate at ephemeral amphibian breeding ponds to exploit high-density prey populations as observed at New Mexican spadefoot toads (

Spea multiplicata

) breeding sites (d'Orgeix

et al.

2013, pp. 213-215). Auxiliary prey items may also include young Woodhouse's toads (

Anaxyrus woodhousei

), treefrogs (Family Hylidae), earthworms, deermice (

Peromyscus

spp.), lizards of the genera

Aspidoscelis

and

Sceloporus,

larval tiger salamanders (

Ambystoma tigrinum

), and leeches (Rosen and Schwalbe 1988, p. 20; Holm and Lowe 1995, pp. 30-31; Degenhardt

et al.

1996, p. 318; Rossman

et al.

1996, p. 176; Manjarrez 1998, p. 465). Salamanders (

Ambystoma

spp.) may be particularly important as prey for northern Mexican gartersnake populations in northern Mexico, both at lower elevations and along the Sierra Madre Occidental (Lemos-Espinal 2013, pers. comm.).

In situations where native prey species are rare or absent, this snake's diet may be almost completely comprised of nonnative species, including larval and juvenile bullfrogs (

Lithobates catesbeianus

), mosquitofish (

Gambusia affinis

) (Holycross

et al.

2006, p. 23), or subadult green sunfish, bluegill, or largemouth bass (Emmons and Nowak 2013, p. 5; Emmons 2013a, pers. comm.). The most recent observations of northern Mexican gartersnakes attempting to eat predatory fish was discussed in Emmons and Nowak (2013, p. 6) where they found fish inside traps with gartersnakes, and the fish appeared to have been partially consumed and then regurgitated. These observations suggest that, while northern Mexican gartersnakes may attempt to eat predatory fish (at least in the artificial confines of a wire trap), they may often be spontaneously regurtitated, potentially causing harm to the snake (Nowak and Santana-Bendix 2002, p. 24), and may not be compatible prey for northern Mexican gartersnakes. Interestingly, in a 2012 trapping effort along the upper Santa Cruz River, minnow traps that become self-baited with bullfrogs, mosquitofish, or macroinvertebrates captured snakes, but those which contained green sunfish or largemouth bass never caught a single northern Mexican gartersnake (Lashway 2012, p. 6).

Chinese mystery snails (

Cipangopaludina chinensis

) have also been reported as a prey item for northern Mexican gartersnakes at the Page Springs and Bubbling Ponds State Fish Hatcheries in Arizona, but some predation attempts on snails have proven fatal for gartersnakes because of their lower jaw becoming permanently lodged in the snails' shell (Young and Boyarski 2012a, p. 498). Venegas-Barrera and Manjarrez (2001, p. 187) reported the first observation of a snake in the natural diet of any species of

Thamnophis

after documenting the consumption of a Mexican alpine blotched gartersnake (

Thamnophis scalaris

) by a Mexican gartersnake (

T. eques;

subspecies not reported); a behavior termed ophiophagy. Ophiophagy has not been specifically reported in northern Mexican gartersnakes, although they are a subspecies of the Mexican gartersnake.

Marcías-García and Drummond (1988, pp. 129-134) sampled the stomach contents of Mexican gartersnakes and the prey populations at (ephemeral) Lake Tecocomulco, Hidalgo, Mexico. Field observations indicated, with high statistical significance, that larger Mexican gartersnakes fed primarily upon aquatic vertebrates (fishes, frogs, and larval salamanders) and leeches, whereas smaller Mexican gartersnakes fed primarily upon earthworms and leeches (Marcías-García and Drummond 1988, p. 131). Marcías-García and Drummond (1988, p. 130) also found that the birth of newborn

T. eques

tended to coincide with the annual peak density of annelids (earthworms and leeches). There is also preliminary evidence that birth may coincide with a pronounced influx of available prey in a given area, especially with that of explosive breeders, such as toads, but more research is needed to confirm such a relationship (Boyarski 2012, pers. comm.). Positive correlations were also made with respect to capture rates (which are correlated with population size) of

T. eques

to lake levels and to prey scarcity; that is, when lake levels were low and prey species scarce, Mexican gartersnake capture rates declined (Marcías-García and Drummond 1988, p. 132). While prey scarcity could have driven snakes to become active or take shelter underground, their results suggest the importance of available water and an adequate prey base to maintaining viable populations of Mexican gartersnakes. Marcías-García and Drummond (1988, p. 133) found that, while certain prey items were positively associated with size classes of snakes, the largest of specimens consume any prey available.

Native predators of the northern Mexican gartersnake include birds of prey, other snakes (kingsnakes (

Lampropeltis

sp.), whipsnakes (

Coluber

sp.), regal ring-necked snakes (

Diadophis punctatus regalis

), etc.), wading birds, mergansers (

Mergus merganser

), belted kingfishers (

Megaceryle alcyon

), raccoons (

Procyon lotor

), skunks (

Mephitis

sp.), and coyotes (

Canis latrans

) (Rosen and Schwalbe 1988, pp. 18, 39; Brennan

et al.

2009, p. 123). Historically, large, highly predatory native fish species such as Colorado pikeminnow (

Ptychocheilus lucius

) may have preyed upon northern Mexican gartersnake where the subspecies co-occurred. Native chubs (

Gila

sp.) may also prey on neonatal gartersnakes, but has not been documented in the literature to our knowledge.

Sexual maturity in northern Mexican gartersnakes occurs at 2 years of age in males and at 2 to 3 years of age in females (Rosen and Schwalbe 1988, pp. 16-17). Northern Mexican gartersnakes are viviparous (bringing forth living young rather than eggs). Mating has been documented in April and May followed by the live birth of between 7 and 38 newborns (average is 13.6) in June, July, and August (Rosen and Schwalbe 1988, p. 16; Nowak and Boyarski 2012, pp. 351-352; Boyarski 2013, pers. comm.). However, field observations in Arizona provide preliminary evidence that mating may also occur during the fall, but further research is required to confirm this hypothesis (Boyarski 2012, pers. comm.). Unlike other gartersnake species, which typically breed annually, one study suggests that only half of the sexually mature females within a population of northern Mexican gartersnake might reproduce in any one season (Rosen and Schwalbe 1988, p. 17). We found no information on the longevity of northern Mexican gartersnakes but presume they may live as long as 10 years in the wild.

Historical Distribution

Within the United States, the northern Mexican gartersnake historically occurred predominantly in Arizona at elevations ranging from 130 to 6,150 ft (40 to 1,875 m). It was generally found where water was relatively permanent and supported suitable habitat. The northern Mexican gartersnake has been documented historically in every county and nearly every subbasin within Arizona, but its historical distribution was essentially the southern two-thirds of Arizona. It was known from several perennial or intermittent creeks, streams, and rivers as well as lentic (still, non-flowing water) wetlands such as cienegas, ponds, or stock tanks. Records documenting northern Mexican

gartersnake exist within the following subbasins in Arizona: Colorado River, Bill Williams River, Agua Fria River, Salt River, Tonto Creek, Verde River, Santa Cruz River, Cienega Creek, San Pedro River, Babocomari River, and the Rio San Bernardino (Black Draw) (Woodin 1950, p. 40; Nickerson and Mays 1970, p. 503; Bradley 1986, p. 67; Rosen and Schwalbe 1988, Appendix I; 1995, p. 452; 1997, pp. 16-17; Holm and Lowe 1995, pp. 27-35; Sredl

et al.

1995b, p. 2; 2000, p. 9; Rosen

et al.

2001, Appendix I; Holycross

et al.

2006, pp. 1-2, 15-51; Brennan and Holycross 2006, p. 123; Radke 2006, pers. comm.; Rosen 2006, pers. comm.; Holycross 2006, pers. comm.; Cotton

et al.

2013, p. 111). Numerous records for the northern Mexican gartersnake (through 1996) in Arizona are maintained in the Arizona Game and Fish Department's (AGFD) Heritage Database (1996a).

Historically, the northern Mexican gartersnake had a limited distribution in New Mexico that consisted of scattered locations throughout the Upper Gila River watershed in Grant and western Hidalgo Counties, including the Upper Gila River, Mule Creek in the San Francisco River subbasin, and the Mimbres River (Price 1980, p. 39; Fitzgerald 1986, Table 2; Degenhardt

et al.

1996, p. 317; Holycross

et al.

2006, pp. 1-2).

One record for the northern Mexican gartersnake exists for the State of Nevada, opposite Fort Mohave, in Clark County along the shore of the Colorado River that was dated 1911 (De Queiroz and Smith 1996, p. 155). The subspecies may have occurred historically in the lower Colorado River region of California, although we were unable to verify any museum records for California. Any populations of northern Mexican gartersnakes that may have historically occurred in either Nevada or California were likely associated directly with the Colorado River, and we believe the northern Mexican gartersnake to be currently extirpated in Nevada and California.

Within Mexico, northern Mexican gartersnakes historically occurred within the Sierra Madre Occidental and the Mexican Plateau in the Mexican states of Sonora, Chihuahua, Durango, Coahuila, Zacatecas, Guanajuato, Nayarit, Hidalgo, Jalisco, San Luis Potosí, Aguascalientes, Tlaxacala, Puebla, México, Veracruz, and Querétaro, comprising approximately 85 percent of the total rangewide distribution of the subspecies (Conant 1963, p. 473; 1974, pp. 469-470; Van Devender and Lowe 1977, p. 47; McCranie and Wilson 1987, p. 15; Rossman

et al.

1996, p. 173; Lemos-Espinal

et al.

2004, p. 83). We are not aware of any systematic, rangewide survey effort for the northern Mexican gartersnake in Mexico. Therefore, we use other related ecological surrogates (such as native freshwater fish) to inform discussion on the status of aquatic communities and aquatic habitat in Mexico, and therefore on the likely status of northern Mexican gartersnake populations. We believe that gartersnakes and native fish are closely ecologically connected because of the high level of dependency of the gartersnakes on the fish as a food source. This discussion is found below in the subheadings pertinent to Mexico.

Current Distribution and Population Status

Data on population status of northern Mexican gartersnakes in the United States are largely summarized in unpublished agency reports. In our literature review we found that reductions in range and population densities have affected the status of the northern Mexican gartersnake significantly in the last 30 years. We found that, in as much as 90 percent of the northern Mexican gartersnakes' historical distribution in the United States, the subspecies occurs at low to very low population densities or may even be extirpated. For example, Holycross

et al.

(2006, p. 66) detected the northern Mexican gartersnake at only 2 of 11 historical localities within the northern-most part of its range in the United States. The degraded status of the northern Mexican gartersnake, in a rangewide context, is primarily the result of predation by and competition with harmful nonnative species, that have been legally released, illegally released, or have naturally dispersed (explained below). However, ecological circumstances and potential threats vary from site to site, and the same threats do not affect every population with the same magnitude across their range. Regardless of how they got into the wild, harmful nonnative species are now widespread and present throughout the range of the northern Mexican gartersnake. Land uses that result in the dewatering of habitat, combined with increasing drought, have destroyed significant amounts of habitat throughout the northern Mexican gartersnake's range and have, therefore, reduced its distribution within several subbasins.

Where northern Mexican gartersnakes are locally abundant, they are usually reliably detected with significantly less effort than populations characterized as having low densities. Northern Mexican gartersnakes are well-camouflaged, secretive, and can be very difficult to detect in structurally complex, dense habitat (Emmons and Nowak 2013, p. 13) or where they occur at very low population densities, which characterizes most occupied sites in lotic habitat. We considered factors such as the date of the last known records for northern Mexican gartersnakes in an area, as well as records of one or more native prey species in making a conclusion on occupancy of the subspecies. We used the year 1980 to qualify occupancy because the 1980s marked the first systematic survey efforts for northern Mexican gartersnakes across their range in the United States (see Rosen and Schwalbe (1988, entire) and Fitzgerald (1986, entire)) and the last, previous records were often dated several decades prior and may not accurately represent the likelihood for current occupation. Several areas where northern Mexican gartersnakes were known to occur have received no, or very little, survey effort in the past several decades. Variability in survey design and effort makes it difficult to compare population sizes or trends among sites and between sampling periods. For each of the sites discussed in Appendix A (available at

http://www.regulations.gov,

Docket No. FWS-R2-ES-2013-0071), we have attempted to translate and quantify search and capture efforts into comparable units (

i.e.,

person-search hours and trap-hours) and have cautiously interpreted those results. Because the presence of suitable prey species in an area may provide evidence that the northern Mexican gartersnake may still persist in low density where survey data are sparse, a record of a native prey species was considered in our determination of occupancy of this subspecies.

Currently, there are only five northern Mexican gartersnake populations in the United States, where the subspecies remains reliably detected and is considered viable, and all are located in Arizona. The five known populations are: (1) The Page Springs and Bubbling Ponds State Fish Hatcheries along Oak Creek, (2) lower Tonto Creek, (3) the upper Santa Cruz River in the San Rafael Valley, (4) the Bill Williams River, and (5) the upper and middle Verde River. In New Mexico, the northern Mexican gartersnake was last documented in 2013 along the Gila River in the vicinity of the Highway 180 crossing (Hotle 2013, entire) and is considered to occur in extremely low population densities within its historical distribution along the Gila River and Mule Creek. While

historically known to occur on tribal lands, the status of the northern Mexican gartersnake on tribal lands, such as those owned by the White Mountain or San Carlos Apache Tribes, is poorly known due to limited survey access. As stated previously, less is known specifically about the current distribution of the northern Mexican gartersnake in Mexico due to limited access to information on survey efforts and field data from Mexico.

In Table 1 below, we summarize the population status of northern Mexican gartersnakes at all known 29 historical localities throughout their United States distribution, as supported by museum records or reliable observations. We categorized each population as either likely viable, likely not viable, or likely extirpated based on the historical survey records, suitable habitat, presence of native prey species, and the presence of harmful nonnative species. For a detailed discussion that explains the rationale for site-by-site conclusions on occupancy, please see Appendix A (available at

http://www.regulations.gov,

Docket No. FWS-R2-ES-2013-0071). General rationale is provided in the introductory paragraph to this section, “Current Distribution and Population Status.”

Table 1—Current Population Status of the Northern Mexican Gartersnake in the United States

[References for This Information Are Provided in Appendix A]

Location

Last record

Suitable physical

habitat

present

Native prey species present

Harmful

nonnative

species

present

Population status

Gila River (NM, AZ)

2013

Yes

Yes

Yes

Likely not viable.

Spring Canyon (NM)

1937

Yes

Possible

Likely

Likely extirpated.

Mule Creek (NM)

1983

Yes

Yes

Yes

Likely not viable.

Mimbres River (NM)

Likely early 1900s

Yes

Yes

Yes

Likely extirpated.

Lower Colorado River (AZ)

1904

Yes

Yes

Yes

Likely extirpated.

Bill Williams River (AZ)

2012

Yes

Yes

Yes

Likely viable.

Agua Fria River (AZ)

1986

Yes

Yes

Yes

Likely not viable.

Little Ash Creek (AZ)

1992

Yes

Yes

Yes

Likely not viable.

Lower Salt River (AZ)

1964

Yes

Yes

Yes

Likely extirpated.

Black River (AZ)

1982

Yes

Yes

Yes

Likely not viable.

Big Bonito Creek (AZ)

1986

Yes

Yes

Yes

Likely not viable.

Tonto Creek (AZ)

2005

Yes

Yes

Yes

Likely viable.

Upper Verde River (AZ)

2012

Yes

Yes

Yes

Likely viable.

Oak Creek (AZ)

(Page Springs and Bubbling Ponds State Fish Hatcheries)

2012

Yes

Yes

Yes

Likely viable.

Spring Creek (AZ)

1986

Yes

Yes

Yes

Likely not viable.

Sycamore Creek (Yavapai/Coconino Co., AZ)

1954

Yes

Possible

Yes

Likely extirpated.

Upper Santa Cruz River/San Rafael Valley (AZ)

2013

Yes

Yes

Yes

Likely viable.

Redrock Canyon (AZ)

2008

Yes

Yes

Yes

Likely not viable.

Sonoita Creek (AZ)

2013

Yes

Possible

Yes

Likely not viable.

Scotia Canyon (AZ)

2009

Yes

Yes

No

Likely not viable.

Parker Canyon (AZ)

1986

Yes

Possible

Yes

Likely not viable.

Las Cienegas National Conservation Area and Cienega Creek Natural Preserve (AZ)

2012

Yes

Yes

Possible

Likely not viable.

Lower Santa Cruz River (AZ)

1956

Yes

Yes

Yes

Likely extirpated.

Buenos Aires National Wildlife Refuge (AZ)

2000

Yes

Yes

Yes

Likely not viable.

Bear Creek (AZ)

1987

Yes

Yes

Yes

Likely not viable.

San Pedro River (AZ)

1996

Yes

Yes

Yes

Likely not viable.

Babocomari River and Cienega (AZ)

1986

Yes

Possible

Yes

Likely not viable.

Canelo Hills-Sonoita Grasslands Area (AZ)

2012

Yes

Yes

Yes

Likely not viable.

San Bernardino National Wildlife Refuge (AZ)

1997

Yes

Yes

Yes

Likely not viable.

Notes:

“Possible” means there were no conclusive data found. “Likely extirpated” means the last record for an area pre-dated 1980, and existing threats suggest the species is likely extirpated. “Likely not viable” means there is a post-1980 record for the species, it is not reliably found with minimal to moderate survey effort, and threats exist which suggest the population may be low density or could be extirpated, but there is insufficient evidence to support extirpation. “Likely viable” means that the species is reliably found with minimal to moderate survey effort, and the population is generally considered to be somewhat resilient.

We conclude that as many as 24 of 29 known northern Mexican gartersnake localities in the United States (83 percent) are likely not viable and may exist at low population densities that could be threatened with extirpation or may already be extirpated. In most localities where the species may occur at low population densities, existing survey data are insufficient to support a conclusion of extirpation. Only five populations of northern Mexican gartersnakes in the United States are considered likely viable where the species remains reliably detected. In our November 25, 2008, 12-month finding, we evaluated the total number of stream miles in the United States that historically supported the northern Mexican gartersnake that are now permanently dewatered (except in the case of temporary flows in response to heavy precipitation), and we concluded that the subspecies has been extirpated from or occurs at low densities in as much as 90 percent of its historical range in the United States (73 FR 71788, pp. 71792-71793). As shown in Table 1, harmful nonnative species are present in all but one northern Mexican gartersnake locality in the United States.

The northern Mexican gartersnake is listed as threatened throughout its range in Mexico by the Mexican Government. However, our understanding of the northern Mexican gartersnake's specific population status throughout its range in Mexico is less precise than that known for its United States distribution because survey efforts are less and available records do not exist or are difficult to obtain for many regions. Some specific geographic distribution records for the Mexican states of Sonora, Chihuahua, and San Luis Potosí were presented in Lemos-Espinal (2013, pers. comm.). Lemos-Espinal (2013 pers. comm), a Mexican herpetologist whose work is focused on the states of Sonora, Chihuahua, and Coahuila, commented that the number and magnitude of threats are not equal across the subspecies' range in Mexico. Habitat alteration or removal, as a circumstance of human population growth in Mexico, is reported as a primary concern for populations that occur in the Sierra Madre Occidental (Lemos-Espinal 2013, pers. comm.). In other regions of Mexico, such as the states of Sonora and Chihuahua, Lemos-Espinal (2013, pers. comm.) observed the northern Mexican gartersnake to be quite common. Another gartersnake researcher from Mexico has observed the decline or disappearance of some populations in central Mexico (Manjerrez 2008).

Narrow-Headed Gartersnake

Species Description

The narrow-headed gartersnake is a small to medium-sized gartersnake with a maximum total length of 44 in (112 cm) (Painter and Hibbitts 1996, p. 147). Its eyes are set high on its unusually elongated head, which narrows to the snout, and it lacks striping on the dorsum (top) and sides, which distinguishes its appearance from other gartersnake species with which it could co-occur (Rosen and Schwalbe 1988, p. 7). The base color is usually tan or grey-brown (but may darken) with conspicuous brown, black, or reddish spots that become indistinct towards the tail (Rosen and Schwalbe 1988, p. 7; Boundy 1994, p. 126). The scales are keeled. Degenhardt

et al.

(1996, p. 327), Rossman

et al.

(1996, pp. 242-244), and Ernst and Ernst (2003, p. 416) further describe the species.

Taxonomy

We recognize the narrow-headed gartersnake,

Thamnophis rufipunctatus,

as a monotypic species (no currently recognized subspecies exist). The narrow-headed gartersnake is a member of the family Colubridae and subfamily Natricinae (harmless live-bearing snakes) (Lawson

et al.

2005, p. 596). The taxonomy of the genus

Thamnophis

has a complex history partly because many of the species are similar in appearance and scutelation (arrangement of scales) and because many of the early museum specimens were in such poor and faded condition that it was difficult to study them (Conant 2003, p. 6). There are approximately 30 species described in the gartersnake genus

Thamnophis

(Rossman

et al.

1996, pp. xvii-xviii). Two large overlapping clades (related taxonomic groups) of gartersnakes have been identified called the “Mexican” and “widespread” clades, supported by allozyme and mitochondrial DNA genetic analyses (de Queiroz

et al.

2002, p. 321). The narrow-headed gartersnake (

Thamnophis rufipunctatus

) is a member of the “Mexican” clade and is most closely related taxonomically to the southern Durango spotted gartersnake (

Thamnophis nigronuchalis

) (de Queiroz and Lawson 1994, p. 217; de Queiroz

et al.

2002; p. 321).

Due to the narrow-headed gartersnake's morphology and feeding habits, there has been considerable deliberation among taxonomists about the correct association of this species within seven various genera over time (Rosen and Schwalbe 1988, pp. 5-6); chiefly, between the genera

Thamnophis

(the “gartersnakes”) and

Nerodia

(the “watersnakes”) (Pierce 2007, p. 5). Chaisson and Lowe (1989, pp. 110-118) argued that the pattern of ultrastructural (as revealed by an electron microscope) pores in the scales of narrow-headed gartersnakes provided evidence that the species is more appropriately placed within the genus

Nerodia.

However, De Queiroz and Lawson (1994, p. 217) rejected this premise using mitochondrial DNA (mtDNA) genetic analyses to refute the inclusion of the narrow-headed gartersnake in the genus

Nerodia

and maintain the species within the genus

Thamnophis.

The narrow-headed gartersnake was first described as

Chilopoma rufipunctatum

by E. D. Cope (in Yarrow, 1875). Recently,

Thamnophis

rufipunctatus nigronuchalis

and

T. r. unilabialis

were recognized as subspecies under

T. rufipunctatus

and comprised what was considered the

T. rufipunctatus

complex (Rossman

et al.

1996, p. 245). However, Rossman

et al.

(1996, pp. 244-246) elevated

T. r. nigronuchalis

to full species designation and argued that recognition of

T. r. unilabialis

be discontinued due to the diagnostic differences being too difficult to discern. Wood

et al.

(2011, p. 14) used genetic analysis of the

T. rufipunctatus

complex to propose the elevation of these three formerly recognized subspecies as three distinct species, as a result of a combination of interglacial warming, ecological and life-history constraints, and genetic drift, which promoted differentiation of these three species throughout the warming and cooling periods of the Pleistocene epoch (Wood

et al.

2011, p. 15). We use these most recent and complete data in acknowledging these three entities as unique species:

T. rufipunctatus

(along the Mogollon Rim of Arizona and New Mexico, the narrow-headed gartersnake, which is the subject of this rule),

T. unilabialis

(Chihuahua, eastern Sonora, and northern Durango, Mexico), and

T. nigronuchalis

(southern Durango, Mexico).

Several common names have been used for this species including the red-spotted gartersnake, the brown-spotted gartersnake, and the currently used, narrow-headed gartersnake (Rosen and Schwalbe 1988, p. 5). Further discussion of the taxonomic history of the narrow-headed gartersnake is available in Crother (2012, p. 71), Degenhardt

et al.

(1996, p. 326), Rossman

et al.

(1996, p. 244), De Queiroz and Lawson (1994, pp. 213-229), Rosen and Schwalbe (1988, pp. 5-7), and De Queiroz

et al.

(2002, p. 321).

Habitat and Natural History

The narrow-headed gartersnake, distributed across the Mogollon Rim of Arizona and New Mexico, is widely considered to be one of the most aquatic of the gartersnakes (Drummond and Marcias Garcia 1983, pp. 24, 27; Rossman

et al.

1996, p. 246). This species is strongly associated with clear, rocky streams, using predominantly pool and riffle habitat that includes cobbles and boulders (Rosen and Schwalbe 1988, pp. 33-34; Degenhardt

et al.

1996, p. 327; Rossman

et al.

1996, p. 246; Nowak and Santana-Bendix 2002, pp. 26-37; Ernst and Ernst 2003, p. 417). Rossman

et al.

(1996, p. 246) also note the species has been observed using lake shoreline habitat in New Mexico. Narrow-headed gartersnakes occur at elevations from approximately 2,300 to 8,000 ft (701 to 2,430 m), inhabiting Petran Montane Conifer Forest, Great Basin Conifer Woodland, Interior Chaparral, and the Arizona Upland subdivision of Sonoran Desertscrub communities (Rosen and Schwalbe 1988, p. 33; Brennan and Holycross 2006, p. 122).

An extensive evaluation of habitat use of narrow-headed gartersnakes along Oak Creek in Arizona is provided in Nowak and Santana-Bendix (2002, pp. 26-37). In the upper reaches of Oak Creek, occupied habitat is found in a steep-walled, confined canyon with shallow, braided stream segments, minimal silt, and good canopy coverage, vegetated islands and significant amounts of aquatic vegetation (Nowak and Santana-Bendix 2002, pp. 29-30). In the middle reaches of Oak Creek, occupied habitat is found in a wider canyon with less stream braiding, deeper pools, more silt, and high canopy coverage and stream-side vegetation, but less aquatic vegetation (Nowak and Santana-Bendix 2002, pp. 30-31). In the lower reaches of Oak Creek, historically occupied habitat occurred outside of the canyon proper, with predominant pool-run sequences, rare channel braiding, much silt, significantly less canopy coverage or streamside vegetation and few areas with aquatic vegetation (Nowak and Santana-Bendix 2002, p. 31).

Nowak and Santana-Bendix (2002, pp. 29-31) found the most narrow-headed gartersnakes in the upper reaches of Oak Creek, followed by the middle reaches; no narrow-headed gartersnakes were found in the lower reaches. Nowak and Santana-Bendix (2002, p. 33) found that, in general, narrow-headed gartersnakes in Oak Creek were more likely to be found within reaches without crayfish and without silt. Population densities of warm-water predatory fish increase on a gradient from the upper to the lower reaches of Oak Creek, while the inverse is true for native fish populations, and their presence confounds the analysis of physical habitat preference of narrow-headed gartersnakes. Rosen and Schwalbe (1988, p. 35) found that the relative abundance of narrow-headed gartersnakes may be highest at the conjunction of cascading riffles with pools, where waters were deeper than 20 in (0.5 m) in the riffle and deeper than 40 in (1 m) in the immediately adjoining area of the pool. However, more than twice the number of snakes was found in pools rather than riffles, but this observation may not translate for smaller streams. Despite their highly aquatic behavior, narrow-headed gartersnakes in Oak Creek have been shown to use upland habitat within 328 feet (100 m) during early fall and spring months, strongly associate with boulders in the floodplain during summer months, and use upland habitat up to 656 feet (200 m) out of the floodplain as hibernation sites (Nowak 2006, pp. 20, 26).

Bank-line vegetation is an important component to suitable habitat for this species (Nowak and Santana-Bendix 2002, pp. 26-37). Narrow-headed gartersnakes will usually bask in situations where a quick escape can be made, whether that is into the water or under substrate such as rocks (Fleharty 1967, p. 16). Common plant species associations include Arizona alder (

Alnus oblongifolia

) (highest correlation with occurrence of the narrow-headed gartersnake), velvet ash (

Fraxinus pennsylvanica

), willows (

Salix

ssp.), canyon grape (

Vitis arizonica

), blackberry (

Rubus

ssp.), Arizona sycamore (

Platanus wrightii

), Arizona black walnut (

Juglans major

), Freemont cottonwood (

Populus fremontii

), Gambel oak (

Quercus gambelii

), and ponderosa pine (

Pinus ponderosa

) (Rosen and Schwalbe 1988, pp. 34-35). Rosen and Schwalbe (1988, p. 35) noted that the composition of bank-side plant species and canopy structure may be less important to the species' needs than was the size class of the plant species present; narrow-headed gartersnakes use shrub- and sapling-sized plants for thermoregulating (basking) at the waters' edge (Degenhardt

et al.

1996, p. 327), as well as islands within the stream channel that are created by sedge (

Carex

spp.) tussocks (Nowak and Santana-Bendix 2002, p. 34).

Narrow-headed gartersnakes may opportunistically forage within dammed reservoirs formed by streams that are occupied habitat, such as at Wall Lake, New Mexico, (located at the confluence of Taylor Creek, Hoyt Creek, and the East Fork Gila River) (Fleharty 1967, p. 207) and most recently at Snow Lake in 2012 (located near the confluence of Snow Creek and the Middle Fork Gila River) (Hellekson 2012b, pers. comm.) in New Mexico, but records from impoundments are rare. The species evolved in the absence of such habitat, and impoundments are generally managed as sport fisheries (Wall Lake and Snow Lake are) and often maintain populations of harmful nonnative species that are incompatible with narrow-headed gartersnakes.

The narrow-headed gartersnake is surface-active generally between March and November (Nowak 2006, p. 16). Little information on suitable temperatures for surface activity of the narrow-headed gartersnake exists;

however, it is presumed to be rather cold-tolerant based on its natural history and foraging behavior that often involves clear, cold streams at higher elevations. Along Oak Creek in Arizona, Nowak (2006, Appendix 1) found the species to be active in air temperatures ranging from 52 to 89 °F (11 to 32 °C) and water temperatures ranging from 54 to 72 °F (12 to 22 °C). Jennings and Christman (2011, pp. 12-14) found body temperatures of narrow-headed gartersnakes along the Tularosa River averaged approximately 68 °F (20 °C) during the mid-morning hours and 81 °F (27 °C) in the late afternoon during the period from late July and August. Variables that affect their body temperature include the temperature of the microhabitat used and water temperature (most predictive), but slope aspect and the surface area of cover used also influenced body temperatures (Jennings and Christman 2011, p. 13). Narrow-headed gartersnakes have a lower preferred temperature for activity as compared to other species of gartersnakes (Fleharty 1967, p. 228), which may facilitate their highly aquatic nature in cold streams.

Narrow-headed gartersnakes specialize on fish as their primary prey item (Rosen and Schwalbe 1988, p. 38; Degenhardt

et al.

1996, p. 328; Rossman

et al.

1996, p. 247; Nowak and Santana-Bendix 2002, pp. 24-25; Nowak 2006, p. 22). They are believed to be mainly visual hunters (Hibbitts and Fitzgerald 2005, p. 364) heavily dependent on visual cues when foraging based on comparative analyses among other species of gartersnakes (de Queiroz 2003, p. 381). Unlike many other species of gartersnakes that are active predators (actively crawl about in search of prey), narrow-headed gartersnakes are considered to be ambush predators (sit-and-wait method) (Brennan and Holycross 2006, p. 122; Pierce

et al.

2007, p. 8). The specific gravity (ratio of the mass of a solid object to the mass of the same volume of water) of the narrow-headed gartersnake was found to be nearly 1, which means that the snake can maintain its desired position in the water column with ease, an adaptation to facilitate foraging on the bottom of streams (Fleharty 1967, pp. 218-219).

Native fish species most often associated as prey items for the narrow-headed gartersnake include Sonora sucker (

Catostomus insignis

), desert sucker (

C. clarki

), speckled dace (

Rhinichthys osculus

), roundtail chub (

Gila robusta

), Gila chub (

Gila intermedia

), and headwater chub (

Gila nigra

) (Rosen and Schwalbe 1988, p. 39; Degenhardt

et al.

1996, p. 328). Nonnative predatory fish species in their fingerling size classes are also used as prey by narrow-headed gartersnakes, including brown trout (Rosen and Schwalbe 1988, p. 39; Nowak and Santana-Bendix 2002, p. 24; Nowak 2006, pp. 22-23), green sunfish (Fleharty 1967, p. 223), and smallmouth bass (

Micropterus dolomieu

) (M. Lopez, 2010, pers. comm.). Reports suggest that brown trout are consumed more frequently than smallmouth bass. Trout species are commonly stocked in, or near, occupied narrow-headed gartersnake habitat. Fleharty (1967, p. 223) reported narrow-headed gartersnakes eating green sunfish. But nonnative fish with spiny dorsal fins are not generally considered suitable prey items due to the risk of injury to the gartersnake during ingestion and because of where they tend to occur in the water column (see discussion in the subsection “Fish” under the subheading “Decline of the Gartersnake Prey Base” and Nowak and Santana-Bendix (2002, p. 24)).

Although the narrow-headed gartersnake has been reported to also prey upon amphibians such as frogs, tadpoles, and salamanders (Stebbins 1985, p. 199; Deganhardt

et al.

1996, p. 328; Ernst and Ernst 2003, p. 418), we believe these are not important items in their diet. Despite several studies focusing on the ecology of narrow-headed gartersnakes in recent times, there are no other records of narrow-headed gartersnakes, under current taxonomic recognition, feeding on prey items other than fish. Fitzgerald (1986, p. 6) referenced the Stebbins (1985) account as the only substantiated account of the species eating something other than fish as prey, apparently as the result of finding a small salamander larvae in the stomach of an individual in Durango, Mexico. Formerly recognized as a subspecies of

Thamnophis rufipunctatus,

that individual is now recognized as

T. unilabialis

(Wood

et al.

2011, p. 3). We found one account of narrow-headed gartersnakes consuming red-spotted toads in captivity (Woodin 1950, p. 40). Amphibian larvae (i.e.

Hyla

sp.,

Anaxyrus

sp.,

Ambystoma

sp.) are generally available to narrow-headed gartersnakes as prey, yet observations of narrow-headed gartersnakes using them are rare. Therefore, we do not consider amphibians as ecologically important prey for this species.

Native predators of the narrow-headed gartersnake include birds of prey, such as black-hawks (Etzel

et al.

2014, p. 56), other snakes such as regal ring-necked snakes (Brennan

et al.

2009, p. 123), wading birds, mergansers, belted kingfishers, raccoons (Rosen and Schwalbe 1988, p. 39), and possibly other generalist mammalian predators. Historically, large, highly predatory native fish species, such as Colorado pikeminnow, may have preyed upon narrow-headed gartersnakes where the species co-occurred. Native chubs (

Gila

spp.) may also prey on neonatal gartersnakes.

Sexual maturity in narrow-headed gartersnakes occurs at 2.5 years of age in males and at 2 years of age in females (Deganhardt

et al.

1996, p. 328). Narrow-headed gartersnakes are viviparous. Narrow-headed gartersnakes breed annually, and females give birth to 4 to 17 offspring from late July into early August, perhaps earlier at lower elevations (Rosen and Schwalbe 1988, pp. 35-37). Narrow-headed gartersnakes may live as long as 10 years in the wild (Rosen and Schwalbe 1988, p. 38).

Historical Distribution

The historical distribution of the narrow-headed gartersnake ranged across the Mogollon Rim and along associated perennial stream drainages from central and eastern Arizona, southeast to southwestern New Mexico at elevations ranging from 2,300 to 8,000 ft (700 to 2,430 m) (Rosen and Schwalbe 1988, p. 34; Rossman

et al.

1996, p. 242; Holycross

et al.

2006, p. 3). The species was historically distributed in headwater streams of the Gila River subbasin that drain the Mogollon Rim and White Mountains in Arizona, and the Gila Wilderness in New Mexico. Major subbasins in its historical distribution included the Salt and Verde River subbasins in Arizona, and the San Francisco and Gila River subbasins in New Mexico (Holycross

et al.

2006, p. 3). Holycross

et al.

(2006, p. 3) suspect the species was likely not historically present in the lowest reaches of the Salt, Verde, and Gila Rivers, even where perennial flow persists. Numerous records for the narrow-headed gartersnake (through 1996) in Arizona are maintained in the AGFD's Heritage Database (1996b). The narrow-headed gartersnake as currently recognized does not occur in Mexico.

Current Distribution and Population Status

Population status information suggests that the narrow-headed gartersnake has experienced significant declines in population density and distribution along streams and rivers where it was formerly well-documented and reliably detected. Many areas where the species may occur likely rely on emigration of individuals from occupied habitat into those areas to maintain the species, provided there are no potential

barriers to movement, such as extensive stretches of dewatered habitat, or high densities of harmful nonnative species. Holycross

et al.

(2006, entire) represents the most recent, comprehensive survey effort for narrow-headed gartersnakes in Arizona. Narrow-headed gartersnakes were detected in 5 of 16 historical localities in Arizona and New Mexico surveyed by Holycross

et al.

(2006) in 2004 and 2005. Population densities have noticeably declined in many populations, as compared to previous survey efforts (Holycross

et al.

2006, p. 66). Holycross

et al.

(2006, pp. 66-67) compared narrow-headed gartersnake detections based on results from their effort and that of previous efforts in the same locations and found that significantly more effort is required to detect this species in areas where it was formerly robust, such as along Eagle Creek (AZ), the East Verde River (AZ), the San Francisco River (NM), the Black River (AZ), and the Blue River (AZ).

Where narrow-headed gartersnakes are locally abundant, they can usually be detected reliably and with significantly less effort than populations characterized as having low densities. Narrow-headed gartersnakes are well-camouflaged, secretive, and very difficult to detect in structurally complex, dense habitat where they could occur at very low population densities, which characterizes most occupied sites. We considered factors such as the date of the last known records for narrow-headed gartersnakes in an area, as well as records of one or more native prey species, in making a conclusion on species occupancy. We used all records that were dated 1980 or later because the 1980s marked the first systematic survey efforts for narrow-headed gartersnake species across their range (see Rosen and Schwalbe (1988, entire) and Fitzgerald (1986, entire)), and the last, previous records were often dated several decades prior and may not accurately represent the likelihood for current occupation. Several areas where narrow-headed gartersnakes were known to occur have received no, or very little, survey effort in the past several decades. Variability in survey design and effort makes it difficult to compare population sizes or trends among sites and between sampling periods. Thus, for each of the sites discussed in Appendix A (available at

http://www.regulations.gov,

Docket No. FWS-R2-ES-2013-0071), we have attempted to translate and quantify search and capture efforts into comparable units (

i.e.,

person-search hours and trap-hours) and have cautiously interpreted those results. Where survey data are sparse, the presence of suitable prey species in an area may provide evidence that narrow-headed gartersnakes may still persist at low densities. Therefore, a record of a native prey species was considered in our determination of occupancy of this species.

As of 2011, the only remaining narrow-headed gartersnake populations where the species could reliably be found were located at: (1) Whitewater Creek (NM), (2) Tularosa River (NM), (3) Diamond Creek (NM), (4) Middle Fork Gila River (NM), and (5) Oak Creek Canyon (AZ). However, populations found in Whitewater Creek and the Middle Fork Gila River were likely significantly affected by the large Whitewater-Baldy Complex Fire, which occurred in June 2012. In addition, salvage efforts were initiated for these two populations, which included the removal of 25 individuals from Whitewater Creek and 14 individuals from the Middle Fork Gila River before the onset of summer rains in 2012. These 39 individuals were transported to the Albuquerque BioPark where 22 remain in captivity. The other 17 of the salvaged individuals were translocated to Saliz Creek, where the resident native prey base appears adequate, and beyond the effects from the Whitewater-Baldy Complex Fire. The status of those populations in Whitewater Creek and the Middle Fork Gila River has likely deteriorated as a result of subsequent declines in resident fish communities due to heavy ash and sediment flows, resulting fish kills, and the removal of snakes, but subsequent survey data have not been collected. If the Whitewater Creek and Middle Fork Gila River populations did decline as a result of these factors, only three remaining populations of this species remain viable today across their entire distribution. While historical records confirm the narrow-headed gartersnake was found on tribal lands, its current status on tribal land is poorly known due to limited survey access.

In Table 2 below, we summarize the population status of the narrow-headed gartersnake at all known localities throughout its distribution, as supported by museum records or reliable observations. For a detailed discussion that explains the rationale for site-by-site conclusions on occupancy and status, please see Appendix A (available at

http://www.regulations.gov,

Docket No. FWS-R2-ES-2013-0071). General rationale is provided in the introductory paragraph to this section, “Current Distribution and Population Status.”

Table 2—Current Population Status of the Narrow-Headed Gartersnake

[References for this information are provided in appendix A]

Location

Last record

Suitable physical habitat present

Native prey species present

Harmful nonnative species present

Population status

West Fork Gila River (NM)

2011

Yes

Yes

Yes

Likely not viable.

Middle Fork Gila River (NM)

2012

Yes

Yes

Yes

Likely not viable.

East Fork Gila River (NM)

2006

Yes

Yes

Yes

Likely not viable.

Gila River (AZ, NM)

2009

Yes

Yes

Yes

Likely not viable.

Snow Creek/Snow Lake (NM)

2012

Yes

No

Yes

Likely not viable.

Gilita Creek (NM)

2009

Yes

Yes

No

Likely not viable.

Iron Creek (NM)

2009

Yes

Yes

No

Likely not viable.

Little Creek (NM)

2010

Yes

Possible

Yes

Likely not viable.

Turkey Creek (NM)

1985

Yes

Yes

Possible

Likely not viable.

Beaver Creek (NM)

1949

Yes

Possible

Yes

Likely extirpated.

Black Canyon (NM)

2010

Yes

Yes

Yes

Likely not viable.

Taylor Creek (NM)

1960

Yes

No

Yes

Likely extirpated.

Diamond Creek (NM)

2011

Yes

Yes

Yes

Likely viable.

Tularosa River (NM)

2012

Yes

Yes

Yes

Likely viable.

Whitewater Creek (NM)

2012

Yes

Yes

Yes

Likely not viable.

San Francisco River (NM)

2011

Yes

Yes

Yes

Likely not viable.

South Fork Negrito Creek (NM)

2011

Yes

Possible

Yes

Likely not viable.

Blue River (AZ)

2007

Yes

Yes

Yes

Likely not viable.

Dry Blue Creek (AZ, NM)

2010

Yes

Possible

Yes

Likely not viable.

Campbell Blue Creek (AZ, NM)

2010

Yes

Possible

Yes

Likely not viable.

Saliz Creek (NM)

2013

Yes

Possible

Yes

Likely not viable.

Eagle Creek (AZ)

2013

Yes

Possible

Yes

Likely not viable.

Black River (AZ)

2013

Yes

Yes

Yes

Likely not viable.

East Fork Black River (AZ)

2004

Yes

Possible

Yes

Likely not viable.

Fish Creek (Tributary to East Fork Black River; AZ)

2004

Yes

Yes

Possible

Likely viable.

White River (AZ)

1986

Yes

Yes

Possible

Likely not viable.

Diamond Creek (AZ)

1986

Yes

Possible

Possible

Likely not viable.

Tonto Creek (tributary to Big Bonita Creek, AZ)

1915

Yes

Possible

Possible

Likely extirpated.

Canyon Creek (AZ)

1991

Yes

Yes

No

Likely not viable.

Upper Salt River (AZ)

1985

Yes

Yes

Yes

Likely not viable.

Cibeque Creek (AZ)

1991

Yes

Yes

Possible

Likely not viable.

Carrizo Creek (AZ)

1997

Yes

Yes

Possible

Likely not viable.

Big Bonito Creek (AZ)

1957

Yes

Yes

Yes

Likely extirpated.

Haigler Creek (AZ)

2008

Yes

Yes

Yes

Likely not viable.

Houston Creek (AZ)

2005

Yes

Yes

Yes

Likely not viable.

Tonto Creek (tributary to Salt River, AZ)

2005

Yes

Yes

Yes

Likely not viable.

Deer Creek (AZ)

1995

No

No

No

Likely extirpated.

Upper Verde River (AZ)

2012

Yes

Yes

Yes

Likely not viable.

Oak Creek (AZ)

2012

Yes

Yes

Yes

Likely viable.

West Fork Oak Creek (AZ)

2012

Yes

Yes

Yes

Likely viable.

East Verde River (AZ)

1992

Yes

Yes

Yes

Likely not viable.

Notes:

“Possible” means there were no conclusive data found. “Likely extirpated” means the last record for an area pre-dated 1980, and existing threats suggest the species is likely extirpated. “Likely not viable” means there is a post-1980 record for the species, it is not reliably found with minimal to moderate survey effort, and threats exist which suggest the population may be low density or could be extirpated, but there is insufficient evidence to support extirpation. “Likely viable” means that the species is reliably found with minimal to moderate survey effort, and the population is generally considered to be somewhat resilient.

Table 2 lists the 41 known localities for narrow-headed gartersnakes throughout their range. We have concluded that, in as many as 31 of 41 known localities (76 percent), the narrow-headed gartersnake population is likely not currently viable and may exist at low population densities that could be threatened with extirpation or may already be extirpated, but survey data are lacking in areas where access is restricted. In most localities where the species may occur at low population densities, existing survey data are insufficient to conclude extirpation. As of 2014, narrow-headed gartersnake populations are considered currently likely viable in five localities (12 percent). The remaining five populations (12 percent) are considered currently likely extirpated. As displayed in Table 2, harmful nonnative species are a concern for all but four narrow-headed gartersnake populations. The status of these populations is expected to continue to decline.

Summary of Biological Status and Threats

Section 4 of the Act (16 U.S.C. 1533), and its implementing regulations at 50 CFR part 424, set forth the procedures for adding species to the Federal Lists of Endangered and Threatened Wildlife and Plants. Under section 4(a)(1) of the Act, we may list a species 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; and (E) other natural or manmade factors affecting its continued existence. Listing actions may be warranted based on any of the above threat factors, singly or in combination.

In the following threats analysis, we treat both gartersnake species in a combined discussion because of partially overlapping ranges, similar natural histories, similar responses to threats, and the fact that many threats are shared in common throughout their ranges.

Weakened Status of Native Aquatic Communities (Northern Mexican and Narrow-Headed Gartersnakes) (Factors A, C, and E)

The presence of harmful nonnative species constitutes the most significant threat to the two gartersnake species. Harmful nonnative species directly prey upon both species of gartersnake and compete with them for prey. Harmful nonnative species also compete with gartersnake prey species as well as modify habitat for both the gartersnakes and their prey, to the detriment of both gartersnakes. Landscape-level effects from the continued expansion of harmful nonnative species have changed the spatial orientation of these gartersnakes' distributions, creating greater isolation between populations. We expect the viability of extant gartersnake populations to continue to degrade into the foreseeable future as a result of ecological interactions with harmful nonnative species. Riparian and aquatic communities in both the southwestern United States and Mexico have been significantly impacted by a shift in species' composition, from one of primarily native fauna, to one dominated by an expanding assemblage of harmful nonnative animal species. Harmful nonnative species have been introduced or have spread into new areas through a variety of mechanisms, including intentional and accidental releases, sport stocking, aquaculture, aquarium releases, bait-bucket releases, or natural dispersal (Welcomme 1984, entire). The ecological ramifications of

the adversarial relationships within southwestern aquatic communities have been discussed and described in a broad body of literature, extending from 1985 to the present (Meffe 1985, pp. 179-185; Propst

et al.

1986, pp. 14-31, 82; 1988, p. 64; 2009, pp. 5-17; Rosen and Schwalbe 1988, pp. 28, 32; 1997, p. 1; Clarkson and Rorabaugh 1989, pp. 531, 535; Douglas

et al.

1994, pp. 9-19; Rosen

et al.

1995, pp. 257-258; 2001, p. 2; Degenhardt

et al.

1996, p. 319; Fernandez and Rosen 1996, pp. 8, 23-27, 71, 96; Richter

et al.

1997, pp. 1089, 1092; Inman

et al.

1998, p. 17; Rinne

et al.

1998, pp. 4-6; Nowak and Santana-Bendix 2002, Table 3; Propst 2002, pp. 21-25; DFT 2003, pp. 1-3, 5-6, 19; 2004, pp. 1-2, 4-5, 10, Table 1; Bonar

et al.

2004, pp. 13, 16-21; Rinne 2004, pp. 1-2; Clarkson

et al.

2005, p. 20; Fagan

et al.

2005, pp. 34, 34-41; Knapp 2005, pp. 273-275; Olden and Poff 2005, pp. 82-87; Turner 2007, p. 41; Holycross

et al.

2006, pp. 13-15; Brennan 2007, pp. 5, 7; Caldwell 2008a, 2008b; d'Orgeix 2008; Luja and Rodríguez-Estrella 2008, pp. 17-22; Propst

et al.

2008, pp. 1242-1243; Rorabaugh 2008a, p. 25; Brennan and Rosen 2009, pp. 8-9; Minckley and Marsh 2009, pp. 50-51; Pilger

et al.

2010, pp. 311-312; Stefferud

et al.

2009, pp. 206-207; 2011, pp. 11-12; Young and Boyarski 2013, pp. 159-160).

Decline of the Gartersnake Prey Base (Northern Mexican and Narrow-Headed Gartersnakes) (Factors A and E)

The prey base of these gartersnakes includes native amphibians and fish populations. Declines in their prey base have led to subsequent declines in the distribution and density of gartersnake populations. In most areas across their ranges, prey base declines are largely attributed to the introduction and expansion of harmful nonnative species.

Northern Mexican and narrow-headed gartersnakes may be particularly vulnerable to the loss of native prey species (Rosen and Schwalbe 1988, pp. 20, 44-45). Rosen

et al.

(2001, pp. 10, 13, 19) theorized that the northern Mexican gartersnake: (1) Is unlikely to increase foraging efforts at the risk of increased predation; and (2) needs adequate food on a regular basis to maintain its weight and health. If forced to forage more often for smaller prey items, a reduction in growth and reproductive rates can result (Rosen

et al.

2001, pp. 10, 13). Rosen

et al.

(2001, p. 22) hypothesized that the presence and expansion of nonnative predators (mainly bullfrogs, crayfish, and green sunfish (

Lepomis cyanellus

)) are the primary causes of decline in northern Mexican gartersnakes and in their prey in southeastern Arizona. In another example, Drummond and Macías Garcia (1989, pp. 25, 30) found that Mexican gartersnakes fed primarily on frogs, and when frogs became unavailable, the species simply ceased major foraging activities. This led the authors to conclude that frog abundance is probably the most important correlate, and main determinant, of foraging behavior in northern Mexican gartersnakes.

With respect to narrow-headed gartersnakes, the relationship between harmful nonnative species, a declining prey base, and gartersnake populations is clearly depicted in one population along Oak Creek. Nowak and Santana-Bendix (2002, Table 3) found a strong correlation in the distribution of fish communities and narrow-headed gartersnake communities in the vicinity of Midgely Bridge. Downstream of that point, nonnative, predatory fish species increase in abundance, and narrow-headed gartersnakes notably decrease in abundance. Upstream of that point, native fish and nonnative, soft-rayed fish species increase in abundance as do narrow-headed gartersnakes (Nowak and Santana-Bendix 2002, p. 23).

Fish (Northern Mexican and Narrow-headed Gartersnakes)

—Fish are an important prey item for the northern Mexican gartersnake and are the only prey for the narrow-headed gartersnake. Native fish communities throughout the range of these gartersnake have been on the decline, both in terms of species composition and biomass, for many decades, and largely as a result of predation and competition from and with nonnative, predatory fish species. Stocked for sport, forage, or biological control, nonnative fishes have been shown to become invasive where released and do not require the natural flow regimes that native species do (Kolar

et al.

2003, p. 9), which has contributed to their expansion in the Gila River basin and elsewhere. Northern Mexican and narrow-headed gartersnakes can successfully use nonnative, soft-rayed fish species as prey, such as mosquitofish, red shiner, and introduced trout species, such as rainbow trout (

Oncorynchus mykiss

), brook trout (

Salvelinus fontinalis

), or brown trout (Nowak and Santana-Bendix 2002, pp. 24-25; Holycross

et al.

2006, p. 23). However, predatory fish are not generally considered prey species for northern Mexican or narrow-headed gartersnakes and, in addition, are known to prey on neonatal and juvenile gartersnakes (Young and Boyarski 2013, pp. 158-159). Nowak and Santana-Bendix (2002, p. 24) propose two hypotheses regarding the reluctance of narrow-headed gartersnakes to prey on nonnative, predatory fish: (1) The laterally compressed shape and presence of sharp, spiny dorsal spines of many nonnative, predatory fish present a choking hazard to gartersnakes that can be fatal; and (2) nonnative, predatory fish (with the exception of catfish) tend to occupy the middle and upper zones in the water column, while narrow-headed gartersnakes typically hunt along the bottom (where native suckers and minnows often occur). As a result, nonnative, predatory fish may be less ecologically available as prey.

Brown trout are highly predatory in all size classes in a wide range of water temperatures, and they adversely affect native fish communities wherever they are introduced (Taylor

et al.

1984, pp. 343-344). Predation on gartersnakes by adult brown trout may be a particular problem for narrow-headed gartersnakes due to their overlapping distributions and habitat preferences, both in terms of direct predation on neonatal gartersnakes and through competitive pressures for gartersnakes by preying on their food source. Specifically, the younger age classes of brown trout present competition problems for the narrow-headed gartersnake by eating small fish. As brown trout mature into the medium to larger size classes, they may prey upon neonatal narrow-headed gartersnakes. These issues are confounded by the fact that young brown trout are also eaten by narrow-headed gartersnakes and may represent an important component of their prey base, depending on fish species composition and age classes represented within the resident fish community. However, whatever benefits fingerling brown trout present for narrow-headed gartersnakes are likely off-set by effects of brown trout predation on important native fish species, and possible effects to recruitment of narrow-headed gartersnakes through predation.

Harmful nonnative species invasions can indirectly affect the health, maintenance, and reproduction of northern Mexican and narrow-headed gartersnakes by altering their foraging strategy and compromising foraging success. Rosen

et al.

(2001, p. 19), in addressing the northern Mexican gartersnake, proposed that an increase in energy expended in foraging, coupled by the reduced number of small to medium-sized prey fish available, results in deficiencies in nutrition, affecting growth and reproduction. This occurs because energy is allocated to maintenance and the increased energy costs of intense foraging activity, rather than to growth and reproduction. In

contrast, a northern Mexican gartersnake diet that includes both fish and amphibians, such as leopard frogs, reduces the necessity to forage at a higher frequency, allowing metabolic energy gained from larger prey items to be allocated instead to growth and reproductive development. Myer and Kowell (1973, p. 225) experimented with food deprivation in common gartersnakes, and found significant reductions in lengths and weights of juvenile snakes that were deprived of regular feedings versus the control group that were fed regularly at natural frequencies. Reduced foraging success of both northern Mexican and narrow-headed gartersnakes means that individuals are likely to become vulnerable to effects from starvation, which may increase fatality rates of juveniles and, consequently, affect recruitment.

Northern Mexican gartersnakes have a more varied diet than narrow-headed gartersnakes. We are not aware of any studies that have addressed the direct relationship between prey base diversity and northern Mexican gartersnake recruitment and survivorship. However, Krause and Burghardt (2001, pp. 100-123) discuss the benefits and costs that may be associated with diet variability in the common gartersnake (

Thamnophis sirtalis

), an ecologically similar species to the northern Mexican gartersnake. Foraging for mixed-prey species may impede predator learning, as compared to specialization on a certain prey species, but it may also provide long-term benefits such as the ability to capture prey throughout their lifespan (Krause and Burghardt 2001, p. 101).

A wide variety of native fish species (many of which are now listed as endangered, threatened, or candidates for listing under the Act) were historically primary prey species for northern Mexican and narrow-headed gartersnakes (Rosen and Schwalbe 1988, pp. 18, 39). Marsh and Pacey (2005, p. 60) predict that, despite the significant physical alteration of aquatic habitat in the southwestern United States, native fish species could flourish in these altered environments but for the presence of harmful nonnative fish species. Northern Mexican and, in particular, narrow-headed gartersnakes depend largely on native fish as a principal part of their prey base, although nonnative, soft-rayed predatory fish have also been documented as prey where they overlap in distribution with these gartersnakes (Nowak and Santana-Bendix 2002, pp. 24-25; Holycross

et al.

2006, p. 23; Emmons and Nowak 2013, p. 6). Nonnative, predatory fish compete with northern Mexican and narrow-headed gartersnakes for prey. In their extensive surveys, Rosen and Schwalbe (1988, p. 44) only found narrow-headed gartersnakes in abundance where native fish species predominated but did not find them abundant in the presence of robust nonnative, predatory fish populations. Minckley and Marsh (2009, pp. 50-51) found nonnative fishes to be the single-most significant factor in the decline of native fish species and also their primary obstacle to recovery. Of the 48 conterminous States in the United States, Arizona has the highest proportion of nonnative fish species (66 percent) represented by approximately 68 species (Turner and List 2007, p. 13).

Collier

et al.

(1996, p. 16) note that interactions between native and nonnative fish have significantly contributed to the decline of many native fish species from direct predation and, indirectly, from competition (which has adversely affected the prey base for northern Mexican and narrow-headed gartersnakes). Holycross

et al.

(2006, pp. 52-61) documented depressed or extirpated native fish prey bases for northern Mexican and narrow-headed gartersnakes along the Mogollon Rim in Arizona and New Mexico. Rosen

et al.

(2001, Appendix I) documented the decline of several native fish species in several locations visited in southeastern Arizona, further affecting the prey base of northern Mexican gartersnakes in that area.

Harmful nonnative fish species tend to be nest-builders and actively guard their young, which may provide them another ecological advantage over native species that are broadcast spawners and provide no parental care to their offspring (Marsh and Pacey 2005, p. 60). In fact, nesting smallmouth bass will attack gartersnakes (Winemiller and Taylor 1982, p. 270). It is, therefore, likely that recruitment and survivorship is greater in nonnative species than native species where they overlap, providing nonnative species with an ecological advantage. Table 2-1 in Kolar

et al.

(2003, p. 10) provides a map depicting the high degree of overlap in the distribution of native and nonnative fishes within the Gila River basin of Arizona and New Mexico as well as watersheds thought to be dominated by nonnative fish species.

The widespread decline of native fish species from the arid southwestern United States and Mexico has resulted largely from interactions with nonnative species and has been noted in the listing rules of 11 fishes under the Act, and their historical ranges overlap with the historical distribution of northern Mexican and narrow-headed gartersnakes. Native fish species that were likely prey species for these gartersnakes and are now listed under the Act, include the bonytail chub (

Gila elegans,

45 FR 27710, April 23, 1980), Yaqui chub (

Gila purpurea,

49 FR 34490, August 31, 1984), Yaqui topminnow (

Poeciliopsis occidentalis sonoriensis,

32 FR 4001, March 11, 1967), beautiful shiner (

Cyprinella formosa,

49 FR 34490, August 31, 1984), Gila chub (

Gila intermedia,

70 FR 66663, November 2, 2005), Colorado pikeminnow (

Ptychocheilus lucius,

32 FR 4001, March 11, 1967), spikedace (

Meda fulgida,

77 FR 10810, February 23, 2012), loach minnow (

Tiaroga cobitis,

77 FR 10810, February 23, 2012), razorback sucker (

Xyrauchen texanus,

56 FR 54957, October 23, 1991), desert pupfish (

Cyprinodon macularius,

51 FR 10842, March 31, 1986), woundfin (

Plagopterus argentissiums,

35 FR 16047, October 13, 1970), and Gila topminnow (

Poeciliopsis occidentalis,

32 FR 4001, March 11, 1967). In total within Arizona, 19 of 31 (61 percent) native fish species are listed under the Act. Arizona ranks the highest of all 50 States in the percentage of native fish species with declining trends (85.7 percent), and New Mexico ranks sixth (48.1 percent) (Stein 2002, p. 21; Warren and Burr 1994, p. 14).

The fastest expanding nonnative species are red shiner (

Cyprinella lutrensis

), fathead minnow (

Pimephales promelas

), green sunfish, largemouth bass (

Micropterus salmoides

), western mosquitofish, and channel catfish (

Ictalurus punctatus

). A nonnative species can become invasive if ecological advantages exist for broad physical tolerances, feeding habits and diet, or reproductive behavior (Taylor

et al.

1984, Table 16-1). These species are considered to be the most invasive in terms of their negative impacts on native fish communities (Olden and Poff 2005, p. 75). Many nonnative fishes, in addition to those listed immediately above, including yellow and black bullheads (

Ameiurus

sp.), flathead catfish (

Pylodictis olivaris

), and smallmouth bass, have been introduced into formerly and currently occupied northern Mexican or narrow-headed gartersnake habitat and are predators on these species (Young and Boyarski 2013, pp. 158-159) and their prey (Bestgen and Propst 1989, pp. 409-410; Marsh and Minckley 1990, p. 265; Sublette

et al.

1990, pp. 112, 243, 246, 304, 313, 318; Abarca and Weedman 1993, pp. 6-12; Stefferud and Stefferud 1994, p. 364; Weedman and Young 1997, pp. 1,

Appendices B, C; Rinne

et al.

1998, pp. 3-6; Voeltz 2002, p. 88; Bonar

et al.

2004, pp. 1-108; Fagan

et al.

2005, pp. 34, 38-39, 41; Propst

et al.

2008, pp. 1242-1243). Nonnative, predatory fish species, such as flathead catfish, may be especially dangerous to narrow-headed gartersnake populations through competition and direct predation because they are primarily piscivorous (fish-eating) (Pilger

et al.

2010, pp. 311-312), have large mouths, and have a tendency to occur along the stream bottom, where narrow-headed gartersnakes principally forage.

Rosen

et al.

(2001, Appendix I) and Holycross

et al.

(2006, pp. 15-51) conducted large-scale surveys for northern Mexican gartersnakes in southeastern and central Arizona and narrow-headed gartersnakes in central and east-central Arizona, and documented the presence of nonnative fish at many locations. Holycross

et al.

(2006, pp. 14-15) found nonnative fish species in 64 percent of the sample sites in the Agua Fria subbasin, 85 percent of the sample sites in the Verde River subbasin, 75 percent of the sample sites in the Salt River subbasin, and 56 percent of the sample sites in the Gila River subbasin. In total, nonnative fish were observed at 41 of the 57 sites surveyed (72 percent) across the Mogollon Rim (Holycross

et al.

2006, p. 14). Entirely native fish communities were presumed in only 8 of 57 sites surveyed (14 percent) (Holycross

et al.

2006, p. 14). It is well documented that nonnative fish have now infiltrated the majority of aquatic communities in the southwestern United States as depicted in Tables 1 and 2, above, as well as in Appendix A (available at

http://www.regulations.gov,

Docket No. FWS-R2-ES-2013-0071).

Several authors have identified both the presence of nonnative fish as well as their deleterious effects on native species within Arizona. Many areas have seen a shift from a predominance of native fishes to a predominance of nonnative fishes. On the upper Verde River, native species dominated the total fish community at greater than 80 percent from 1994 to 1996, before dropping to approximately 20 percent in 1997 and 19 percent in 2001. At the same time, three nonnative species increased in abundance between 1994 and 2000 (Rinne

et al.

2005, pp. 6-7). In an assessment of the Verde River, Bonar

et al.

(2004, p. 57) found that, in the Verde River mainstem, nonnative fishes were approximately 2.6 times more dense per unit volume of river than native fishes, and their populations were approximately 2.8 times that of native fishes per unit volume of river. Similar changes in the dominance of nonnative fishes have occurred on the Middle Fork Gila River, with a 65 percent decline of native fishes between 1988 and 2001 (Propst 2002, pp. 21-25). Abarca and Weedman (1993, pp. 6-12) found that the number of nonnative fish species was twice the number of native fish species in Tonto Creek in the early 1990s, with a stronger nonnative species influence in the lower reaches, where the northern Mexican gartersnake is considered to still occur (Burger 2010, p. 1, Madera-Yagla 2010, p. 6, 2011, p. 6).

Beginning in 2014, the AGFD plans to stock 4.6 million Florida-strain largemouth bass, 3.3 million bluegill, and 4.5 million black crappie annually into Roosevelt Lake in order to control the gizzard shad (

Dorosoma cepedianum

) population, which is currently the most prevalent fish species in the lake and is thought to be depressing sport fish populations in the reservoir (AGFD 2014, p. 3). Roosevelt Lake is not, and will never be, suitable habitat for the northern Mexican gartersnake because of its management as a sport fishery. However, if the goal of this effort is achieved, we expect a higher risk of predation of gartersnakes in lower Tonto Creek when a suitable hydrologic connection is made between Tonto Creek and the lake body (providing the opportunity for predatory nonnative fish to move into lower Tonto Creek). We also expect high risk of predation of individual snakes that may disperse downstream into the lake itself. Fish surveys in the Salt River above Lake Roosevelt already indicate a decline of roundtail chub and other native fishes, with an increase in flathead and channel catfish numbers (Voeltz 2002, p. 49).

In New Mexico, nonnative fish have been identified as the main cause for declines observed in native fish populations (Voeltz 2002, p. 40; Propst

et al.

2008, pp. 1242-1243). Fish experts from the U.S. Forest Service, U.S. Bureau of Reclamation, U.S. Bureau of Land Management (BLM), University of Arizona, Arizona State University, The Nature Conservancy, and others declared the native fish fauna of the Gila River basin to be critically imperiled, and they cite habitat destruction and nonnative species as the primary factors for the declines (DFT 2003, p. 1). They call for the control and removal of nonnative fish as an overriding need to prevent the decline, and possible extinction, of native fish species within the basin (DFT 2003, p. 1). In some areas, nonnative fishes may not dominate the system, but their abundance has increased. This is the case for the Cliff-Gila Valley area of the Gila River where nonnative fishes increased from 1.1 percent to 8.5 percent, while native fishes declined steadily over a 40-year period (Propst

et al.

1986, pp. 27-32). At the Redrock and Virden Valleys on the Gila River, the relative abundance in nonnative fishes in the same time period increased from 2.4 percent to 17.9 percent (Propst

et al.

1986, pp. 32-34). Four years later, the relative abundance of nonnative fishes increased to 54.7 percent at these sites (Propst

et al.

1986, pp. 32-36). The percentage of nonnative fishes increased by almost 12 percent on the Tularosa River between 1988 and 2003, while on the East Fork Gila River, nonnative fishes increased to 80.5 percent relative abundance in 2003 (Propst 2005, pp. 6-7, 23-24).

In addition to harmful nonnative species, various parasites may affect native fish species that are prey for northern Mexican and narrow-headed gartersnakes. Parasites affecting various species of native fishes within the range of these gartersnakes include Asian tapeworm (U.S. Fish and Wildlife Service (USFWS) National Wild Fish Health Survey 2010),

Ichthyophthirius multifiliis

(Ich) (Mpoame 1982, p. 46; Robinson

et al.

1998, p. 603), anchor worm (

Lernaea cyprinacea

) (Robinson

et al.

1998, pp. 599, 603-605; Hoffnagle and Cole 1999, p. 24), yellow grub (

Clinostomum marginatum

) (Amin 1969, p. 436; Mpoame and Rinne 1983, pp. 400-401; Bryan and Robinson 2000, p. 19; Maine Department of Inland Fisheries and Wildlife 2002a, p. 1), and black grub (

Neascus spp.

), also called black spot (Robinson

et al.

1998, p. 603; Bryan and Robinson 2000, p. 21; Lane and Morris 2000, pp. 2-3; Maine Department of Inland Fisheries and Wildlife 2002b, p. 1; Paroz 2011, pers. comm.). However, currently, we have no information on what effect parasite infestation in native fish might have on gartersnake populations.

Decline of Native Fish Communities in Mexico (Northern Mexican Gartersnake)

—The first tabulations of freshwater fish species at risk in Mexico occurred in 1961, when 11 species were identified as being at risk (Contreras-Balderas

et al.

2003, p. 242). As of 2003, of the 506 species of freshwater fish recorded in Mexico, 185 (37 percent) have been listed by the Mexican Federal Government as either endangered, facing extinction, under special protection, or likely extinct (Alvarez-Torres

et al.

2003, p. 323), almost a 17-fold increase in slightly over four decades; 25 species are believed to have gone extinct (Contreras-Balderas

et al.

2003, p. 241). In the lower elevations of

Mexico, within the distribution of the northern Mexican gartersnake, there are approximately 200 species of native freshwater fish documented, with 120 native species under some form of threat and an additional 15 that have gone extinct (Contreras-Balderas and Lozano 1994, pp. 383-384). The Fisheries Law in Mexico empowered the country's National Fisheries Institute to compile and publish the National Fisheries Chart in 2000, which found that Mexico's fish fauna has seriously deteriorated as a result of environmental impacts (pollution), water basin degradation (dewatering, siltation), and the introduction of nonnative species (Alvarez-Torres

et al.

2003, pp. 320, 323). The National Fisheries Chart is regarded as the first time the Mexican Government has openly revealed the status of its freshwater fisheries and described their management policies (Alvarez-Torres

et al.

2003, pp. 323-324).

Industrial, municipal, and agricultural water pollution, dewatering of aquatic habitat, and the proliferation of nonnative species are widely considered to be the greatest threats to freshwater ecosystems in Mexico (Branson

et al.

1960, p. 218; Conant 1974, pp. 471, 487-489; Miller

et al.

1989, pp. 25-26, 28-33; 2005, pp. 60-61; DeGregorio 1992, p. 60; Contreras Balderas and Lozano 1994, pp. 379-381; Lyons

et al.

1995, p. 572; 1998, pp. 10-12; Landa

et al.

1997, p. 316; Mercado-Silva

et al.

2002, p. 180; Contreras-Balderas

et al.

2003, p. 241; Domínguez-Domínguez

et al.

2007, Table 3). A shift in land use policies in Mexico to encourage free market principles in rural, small-scale agriculture has been found to promote land use practices that threaten local biodiversity (Ortega-Huerta and Kral 2007, p. 2; Randall 1996, pp. 218-220; Kiernan 2000, pp. 13-23).

These threats have been documented throughout the distribution of the northern Mexican gartersnake in Mexico and are best represented in the scientific literature in the context of fisheries studies. Contreras-Balderas

et al.

(2003, pp. 241, 243) named Chihuahua (46 species), Coahuila (35 species), Sonora (19 species), and Durango (18 species) as Mexican states that had some of the most reports of freshwater fish species at risk. These states are all within the distribution of the northern Mexican gartersnake, indicating an overlapping trend of declining prey bases and threatened ecosystems within the range of the northern Mexican gartersnake in Mexico. Contreras-Balderas

et al.

(2003, Appendix 1) found various threats to be adversely affecting the status of freshwater fish and their habitat in several states in Mexico: (1) Habitat reduction or alteration (Sonora, Chihuahua, Durango, Coahuila, San Luis Potosí, Jalisco, Guanajuato); (2) water depletion (Chihuahua, Durango, Coahuila, Sonora, Guanajuato, Jalisco, San Luis Potosí); (3) harmful nonnative species (Durango, Chihuahua, Coahuila, San Luis Potosí, Sonora, Veracruz); and (4) pollution (México, Jalisco, Chihuahua, Coahuila, Durango). Within the states of Chihuahua, Durango, Coahuila, Sonora, Jalisco, and Guanajuato water depletion is considered serious, with entire basins having been dewatered, or conditions have been characterized as “highly altered” (Contreras-Balderas

et al.

2003, Appendix 1). All of the Mexican states with the highest numbers of fish species at risk are considered arid, a condition hastened by increasing desertification (Contreras-Balderas

et al.

2003, p. 244).

Aquaculture and Nonnative Fish Proliferation in Mexico (Northern Mexican Gartersnake)

—Nonnative fish compete with and prey upon northern Mexican gartersnakes and their native prey species. The proliferation of nonnative fish species throughout Mexico happened mainly by natural dispersal, intentional stockings, and accidental breaches of artificial or constructed barriers by nonnative fish (Welcomme 1984, entire). Lentic water bodies such as lakes, reservoirs, and ponds are often used for flood control, agricultural purposes, and most commonly to support commercial fisheries. The most recent estimates indicate that Mexico has 13,936 of such water bodies, where approximately 96 percent are between 2.47-247 acres (1-100 hectares) and approximately half are artificial (Sugunan 1997, Table 8.3; Alvarez-Torres

et al.

2003, pp. 318, 322). Areas where these landscape features are most prevalent occur within the distribution of the northern Mexican gartersnake. For example, Jalisco and Zacatecas are listed as two of four states with the highest number of reservoirs, and Chihuahua is one of two states known for a high concentration of lakes (Sugunan 1997, Section 8.4.2).

Based on the data presented in Sugunan (1997, Table 8.5), a total of 422 dammed reservoirs are located within the 16 Mexican states where the northern Mexican gartersnake is thought to occur. Mercado-Silva

et al.

(2006, p. 534) found that, within the state of Guanajuato, “Practically all streams and rivers in the (Laja) basin are truncated by reservoirs or other water extraction and storage structures.” On the Laja River alone, there are two major reservoirs and a water diversion dam; 12 more reservoirs are located on its tributaries (Mercado-Silva

et al.

2006, p. 534). As a consequence of dam operations, the main channel of the Laja remains dry for extensive periods of time (Mercado-Silva

et al.

2006, p. 541). The damming and modification of the lower Colorado River in Mexico, where the northern Mexican gartersnake occurred, has facilitated the replacement of the entire native fishery with nonnative species (Miller

et al.

2005, p. 61). Each reservoir created by a dam is either managed as a nonnative commercial fishery or has become a likely source population of nonnative species, which have naturally or artificially colonized the reservoir, dispersed into connected riverine systems, and damaged native aquatic communities.

Mexico depends in large part on freshwater commercial fisheries as a source of protein for both urbanized and rural human populated areas. Commercial and subsistence fisheries rely heavily on introduced, nonnative species in the largest freshwater lakes (Soto-Galera

et al.

1999, p. 133) down to rural, small ponds (Tapia and Zambrano 2003, p. 252). At least 87 percent of the species captured or cultivated in inland fisheries of Mexico from 1989-1999 included tilapia (

Tilapia

spp.), common carp (

Cyprinus carpio

), channel catfish, trout, and black bass (

Micropterus

sp.), all of which are nonnative (Alvarez-Torres

et al.

2003, pp. 318, 322). In fact, the northern and central plateau region of Mexico (which comprises most of the distribution of the northern Mexican gartersnake's distribution in Mexico) is considered ideal for the production of harmful, predatory species such as bass and catfish (Sugunan 1997, Section 8.3). Largemouth bass are now produced and stocked in reservoirs and lakes throughout the distribution of the northern Mexican gartersnake (Sugunan 1997, Section 8.8.1).

The Secretariat for Environment, Natural Resources and Fisheries (SEMARNAP), formed in 1995, is the Mexican federal agency responsible for management of the country's environment and natural resources. SEMARNAP dictates the stocking rates of nonnative species into the country's lakes and reservoirs. For example, the permitted stocking rate for largemouth bass in Mexico is one fish per square meter in large reservoirs (Sugunan 1997, Table 8.8); therefore, a 247-acre (100-ha) reservoir could be stocked with 1,000,000 largemouth bass. The common carp, the subject of significant aquaculture investment since the 1960s in Mexico, is known for altering aquatic habitat and consuming the eggs and fry of native fish species, and is now

established in 95 percent of Mexico's freshwater systems (Tapia and Zambrano 2003, p. 252).

Basins in northern Mexico, such as the Rio Yaqui, have been found to be significantly compromised by harmful nonnative fish species. Unmack and Fagan (2004, p. 233) compared historical museum collections of nonnative fish species from the Gila River basin in Arizona and the Yaqui River basin in Sonora, Mexico, to gain insight into the trends in distribution, diversity, and abundance of nonnative fishes in each basin over time. They found that nonnative species are slowly, but steadily, increasing in all three parameters in the Yaqui Basin (Unmack and Fagan 2004, p. 233). Unmack and Fagan (2004, p. 233) predicted that, in the absence of aggressive management intervention, significant extirpations or range reductions of native fish species are expected to occur in the Yaqui Basin of Sonora, Mexico, which may have extant populations of the northern Mexican gartersnake, as did much of the Gila Basin before the introduction of nonnative species. Loss of native fishes impacts prey availability for the northern Mexican gartersnake and threatens its persistence in these areas. Black bullheads (

Ameiurus melas

) were reported as abundant, and common carp were detected from the Rio Yaqui in southern Sonora, Mexico (Branson

et al.

1960, p. 219). Bluegill (

Lepomis macrochirus

) were also reported at this location, representing a significant range expansion that the authors expected was the result of escaping nearby farm ponds or irrigation ditches (Branson

et al.

1960, p. 220). Largemouth bass, green sunfish, and an undetermined crappie species have also been reported from this area (Branson

et al.

1960, p. 220).

Documented problems with aquatic habitats in Mexico include water pollution, harmful nonnative species, and physical habitat alteration. All of these factors lead to declines in native fish abundance and, therefore, a decline in the food source for the northern Mexican gartersnake. Domínguez-Domínguez

et al.

(2007, p. 171) sampled 52 localities for a rare freshwater fish, the Picotee goodeid (

Zoogoneticus quitzeoensis

), along the southern portion of the Mesa Central (Mexican Plateau) of Mexico and found 21 localities had significant signs of pollution. Of the 29 localities where the target species was detected, 28 of them also had harmful nonnative species present, such as largemouth bass, cichlids (

Oreochromis

sp.), bluegill, and Pátzcuaro chub (

Algansea lacustris

) (Domínguez-Domínguez

et al.

2007, pp. 171, Table 3). The first assessment of the impacts of largemouth bass on native fishes in Mexico was in 1941 during the examination of their effect in Lago de Pátzcuaro (Contreras and Escalante 1984, p. 102). Other nonnative fish species reported are soft-rayed and small bodied, and may be prey items for younger age classes of gartersnakes.

Several examples of significant aquatic habitat degradation or destruction were also observed by Domínguez-Domínguez

et al.

(2007, Table 3) in this region of Mexico, including the draining of natural lakes and cienegas for conversion to agricultural purposes, modification of springs for recreational swimming, diversions, and dam construction. It should be noted that approximately 17 percent of the localities sampled by Domínguez-Domínguez

et al.

(2007, entire) are within the likely range of the northern Mexican gartersnake; chiefly sites located within the Rio Grande de Santiago and Laja Basin. However, collectively, observations made by Domínguez-Domínguez

et al.

(2007, entire) provide a regional context to potential threats acting on northern Mexican gartersnakes in their southern-most distribution. As of 2006, native fish species dominated the fish community in both species composition and overall abundance in the Laja Basin; however, the basin is now trending toward a nonnative fishery compared to historical data. For example, nonnative species were most recently collected from 16 of 17 sample sites in the basin, with largemouth bass significantly expanding their distribution within the headwaters of the basin and bluegill being widespread in the Laja River (Mercado-Silva

et al.

2006, pp. 537, 542, Table 4). The decline of native fishes in this region of Mexico is likely negatively affecting the status of the northern Mexican gartersnakes there.

Harmful nonnative fish species in Mexico (Contraras and Escalante 1984, pp. 102-125) may be posing a significant threat to the native fish prey base of northern Mexican gartersnakes and to the gartersnakes themselves. The ecological risk of nonnative, freshwater fishes is only expected to increase with increases in aquaculture production, most notably in the country's rural, poorest regions (Tapia and Zambrano 2003, p. 252). Amendments to Mexico's existing fishing regulations imposed by other government regulations have been relaxed, and investment in commercial fishing has expanded to promote growth in Mexico's aquaculture sector (Sugunan 1997, Section 8.7.1). Several areas within the range of the northern Mexican gartersnake in Mexico have experienced adverse effects associated with nonnative species.

Amphibian Decline (Northern Mexican Gartersnake)

—Amphibians are a principle prey item for the northern Mexican gartersnake, and documented declines in amphibian population densities and distributions have significantly contributed to the decline in northern Mexican gartersnakes. As an example of these effects from another region, Matthews

et al.

(2002, p. 16) examined the relationship of gartersnake distributions, amphibian population declines, and nonnative fish introductions in high-elevation aquatic ecosystems in California. Matthews

et al.

(2002, p. 16) specifically examined the effect of nonnative trout introductions on populations of amphibians and mountain gartersnakes (

Thamnophis elegans elegans

). Their results indicated that the probability of observing gartersnakes was 30 times greater in lakes containing amphibians than in lakes where amphibians have been extirpated by nonnative fish. These results supported a prediction by Jennings

et al.

(1992, p. 503) that native amphibian declines will lead directly to gartersnake declines.

Declines in the native leopard frog populations in Arizona have likely been a significant, contributing factor to declines in many northern Mexican gartersnake populations. Native ranid (of the family Ranidae) frog species, such as lowland leopard frogs, northern leopard frogs, and federally threatened Chiricahua leopard frogs, have experienced declines in various degrees throughout their distribution in the Southwest, largely due to predation and competition with nonnative species (Clarkson and Rorabaugh 1989, pp. 531, 535; Hayes and Jennings 1986, p. 490). Rosen

et al.

(1995, pp. 257-258) found that Chiricahua leopard frog distribution in the Chiricahua Mountain region of Arizona was inversely related to nonnative species distribution. Along the Mogollon Rim, Holycross

et al.

(2006, p. 13) found that only 8 sites of 57 surveyed (15 percent) consisted of an entirely native anuran (of the order Anura) community and that native frog populations in another 19 sites (33 percent) had been completely displaced by invading bullfrogs. However, such declines in native frog populations are not necessarily irreversible. Ranid frog populations have been shown to rebound strongly when nonnative fish are removed (Knapp

et al.

2007, pp. 15-18).

Scotia Canyon, in the Huachuca Mountains of southeastern Arizona, is a location where corresponding declines of leopard frog and northern Mexican

gartersnake populations have been documented through repeated survey efforts over time (Holm and Lowe 1995, p. 33). Surveys of Scotia Canyon occurred during the early 1980s and again during the early 1990s. Leopard frogs in Scotia Canyon were infrequently observed during the early 1980s and were nearly extirpated by the early 1990s (Holm and Lowe 1995, pp. 45-46). Northern Mexican gartersnakes were observed in decline during the early 1980s, with low capture rates continuing through the early 1990s (Holm and Lowe 1995, pp. 27-35). Surveys documented further decline of leopard frogs and northern Mexican gartersnakes in 2000 (Rosen

et al.

2001, pp. 15-16).

A former large, local population of northern Mexican gartersnakes at the San Bernardino National Wildlife Refuge (SBNWR) in southeastern Arizona has also experienced a correlative decline of leopard frogs, and northern Mexican gartersnakes are now thought to occur at very low population densities or may be extirpated there (Rosen and Schwalbe 1988, p. 28; 1995, p. 452; 1996, pp. 1-3; 1997, p. 1; 2002b, pp. 223-227; 2002c, pp. 31, 70; Rosen

et al.

1996b, pp. 8-9; 2001, pp. 6-10).

Survey data indicate that declines of leopard frog populations, often correlated with nonnative species introductions, the spread of a chytrid fungus (

Batrachochytrium dendrobatidis,

Bd), and habitat modification and destruction, have occurred throughout much of the northern Mexican gartersnake's U.S. distribution (Nickerson and Mays 1970, p. 495; Vitt and Ohmart 1978, p. 44; Ohmart

et al.

1988, p. 150; Rosen and Schwalbe 1988, Appendix I; 1995, p. 452; 1996, pp. 1-3; 1997, p. 1; 2002b, pp. 232-238; 2002c, pp. 1, 31; Clarkson and Rorabaugh 1989, pp. 531-538; Sredl

et al.

1995a, pp. 7-8; 1995b, pp. 8-9, 1995c, pp. 7-8; 2000, p. 10; Holm and Lowe 1995, pp. 45-46; Rosen

et al.

1996b, p. 2; 2001, pp. 2, 22; Degenhardt

et al.

1996, p. 319; Fernandez and Rosen 1996, pp. 6-20; Drost and Nowak 1997, p. 11; Turner

et al.

1999, p. 11; Nowak and Spille 2001, p. 32; Holycross

et al.

2006, pp. 13-14, 52-61). Holycross

et al.

(2006, pp. 53-57, 59) documented population declines and potential extirpations of lowland leopard frogs (an important prey species of the northern Mexican gartersnake) in most of the Agua Fria subbasin and areas of the Salt and Verde subbasins in the period 1986-2006. Specifically, Holycross

et al.

(2006, pp. 53-57, 59) detected no lowland leopard frogs at several recently, historically, or potentially occupied locations, including the Agua Fria River in the vicinity of Table Mesa Road and Little Grand Canyon Ranch, and at Rock Springs, Dry Creek from Dugas Road to Little Ash Creek, Little Ash Creek from Brown Spring to Dry Creek, Sycamore Creek (Agua Fria subbasin) in the vicinity of the Forest Service Cabin, the Page Springs and Bubbling Ponds fish hatchery along Oak Creek, Sycamore Creek (Verde River subbasin) in the vicinity of the confluence with the Verde River north of Clarkdale, along several reaches of the Verde River mainstem, Cherry Creek on the east side of the Sierra Ancha Mountains, and Tonto Creek from Gisela to “the Box,” near its confluence with Rye Creek. Rosen

et al.

(2013, p. 8) suggested that the decline of leopard frogs in the Empire Valley of southern Arizona is likely largely responsible for the decline of the northern Mexican gartersnake there.

A primary factor in the decline of native amphibians as a food source for northern Mexican gartersnakes in southern Arizona is likely the result of impacts from nonnative species, mainly bullfrogs. Rosen

et al.

(1995, pp. 252-253) sampled aquatic herpetofauna at 103 sites in the Chiricahua Mountains region, which included the Chiricahua, Dragoon, and Peloncillo Mountains, and the Sulphur Springs, San Bernardino, and San Simon valleys. They found that 43 percent of all ectothermic (cold-blooded) aquatic and semi-aquatic vertebrate species detected were nonnative. The most commonly encountered nonnative species was the bullfrog (Rosen

et al.

1995, p. 254). Witte

et al.

(2008, p. 1) found that the disappearance of ranid frog populations in Arizona were 2.6 times more likely in the presence of crayfish. Witte

et al.

(2008, p. 7) emphasized the significant influence of nonnative species on the disappearance of ranid frogs in Arizona. In one area, Rosen

et al.

(2001, p. 22) identified the expansion of bullfrogs into the Sonoita grasslands, which contain occupied northern Mexican gartersnake habitat, and the introduction of crayfish into Lewis Springs as being of particular concern for the northern Mexican gartersnake in that area.

In addition to harmful nonnative species, disease and nonnative parasites have been implicated in the decline of the prey base of the northern Mexican gartersnake. In particular, the outbreak of chytridiomycosis or “Bd,” a skin fungus, has been identified as a chief causative agent in the significant declines of many of the native ranid frogs and other amphibian species. As indicated, Bd has been implicated in both large-scale declines and local extirpations of many amphibians, chiefly anuran species, around the world (Johnson 2006, p. 3011). Lips

et al.

(2006, pp. 3166-3169) suggest that the high virulence and large number of potential hosts make Bd a serious threat to amphibian diversity. In Arizona, Bd infections have been reported in several of the native prey species of the northern Mexican gartersnake within the distribution of the snake (Morell 1999, pp. 731-732; Sredl and Caldwell 2000, p. 1; Hale 2001, pp. 32-37; Bradley

et al.

2002, p. 207; USFWS 2002, pp. 40802-40804; USFWS 2007a, pp. 26, 29-32). Declines of native prey species of the northern Mexican gartersnake from Bd infections have contributed to the decline of this species in the United States (Morell 1999, pp. 731-732; Sredl and Caldwell 2000, p. 1; Hale 2001, pp. 32-37; Bradley

et al.

2002, p. 207; USFWS 2002, pp. 40802-40804; USFWS 2007a, pp. 26, 29-32).

Evidence of Bd-related amphibian declines has been confirmed in portions of southern Mexico (just outside the range of northern Mexican gartersnakes), and data suggest declines are more prevalent at higher elevations where northern Mexican gartersnakes can occur (Lips

et al.

2004, pp. 560-562). However, much less is known about the role of Bd in amphibian declines across much of Mexico, in particular the mountainous regions of Mexico (including much of the range of northern Mexican gartersnakes in Mexico) as the region is significantly understudied (Young

et al.

2000, p. 1218). Because narrow-headed gartersnakes feed on fish, Bd has not affected their prey base. A recent study in Panama by Kilburn

et al.

(2011, p. 132) found that reptiles may act as reservoirs for

Bd

(at least in environments such as Panama) based on the presence of the fungus at non-pathological levels on lizards that occur in areas with significant

Bd

outbreaks in resident amphibians. Their study did not conclude that

Bd

is a virulent reptile pathogen, or that it causes disease-induced population declines in reptiles (Kilburn

et al.

2011, p. 132).

Effects of Bullfrogs on Native Aquatic Communities (Northern Mexican and Narrow-Headed Gartersnakes) (Factors A, C, and E)

Direct predation by, and competition with, bullfrogs is a serious threat to northern Mexican gartersnakes throughout their range (Conant 1974, pp. 471, 487-489; Rosen and Schwalbe 1988, pp. 28-30; Rosen

et al.

2001, pp. 21-22). Bullfrogs have and do threaten some populations of narrow-headed

gartersnakes, but differing habitat preferences between bullfrogs and narrow-headed gartersnakes lessen their effect on narrow-headed gartersnake populations. Bullfrogs adversely affect northern Mexican and narrow-headed gartersnake populations through direct predation of juveniles and sub-adults. Bullfrogs also compete with northern Mexican gartersnakes for prey species.

Bullfrogs are not native to the southwestern United States or Mexico, and they first appeared in Arizona in 1926 as a result of a systematic introduction effort by the State Game Department (now, the AGFD) for the purposes of sport hunting and as a food source (Tellman 2002, p. 43). The first bullfrog record from New Mexico is dated 1885 (Degenhardt

et al.

1996, p. 85). Bullfrogs are extremely prolific, are strong colonizers, can reach high densities, are persistent via cannibalism, and may disperse distances of up to 10 mi (16 km) across uplands and likely further within drainages (Bautista 2002, p. 131; Rosen and Schwalbe 2002a, p. 7; Casper and Hendricks 2005, p. 582; Suhre 2008, pers. comm.; Rosen

et al.

2013, pp. 35-36).

Bullfrogs are large-bodied, voracious, opportunistic, even cannibalistic predators that readily attempt to consume any living thing smaller than them. Bullfrogs have a highly varied diet, which has been documented to include vegetation, invertebrates, fish, birds, mammals, amphibians, and reptiles, including numerous species of snakes (eight genera, including six different species of gartersnakes, two species of rattlesnakes, and Sonoran gophersnakes (

Pituophis catenifer affinis

)) (Bury and Whelan 1984, p. 5; Clarkson and DeVos 1986, p. 45; Holm and Lowe 1995, pp. 37-38; Carpenter

et al.

2002, p. 130; King

et al.

2002; Hovey and Bergen 2003, pp. 360-361; Casper and Hendricks 2005, pp. 543-544; Combs

et al.

2005, p. 439; Wilcox 2005, p. 306; DaSilva

et al.

2007, p. 443; Neils and Bugbee 2007, p. 443; Rowe and Garcia 2012, pp. 633-634). In one study, three different species of gartersnakes (

Thamnophis sirtalis, T. elegans,

and

T. ordinoides

) totaling 11 snakes were found inside the stomachs of resident bullfrogs from a single region (Jancowski and Orchard 2013, p. 26). Bullfrogs can significantly reduce or eliminate the native amphibian populations (Moyle 1973, pp. 18-22; Conant 1974, pp. 471, 487-489; Hayes and Jennings 1986, pp. 491-492; Rosen and Schwalbe 1988, pp. 28-30; 2002b, pp. 232-238; Rosen

et al.

1995, pp. 257-258; 2001, pp. 2, Appendix I; Wu

et al.

2005, p. 668; Pearl

et al.

2004, p. 18; Kupferberg 1994, p. 95; Kupferburg 1997, pp. 1736-1751; Lawler

et al.

1999; Bury and Whelan 1986, pp. 9-10; Hayes and Jennings 1986, pp. 500-501; Jones and Timmons 2010, pp. 473-474), which are vital for northern Mexican gartersnakes.

Different age classes of bullfrogs can affect native ranid populations via different mechanisms. Juvenile bullfrogs affect native ranids through competition; male bullfrogs affect native ranids through predation; and female bullfrogs affect native ranids through both mechanisms depending on body size and microhabitat (Wu

et al.

2005, p. 668). Pearl

et al.

(2004, p. 18) also suggested that the effect of bullfrog introductions on native ranids may be different based on specific habitat conditions but also suggested that an individual ranid frog species' physical ability to escape influences the effect of bullfrogs on each native ranid community. Bullfrogs can also negatively affect native ranid frog populations, both locally and regionally, as carriers or reservoir species for Bd, depending on the strain of Bd (Gervasi

et al.

2013, p. 169).

Bullfrogs have been documented to occur throughout Arizona. Holycross

et al.

(2006, pp. 13-14, 52-61) found bullfrogs at 55 percent of sample sites in the Agua Fria subbasin, 62 percent of sites in the Verde River subbasin, 25 percent of sites in the Salt River subbasin, and 22 percent of sites in the Gila River subbasin. In total, bullfrogs were observed at 22 of the 57 sites surveyed (39 percent) across the Mogollon Rim (Holycross

et al.

2006, p. 13). A number of authors have also documented the presence of bullfrogs throughout many subbasins in Arizona and New Mexico adjacent to the historical distribution of the northern Mexican or narrow-headed gartersnake, including northern Arizona (Sredl

et al.

1995a, p. 7; 1995c, p. 7), central Arizona and along the Mogollon Rim of Arizona and New Mexico (Nickerson and Mays 1970, p. 495; Hulse 1973, p. 278; Sredl

et al.

1995b, p. 9; Drost and Nowak 1997, p. 11; Nowak and Spille 2001, p. 11; Holycross

et al.

2006, pp. 15-51; Wallace

et al.

2008; pp. 243-244; Hellekson 2012a, pers. comm.), southern Arizona (Rosen and Schwalbe 1988, Appendix I; 1995, p. 452; 1996, pp. 1-3; 1997, p. 1; 2002b, pp. 223-227; 2002c, pp. 31, 70; Holm and Lowe 1995, pp. 27-35; Rosen

et al.

1995, p. 254; 1996a, pp. 16-17; 1996b, pp. 8-9; 2001, Appendix I; Turner

et al.

1999, p. 11; Sredl

et al.

2000, p. 10; Turner 2007; p. 41), and along the Colorado River (Vitt and Ohmart 1978, p. 44; Clarkson and DeVos 1986, pp. 42-49; Ohmart

et al.

1988, p. 143). In one of the more conspicuous examples, bullfrogs were identified as the primary cause for collapse of the northern Mexican gartersnake and its prey base on the SBNWR (Rosen and Schwalbe 1988, p. 28; 1995, p. 452; 1996, pp. 1-3; 1997, p. 1; 2002b, pp. 223-227; 2002c, pp. 31, 70; Rosen

et al.

1996b, pp. 8-9).

Once established, bullfrogs are persistent in an area and very difficult to eradicate. Rosen and Schwalbe (1995, p. 452) experimented with bullfrog removal at various sites on the SBNWR, in addition to a control site with no bullfrog removal in similar habitat on the Buenos Aires National Wildlife Refuge (BANWR). Removal of adult bullfrogs, without removal of eggs and tadpoles, resulted in a substantial increase in younger age-class bullfrogs where removal efforts were the most intensive (Rosen and Schwalbe 1997, p. 6). Contradictory to the goals of bullfrog eradication, evidence from dissection samples from young adult and subadult bullfrogs indicated these age-classes readily prey upon juvenile bullfrogs (up to the average adult leopard frog size) as well as juvenile gartersnakes, which suggests that the selective removal of only the large adult bullfrogs (presumed to be the most dangerous size class to leopard frogs and gartersnakes), favoring the young adult and sub-adult age classes, could indirectly lead to increased predation of leopard frogs and juvenile gartersnakes (Rosen and Schwalbe 1997, p. 6). These findings illustrate that, in addition to large adults, sub-adult bullfrogs also negatively impact northern Mexican gartersnakes and their prey species. The findings also indicate the importance of including egg mass and tadpole removal during efforts to control bullfrogs and timing removal projects to ensure reproductive bullfrogs are removed prior to breeding. Recent success in regional bullfrog eradication has been found in a few cases described below in the section entitled “

Current Conservation of Northern Mexican and Narrow-headed Gartersnakes.

”

Bullfrogs not only compete with the northern Mexican gartersnake for prey items but directly prey upon juvenile and, occasionally, sub-adult northern Mexican and narrow-headed gartersnakes (Rosen and Schwalbe 1988, pp. 28-31; 1995, p. 452; 2002b, pp. 223-227; Holm and Lowe 1995, pp. 29-29; Rossman

et al.

1996, p. 177; AGFD

In Prep.,

p. 12; 2001, p. 3; Rosen

et al.

2001, pp. 10, 21-22; Carpenter

et al.

2002, p. 130; Wallace 2002, p. 116). A well-circulated photograph of an adult bullfrog in the process of consuming a northern Mexican gartersnake at Parker Canyon Lake, Cochise County, Arizona,

taken by John Carr of the AGFD in 1964, provides photographic documentation of bullfrog predation (Rosen and Schwalbe 1988, p. 29; 1995, p. 452). The most recent, physical evidence of bullfrog predation of northern Mexican gartersnakes is provided in photographs of a dissected bullfrog at Pasture 9 Tank in the San Rafael Valley of Arizona that had a freshly eaten neonatal northern Mexican gartersnake in its stomach (Akins 2012, pers. comm.).

A common observation in northern Mexican gartersnake populations that co-occur with bullfrogs is a preponderance of large, mature adult snakes with conspicuously low numbers of individuals in the newborn and juvenile age size classes. This occurs due to bullfrogs preying on young small snakes more effectively, which leads to reduced survival of young and depressed recruitment within populations (Rosen and Schwalbe 1988, p. 18; Holm and Lowe 1995, p. 34). In lotic (flowing water) systems, bullfrogs prefer sites with low or limited flow, such as backwaters, side channels, and pool habitat. These areas are also used frequently by northern Mexican and narrow-headed gartersnakes, which likely results in increased predation rates and likely depressed recruitment of gartersnakes. Potential recruitment problems for northern Mexican gartersnakes due to effects from nonnative species are suspected at Tonto Creek (Wallace

et al.

2008, pp. 243-244). Rosen and Schwalbe (1988, p. 18) stated that the low recruitment at the SBNWR, a typical characteristic of gartersnake populations affected by harmful nonnative species, is the likely cause of that populations' decline and possibly for declines in populations throughout their range in Arizona. Specific localities within the distribution of northern Mexican and narrow-headed gartersnakes where bullfrogs have been detected are presented in Appendix A (available at

http://www.regulations.gov

, Docket No. FWS-R2-ES-2013-0071).

Effects of Crayfish on Native Aquatic Communities (Northern Mexican and Narrow-Headed Gartersnakes) (Factors A and C)

Crayfish are another nonnative species in Arizona and New Mexico that threaten northern Mexican and narrow-headed gartersnakes through competition by consuming prey species of the gartersnakes and through direct predation on juvenile gartersnakes themselves (Fernandez and Rosen 1996, p. 25; Voeltz 2002, pp. 87-88; USFWS 2007a, p. 22). Rogowski

et al.

(2013, p. 1,280) found Arizona's aquatic communities to be particularly vulnerable to crayfish because many endemic aquatic species never evolved in the presence of crayfish. Fernandez and Rosen (1996, p. 3) studied the effects of crayfish introductions on two stream communities in Arizona, a low-elevation semi-desert stream and a high mountain stream, and concluded that crayfish can noticeably reduce species diversity and destabilize food chains in riparian and aquatic ecosystems through their effect on vegetative structure, stream substrate (stream bottom;

i.e.,

silt, sand, cobble, boulder) composition, and predation on eggs, larval, and adult forms of native invertebrate and vertebrate species. Crayfish fed on embryos, tadpoles, newly metamorphosed frogs, and adult leopard frogs, but they did not feed on egg masses (Fernandez and Rosen 1996, p. 25). However, Gamradt and Kats (1996, p. 1155) found that crayfish readily consumed the egg masses of California newts (

Taricha torosa

). Crayfish are known to also eat fish eggs and larva (Inman

et al.

1998, p. 17), especially those bound to the substrate (Dorn and Mittlebach 2004, p. 2135). Fernandez and Rosen (1996, pp. 6-19, 52-56) and Rosen (1987, p. 5) discussed observations of inverse relationships between crayfish abundance and native reptile and amphibian populations, including narrow-headed gartersnakes, northern leopard frogs, and Chiricahua leopard frogs. Crayfish may also affect native fish populations. Carpenter (2005, pp. 338-340) documented that crayfish may reduce the growth rates of native fish through competition for food and noted that the significance of this impact may vary between species.

Crayfish alter the abundance and structure of aquatic vegetation by grazing on aquatic and semiaquatic vegetation, which reduces the cover needed by frogs and gartersnakes, as well as the food supply for prey species such as tadpoles (Fernandez and Rosen 1996, pp. 10-12). Fernandez and Rosen (1996, pp. 10-12) found that crayfish frequently burrow into stream banks, leading to increased bank erosion, stream turbidity, and siltation of stream bottoms. Creed (1994, p. 2098) found that filamentous alga (

Cladophora glomerata

) was at least 10-fold greater in aquatic habitats that lacked crayfish. Filamentous algae is an important component of aquatic vegetation that provides cover for foraging gartersnakes, as well as microhabitat for prey species, in situations where predation risk is high.

Crayfish have recently been found to also act as a host for the amphibian disease-causing fungus, Bd (McMahon

et al.

2013, pp. 210-213). This could have serious implications for northern Mexican gartersnakes because crayfish can now be considered a source of disease in habitat that is devoid of amphibians but otherwise potentially suitable habitat for immigrating amphibians, such as leopard frogs, which could serve as a prey base. Because crayfish are so widespread throughout Arizona, New Mexico, and portions of Mexico, the scope of this threat is significant for native amphibian populations and, therefore, to northern Mexican gartersnake populations.

Inman

et al.

(1998, p. 3) documented crayfish as widely distributed and locally abundant in a broad array of natural and artificial free-flowing and still-water habitats throughout Arizona, many of which overlap the historical and current distribution of northern Mexican and narrow-headed gartersnakes. Hyatt (undated, p. 71) concluded that the majority of waters in Arizona contained at least one species of crayfish. In surveying for northern Mexican and narrow-headed gartersnakes, Holycross

et al.

(2006, p. 14) found crayfish in 64 percent of the sample sites in the Agua Fria subbasin; in 85 percent of the sites in the Verde River subbasin; in 46 percent of the sites in the Salt River subbasin; and in 67 percent of the sites in the Gila River subbasin. In total, crayfish were observed at 35 (61 percent) of the 57 sites surveyed across the Mogollon Rim (Holycross

et al.

2006, p. 14), most of which were sites historically or currently occupied by northern Mexican or narrow-headed gartersnakes, or sites the investigators believed possessed suitable habitat and may be occupied by these gartersnakes based upon their known historical distributions.

A number of authors have documented the presence of crayfish through their survey efforts throughout Arizona and New Mexico in specific regional areas, drainages, and lentic wetlands within or adjacent to the historical distribution of the northern Mexican or narrow-headed gartersnake, including northern Arizona (Sredl

et al.

1995a, p. 7; 1995c, p. 7), central Arizona and along the Mogollon Rim of Arizona and New Mexico (Sredl

et al.

1995b, p. 9; Fernandez and Rosen 1996, pp. 54-55, 71; Inman

et al.

1998, Appendix B; Nowak and Spille 2001, p. 33; Holycross

et al.

2006, pp. 15-51; Brennan 2007, p. 7; Burger 2008, p. 4; Wallace

et al.

2008; pp. 243-244; Brennan and Rosen 2009, p. 9; Karam

et al.

2009; pp. 2-3; Hellekson 2012a, pers. comm.), southern Arizona (Rosen and Schwalbe 1988, Appendix I; Inman

et al.

1998,

Appendix B; Sredl

et al.

2000, p. 10; Rosen

et al.

2001, Appendix I), and along the Colorado River (Ohmart

et al.

1988, p. 150; Inman

et al.

1998, Appendix B). Specific localities within the distribution of northern Mexican and narrow-headed gartersnakes where crayfish have been detected are presented in Appendix A (available at

http://www.regulations.gov,

Docket No. FWS-R2-ES-2013-0071). Like bullfrogs, crayfish can be very difficult, if not impossible, to eradicate once they have become established in an area, depending on the complexity of the habitat (Rosen and Schwalbe 1996a, pp. 5-8; 2002a, p. 7; Hyatt undated, pp. 63-71).

It is likely that crayfish populations, where they overlap with northern Mexican or narrow-headed gartersnakes, could have a varied influence on gartersnake populations. The size of crayfish can influence their predatory influence on gartersnakes or their prey species; small crayfish are unlikely to pose a significant threat to gartersnakes themselves but may still consume fish eggs or fry, whereas larger crayfish can prey on neonatal gartersnakes directly. The presence of adequate numbers of favorable fish prey for narrow-headed gartersnakes may counter the effects of resident crayfish to some degree. Crayfish densities may also be affected by periodic flooding, which is thought to reduce crayfish population densities temporarily until recolonization occurs from the dispersal of individuals from downstream populations. More field research is needed to fully understand the ecological relationship between crayfish and these gartersnakes, at least at any particular site. However, the best available scientific and commercial information strongly suggests that crayfish in larger size classes or in high densities are a cause for concern for gartersnakes and their prey species, especially with other threats simultaneously affecting gartersnake populations.

Effects of Predation-Related Injuries to Gartersnakes (Northern Mexican and Narrow-Headed Gartersnakes) (Factor C)

The tails of gartersnakes are often broken off during predation attempts by bullfrogs, crayfish, or other predators, and do not regenerate. The incidence of tail breaks in gartersnakes can often be used to assess predation pressure within gartersnake populations. Attempted predation occurs on both sexes and all ages of gartersnakes within a population, although some general trends have been detected. For example, female gartersnakes may be more susceptible to predation as evidenced by the incidence of tail damage (Willis

et al.

1982, pp. 100-101; Rosen and Schwalbe 1988, p. 22; Mushinsky and Miller 1993, pp. 662-664; Fitch 2003, p. 212). This can be explained by higher basking rates associated with pregnant females that increase their visibility to predators. Fitch (2003, p. 212) found that tail injuries in the common gartersnake occurred more frequently in adults than in juveniles. Predation on juvenile snakes likely results in complete consumption of the animal, which would limit observations of tail injury in their age class.

Tail injuries can have negative effects on the health, longevity, and overall success of individual gartersnakes from infection, slower swimming and crawling speeds, or impeding reproduction. Mushinsky and Miller (1993, pp. 662-664) commented that, while tail breakage in gartersnakes can save the life of an individual snake, it also leads to permanent handicapping of the snake, resulting in slower swimming and crawling speeds, which could leave the snake more vulnerable to predation or affect its foraging ability. Willis

et al.

(1982, p. 98) discussed the incidence of tail injury in three species in the genus

Thamnophis

(common gartersnake, Butler's gartersnake (

T. butleri

), and the eastern ribbon snake (

T. sauritus

)) and concluded that individuals that suffered nonfatal injuries prior to reaching a length of 12 in (30 cm) are not likely to survive and that physiological stress during post-injury hibernation may play an important role in subsequent fatality. While northern Mexican or narrow-headed gartersnakes may survive an individual predation attempt from a bullfrog or crayfish with tail damage, secondary effects from infection of the wound may significantly contribute to fatality of individuals. Perry-Richardson

et al.

(1990, p. 77) described the importance of tail-tip alignment in the successful courtship and mating in Thamnophiine snakes and found that missing or shortened tails adversely affected these activities and, therefore, mating success. In researching the role of tail length in mating success in the red-sided gartersnake (

Thamnophis sirtalis parietalis

), Shine

et al.

(1999, p. 2150) found that males that experienced injuries or the partial or whole loss of the tail experienced a three-fold decrease in mating success.

The frequency of tail injuries can be quite high in a given gartersnake population; for example at the SBNWR (Rosen and Schwalbe 1988, pp. 28-31), 78 percent of northern Mexican gartersnakes had broken tails with a “soft and club-like” terminus, which suggests repeated injury from multiple predation attempts by bullfrogs. While medically examining pregnant female northern Mexican gartersnakes, Rosen and Schwalbe (1988, p. 28) noted bleeding from the posterior region, which suggested to the investigators the snakes suffered from “squeeze-type” injuries inflicted by adult bullfrogs. In another example, Holm and Lowe (1995, pp. 33-34) observed tail injuries in 89 percent of northern Mexican gartersnakes during the early 1990s in Scotia Canyon in the Huachuca Mountains, as well as a skewed age class ratio that favored adults over sub-adults, which is consistent with data collected by Willis

et al.

(1982, pp. 100-101) on other gartersnake species. Bullfrogs are largely thought to be responsible for the significant decline of northern Mexican gartersnake and its prey base at this locality, although the latter has improved through recovery actions. In the Black River, crayfish are very abundant and have been identified as the likely cause for a high-frequency of tail injuries to narrow-headed gartersnakes (Brennan 2007, p. 7; Brennan and Rosen 2009, p. 9). Brennan (2007, p. 5) found that, in the Black River, 14 of 15 narrow-headed gartersnakes captured showed evidence of damaged or missing tails (Brennan 2007, p. 5). In 2009, 16 of 19 narrow-headed gartersnakes captured in the Black River showed evidence of damaged or missing tails (Brennan and Rosen 2009, p. 8). In the middle Verde River region, Emmons and Nowak (2013, p. 5) reported that 18 of 49 (37 percent) northern Mexican gartersnakes captured had scars (n = 17) and/or missing tails tips (n = 7).

Vegetation or other forms of protective cover may be particularly important for gartersnakes to reduce the effects of harmful nonnative species on populations. For example, the population of northern Mexican gartersnakes at the Page Springs and Bubbling Ponds State Fish Hatcheries occurs with harmful nonnative species (Boyarski 2008b, pp. 3-4, 8). Yet, only 11 percent of northern Mexican gartersnakes captured in 2007 were observed as having some level of tail damage (Boyarski 2008b, pp. 5, 8). The relatively low occurrence of tail damage, as compared to 78 percent of snakes with tail damage found by Rosen and Schwalbe (1988, pp. 28-31), may indicate: (1) Adequate vegetation density was used by gartersnakes to avoid harmful nonnative species predation attempts; (2) a relatively small population of harmful nonnative species may be at a comparatively lower density than sites sampled by previous studies

(harmful nonnative species population density data were not collected by Boyarski (2008b)); (3) gartersnakes may not have needed to move significant distances at this locality to achieve foraging success, which might reduce the potential for encounters with harmful nonnative species; or (4) gartersnakes infrequently escaped predation attempts by harmful nonnative species, were removed from the population, and were consequently not detected by surveys.

Expansion of the American Bullfrog and Crayfish in Mexico (Northern Mexican Gartersnake) (Factors A, C, and E)

Bullfrogs are a significant threat to native aquatic and riparian species throughout Mexico. Luja and Rodríguez-Estrella (2008, pp. 17-22) examined the invasion of the bullfrog in Mexico. The earliest records of bullfrogs in Mexico were Nuevo Leon (1853), Tamaulipas (1898), Morelos (1968), and Sinaloa (1969) (Luja and Rodríguez-Estrella 2008, p 20). By 1976, the bullfrog was documented in seven more states: Aguacalientes, Baja California Sur, Chihuahua, Distrito Federal, Puebla, San Luis Potosi, and Sonora (Luja and Rodríguez-Estrella 2008, p. 20). The bullfrog was recently verified from the state of Hidalgo, Mexico, at an elevation of 8,970 feet (2,734 m), which indicates the species continues to spread in that country and can exist even at the uppermost elevations inhabited by northern Mexican gartersnakes (Duifhuis Rivera

et al.

2008, p. 479). As of 2008, Luja and Rodríguez-Estrella (2008, p. 20) have recorded bullfrogs in 20 of the 31 Mexican States (65 percent of the states in Mexico) and suspect that they have invaded other States, but were unable to find documentation.

Bullfrogs have been commercially produced for food in Mexico in Yucatan, Nayarit, Morelos, Estado de Mexico, Michoacán, Guadalajara, San Luis Potosi, Tamaulipas, and Sonora, and their use for food was endorsed by the Mexican Secretary of Aquaculture Support (Luja and Rodríguez-Estrella 2008, p. 20). However, frog legs ultimately never gained popularity in Mexican culinary culture (Conant 1974, pp. 487-489), and Luja and Rodríguez-Estrella (2008, p. 22) point out that only 10 percent of these farms remain in production. Luja and Rodríguez-Estrella (2008, pp. 20, 22) document instances where bullfrogs have escaped production farms and suspect the majority of the frogs that were produced commercially in farms that have since ceased operation have assimilated into surrounding habitat.

Luja and Rodríguez-Estrella (2008, p. 20) also state that Mexican people deliberately introduce bullfrogs for ornamental purposes, or “for the simple pleasure of having them in ponds.” The act of deliberately releasing bullfrogs into the wild in Mexico was cited by Luja and Rodríguez-Estrella (2008, p. 21) as being “more common than we can imagine.” Bullfrogs are available for purchase at some Mexican pet stores (Luja and Rodríguez-Estrella 2008, p. 22). Luja and Rodríguez-Estrella (2008, p. 21) state that bullfrog eradication efforts in Mexico are often thwarted by their popularity in rural communities (presumably as a food source). Currently, no regulation exists in Mexico to address the threat of bullfrog invasions or prevent their release into the wild (Luja and Rodríguez-Estrella 2008, p. 22). As a result, the bullfrogs' distribution continues to increase in Mexico, beyond what it would through natural dispersal mechanisms.

Rosen and Melendez (2006, p. 54) report bullfrog invasions to be prevalent in northwestern Chihuahua and northwestern Sonora, where the northern Mexican gartersnake is thought to occur. In many areas, native leopard frogs were completely displaced where bullfrogs were observed. Rosen and Melendez (2006, p. 54) also demonstrated the relationship between fish and amphibian communities in Sonora and western Chihuahua. Native leopard frogs, a primary prey item for the northern Mexican gartersnake, only occurred in the absence of nonnative fish, and were absent from waters containing nonnative species, which included several major waters. In Sonora, Rorabaugh (2008a, p. 25) also considers the bullfrog to be a significant threat to the northern Mexican gartersnake and its prey base, substantiated by field observations made during surveys conducted in Chihuahua and Sonora in 2006 (Rorabaugh 2008b, p. 1).

Few data were found on the presence or distribution of nonnative crayfish species in Mexico. However, in a 2-week gartersnake survey effort in 2006 in northern Mexico, crayfish were observed as “widely distributed” in the valleys of western Chihuahua (Rorabaugh 2008b, p. 1). Based on the invasive nature of crayfish ecology and their distribution in the United States along the Border region, it is reasonable to assume that, at a minimum, crayfish are likely distributed along the entire Border region of northern Mexico, adjacent to where they occur in the United States, and act in a similar fashion on affected northern Mexican gartersnake populations.

Risks to Gartersnakes From Fisheries Management Activities (Northern Mexican and Narrow-Headed Gartersnakes) (Factors A and E)

The decline in native fish communities from the effects of harmful nonnative fish species has spurred resource managers to take action to help recover native fish species. While we fully support activities designed to help recover native fish, recovery actions for native fish, in the absence of thorough planning, can have negative effects on resident gartersnake populations.

Piscicides

—Piscicide is a term that refers to a “fish poison.” The use of piscicides, such as rotenone or antimycin A, for the removal of harmful nonnative fish species has widely been considered invaluable for the conservation and recovery of imperiled native fish species throughout the United States, and in particular the Gila River basin of Arizona and New Mexico (Dawson and Kolar 2003, entire). Antimycin A is rarely used anymore due to limited production and has been largely replaced by rotenone in field applications. Experimentation with ammonia as a piscicide has shown promising results and may ultimately replace rotenone in the future as a desired control method if legally registered for such use (Ward

et al.

2013, pp. 402-404). Currently, rotenone is the most commonly used piscicide. The active ingredient in rotenone is a natural chemical compound extracted from the stems and roots of tropical plants in the family Leguminosae that interrupts oxygen absorption in gill-breathing animals (Fontenot

et al.

1994, pp. 150-151). In the greater Gila River subbasin alone, 57 streams or water bodies have been treated with piscicide, some on several occasions spanning many years (Carpenter and Terrell 2005; Table 6). However, this practice has been the source of recent controversy due to a perceived link between rotenone and Parkinson's disease in humans, as well as potential effects to livestock.

Speculation of the potential role of rotenone in Parkinson's disease was fueled by Tanner

et al.

(2011, entire), which correlated the incidence of the disease with lifetime exposure to certain pesticides, including rotenone. As a result, in 2012, the Arizona State Legislature proposed two bills that called for the development of an environmental impact statement prior to the application of rotenone or antimycin A (S.B. 1453, see State of Arizona Senate (2012b)) and urged the U.S. Environmental Protection Agency to deregister rotenone from use in the

United States (S.B. 1009, see State of Arizona Senate (2012b)). Public safety considerations were fully evaluated by a multidisciplined technical team of specialists that found no correlation between rotenone applications performed, according to product label instructions, and Parkinson's disease (Rotenone Review Advisory Committee 2012, pp. 24-25). Nonetheless, continued anxiety regarding the use of piscicides for conservation and management of fish communities leaves an uncertain future for this important management tool. Should circumstances result in the discontinued practice of using piscicides for fish recovery and management, the likelihood of recovery for listed or sensitive aquatic vertebrates in Arizona, such as northern Mexican and narrow-headed gartersnakes, would be substantially reduced, if not eliminated outright.

The use of piscicides is a vital and scientifically sound tool, the only tool, in most circumstances, for reestablishing native fish communities and removing threats related to nonnative aquatic species in occupied northern Mexican and narrow-headed gartersnake habitat. By extension, the use of piscicides is also invaluable in the recovery and conservation of northern Mexican and narrow-headed gartersnakes. However, without proper planning the amount of time a treated water body remains fishless post-treatment can affect gartersnakes by removing fish, their primary food source. The time period between rotenone applications and the subsequent restocking of native fish is contingent on two basic variables, the time it takes for piscicide levels to reach nontoxic levels and the level of certainty required to ensure that renovation goals and objectives have been met prior to restocking. Implementation of the latter consideration may vary from to a year or longer, depending on the level of certainty required by project proponents. Carpenter and Terrell (2005, p. 14) reported that standard protocols used by the AGFD for Apache trout renovations at that time required two applications of piscicide before repatriating native fish to a stream, waiting a season to see if the renovation was successful, and then continuing to renovate if necessary. Past protocols have included goals for the renovated water body to remain fishless for extended periods, sometimes up to an entire year before restocking (Carpenter and Terrell 2005, p. 14). At a minimum and according to our files, reaches of Big Bonito Creek, the West Fork Black River, West Fork Gila River, Little Creek, and O'Donnell Creek have all been subject to fish renovations using these or similarly accepted protocols (Carpenter and Terrell 2005; Table 6; Paroz and Propst 2009, p. 4; Hellekson 2012a, pers. comm.). Therefore, northern Mexican or narrow-headed gartersnake populations in these streams have likely been negatively affected, due to the eradication of a portion of, or their entire, prey base in these systems for varying periods of time. Big Bonito Creek was restocked with salvaged native fish shortly after renovation occurred. However, we are uncertain how long other stream reaches remained fishless post-treatment, but it was likely to be a minimum of weeks in each instance, and possibly a year or longer in some instances.

Although significant efforts are generally made to salvage as many native fish as possible prior to treatment, logistics of holding fish for several weeks prior to restocking limit the number of individuals that can be held safely. Therefore, not every individual fish is salvaged, and native fish remaining in the stream are subsequently lost during the treatment. The number of fish subsequently restocked is, therefore, smaller than the number of fish that were present prior to the treatment. The full restoration of native fish populations to pre-treatment levels may take several years, depending on the size of the treated area and the size and maturity of the founding populations. Restocking salvaged fish in the fall may allow natural spawning and recruitment to begin in the spring, which would provide a more immediate benefit to resident gartersnake populations.

Several streams within the distribution of narrow-headed gartersnakes in New Mexico have been identified for potential future fish barrier construction, for which piscicide applications are likely necessary. These streams include Little Creek, West Fork Gila River, Middle Fork Gila River, Turkey Creek, Saliz Creek, Dry Blue Creek, Iron Creek, and the San Francisco River (Riley and Clarkson 2005, pp. 4-5, 7, 9, 12; Clarkson and Marsh 2012, p. 8; 2013, pp. 1, 4, 6; Hellekson 2013, pers. comm.). Of these, the Middle Fork Gila River and Turkey Creek appear to the most likely chosen for renovation (Clarkson and Marsh 2013, p. 8). Mule Creek and Cienega Creek, both occupied by northern Mexican gartersnakes, as well as Whitewater Creek (occupied by narrow-headed gartersnakes), are under consideration but ultimately may not be chosen (Clarkson and Marsh 2013, pp. 8-9). Haigler Creek (occupied by narrow-headed gartersnakes) is planned for renovation in 2015 (Burger and Jeager 2013, p. 2) and barrier development.

The current standard operating procedures for piscicide application, as adopted nationally and provided in Finlayson

et al.

(2010, p. 23), provide guidance for assuring that nontarget, baseline environmental conditions (the biotic community) are accounted for in assessing whether mitigation measures are necessary. This procedural protocol states, “Survival and recovery of the aquatic community may be demonstrated by sampling plankton, macroinvertebrates (aquatic insects, crustacea, leeches, and mollusks), and amphibians (frogs, tadpoles, and larval and adult salamanders)” (Finlayson

et al.

2010, p. 23). This protocol, adopted by the AGFD (see AGFD 2012a), does not in itself consider the effects of leaving a treated water body without a prey base for a sensitive species much less for a fish-specialist, such as the narrow-headed gartersnake, for extended periods of time. However, the AGFDs' internal Environmental Assessment Checklist (EAC) addresses considerations for nontarget aquatic reptiles. Thus, we believe that concerns for potential effects of piscicide treatments on these gartersnake species within Arizona should not be substantial in the future.

As of 2012, a new policy was finalized by the AGFD that includes an early and widespread public notification and planning process that involves the approval of several decision-makers within four major stages: (1) Piscicide project internal review and approval; (2) preliminary planning and public involvement; (3) intermediate planning and public involvement; and (4) project implementation and evaluation (AGFD 2012a, p. 3). Within the Internal Review and Approval stage of the process, sensitive, endemic, and listed species potentially impacted by the project must be identified (AGFD 2012a, p. 13), such as northern Mexican or narrow-headed gartersnakes. This change ensures that an analysis of potential effects to nontarget wildlife by fisheries management activities occurs within the same planning document, versus a separate process. In addition, the AGFD's Conservation and Mitigation Program has specifically committed to quickly restocking renovated streams that are occupied by either northern Mexican or narrow-headed gartersnakes (USFWS 2011, Appendix C).

Piscicide application protocols used by the New Mexico Department of Game and Fish are provided in Pierce (2014,

entire) and specify that effects to amphibian species are reviewed prior to application; however, the protocol does not provide for an assessment of potential gartersnake effects from treatment. No specific timeframe, post-treatment, was recommended by the protocol for when native fish are recommended for stocking into treated waters (Pierce 2014, pers. comm.). We intend to coordinate with the New Mexico Department of Game and Fish as active partners in wildlife conservation to ensure potential effects, from piscicide treatments, to either gartersnake are avoided or minimized. However, if proper protocols are not incorporated into future fish restoration projects, these activities will continue to threaten local gartersnake populations.

Mechanical Methods

—In addition to chemical renovation techniques, mechanical methods using electroshocking equipment are often used in fisheries management, both for nonnative aquatic species removal and fisheries survey and monitoring activities that often occur in conjunction with piscicide treatments. Northern Mexican and narrow-headed gartersnakes often flee into the water as a first line of defense when startled. In occupied habitat, gartersnakes present in the water and within the affected radius of electroshockers are often temporarily paralyzed from electrical impulses intended for fish, and are, therefore, readily detected by surveyors (Hellekson 2012a, pers. comm.). We are not aware of any research that has investigated potential short- or long-term consequences to gartersnakes from these events, and so we do not consider electroshock surveys as a substantial threat to either gartersnake.

Trapping methods are also used in fisheries surveys, for other applications in aquatic species management, and for the collection of live baitfish in recreational fishing. One such common method to study aquatic or semi-aquatic wildlife (including populations of aquatic snakes such as gartersnakes) is through the use of wire minnow traps. When used to monitor gartersnake populations, wire minnow traps are anchored to vegetation, logs, etc., along the shoreline (in most applications) and positioned so that half to one-third of the trap, along its lateral line, is above the water surface to allow snakes to surface for air. These traps often attract prey species, such as small fishes and amphibian larvae (when present), and, therefore, become self-baiting. They are then checked according to a predetermined schedule. Because the wire, twine, etc., used to anchor these traps is fixed in length, these traps may become fully submerged if there is a sudden, unanticipated rise in water levels (e.g., storm event). During the monsoon in Arizona and New Mexico, these types of storm events are common, and river hydrographs respond accordingly with rapid and dynamic increases in flow.

We are aware of examples where northern Mexican gartersnakes, intentionally captured in minnow traps, have drowned as a direct result of a rapid, unexpected rise in water levels. Some examples include an adult female northern Mexican gartersnake along lower Tonto Creek in 2004, an adult and two neonates at the Bubbling Ponds State Fish Hatchery in 2009 and 2010, respectively, and an individual of undisclosed age in the upper Santa Cruz River (Holycross

et al.

2006, p. 41, Boyarski 2011, pp. 2-3; Lashway 2012, p. 5). In another example, involving an underwater funnel trap used to survey for lowland leopard frogs (but which are not used for fishery surveys), a large adult female northern Mexican gartersnake was discovered deceased in the trap (Jones 2012a, pers. comm.). Death of that individual was likely due to drowning or predation by numerous crayfish that were also confined in the funnel trap with the gartersnake (Jones 2012a, pers. comm.). Depending on the mesh size of traps, neonatal gartersnakes can become stuck in the mesh of traps (Lashway 2012, p. 5), which could result in injury or death of the individual. There are likely additional cases where northern Mexican or narrow-headed gartersnake fatality from trapping has not been reported, particularly where trapping has occurred in occupied habitat prone to flash flooding.

Minnow traps are often deployed for monitoring fully aquatic species, such as fish, and are, therefore, intentionally positioned in the water column where they are fully under water. Traps used for this purpose may be checked less frequently, because risks to gill-breathing aquatic species are less if held in the trap for longer periods of time. As fish collectively become trapped, the trap becomes incidentally self-baited for gartersnakes and, if deployed in habitat occupied by either northern Mexican or narrow-headed gartersnakes, these traps may accidentally attract, capture, and drown gartersnakes that are actively foraging under water and are lured to the traps because of captured prey species. Neonatal northern Mexican and narrow-headed gartersnakes can also wriggle through the mesh of some wire minnow traps and become lodged halfway through, depending on the pore size of the wire mesh (Jaeger 2012, pers. comm.). If not found in time, this situation would likely result in their death from drowning, predation, or exposure.

The use of minnow traps is also allowed in recreational fishing in Arizona and New Mexico (AGFD 2013a, p. 57; New Mexico Department of Game and Fish (NMDGF) 2013, p. 17). In Arizona and New Mexico, it is lawful to set minnow traps for the collection of live baitfish (AGFD 2013a, pp. 56-57; NMDGF 2013, p. 17). In Arizona, minnow traps used for collecting live baitfish must be checked once daily and the trapping activity must occur where captured bait will be used (AGFD 2013a, pp. 56-57); in New Mexico, there is no stipulation on time intervals in the regulations to check minnow traps (NMDGF 2013, p. 17). In either scenario in either state, these minnow traps are likely to be fully submerged when in use and pose a drowning hazard to resident gartersnakes while foraging underwater, as they can be lured into the traps by fish already caught.

We do not have adequate information to assess the frequency and geographical extent to which accidental drownings of gartersnakes in minnow traps may be occurring. This is mainly because it happens incidentally as a result of trapping efforts for other species, and so it historically did not get reported by researchers. Without additional information, we cannot conclude at this time that deaths from accidental minnow trapping are likely having population-level effects on either gartersnake. However, if even a few adult females are lost from populations that already have low densities and low rates of recruitment, these losses would contribute to population extirpations and the continued decline in the status of the gartersnakes. Working with researchers in the future to minimize the chances of snake drownings and to report any incidental collections of gartersnakes will be important for future conservation of both species.

Intentional Dewatering

—Lastly, dewatering or water fluctuation techniques are sometimes considered for eliminating undesirable fish species from water bodies (Finlayson

et al.

2010, p. 4). Dewatering of occupied northern Mexican or narrow-headed gartersnake habitat would have deleterious effects to affected populations by removing a primary habitat feature and eliminating the prey base. Because northern Mexican gartersnakes often occupy lentic water bodies or intermittently watered canyon bottoms, where this practice is most feasible, effects of dewatering activities may disproportionately affect that

speci

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