DRAFT , occidentalis occidentalis,

Revised RECOVERY PLAN

(Original Approval: March 15, 1984)

Prepared by

Francisco J. Abarca, Arizona Game and Department Phoenix, Arizona

Brian E. Bagley, Phoenix, Arizona

Dean A. Hendrickson, University of Texas Austin, Texas

and

Jeff K. Simms Bureau of Land Management Tucson, Arizona

for

Region 2 U.S. Fish and Wildlife Service Albuqerque, New Mexico

April 1993

Approved: Regional Director, U.S. Fish and Wildlife Service

Date: April-1993 Draft page ii

DISCLAIMER

Recovery plans delineate reasonable actions required to recover and protect the species. The U.S. Fish and Wildlife Service prepares the plans, sometimes with the assistance of recovery teams, contractors, State and Federal Agencies, and others. Objectives are attained and any necessary funds made available subject to budgetary and other constraints affecting the parties involved, as well as the need to address other priorities. Recovery plans do not necessarily represent the views nor official positions or approval of any persons or agencies involved in the plan formulation, other than the USFWS. They represent the official position of the USFWS only after they have been signed by the Regional Director or Director as approved. Approved recovery plans are subject to modification as dictated by new fmdings, changes in species status, and the completion of recovery tasks.

Literature citation should read as follows:

U.S. Fish and Wildlife Service. 199 . Gila topminnow, Poeciliopsis occidentalis occidentalis, Recovery Plan. U.S. Fish and_ Wildlife Service, Region 2, Albuquerque, New Mexico. pages.

Additional copies may be purchased from:

Fish and Wildlife Reference Service 5430 Grosvenor Lane, Suite 110 Bethesda, Maryland 20814 301/492-6403 or 1-800-582-3421

The fee for documents varies depending on the number of pages. April-1993 Draft page iii

ACKNOWLEDGEMENTS

Preparation of the Gila topmhulow Recovery Plan benefited from review and comments by the members of the Desert Recovery Team, and the following individuals:

Many other persons reviewed and provided information for the preparation of this plan. The USFWS appreciates the help provided by team members (past and present), consultants, and other individuals who contributed to the preparations of this document. April-1993 Draft page iv

EXECUTIVE SUMMARY

Current Species Status: The Sonora topminnow, Poeciliopsis occidentalis, is comprised of two subspecies, the Gila topminnow, Poeciliopsis o. occidentalis, and the Yaqui topminnow, Poeciliopsis o. sonoriensis. Both subspecies were listed as endangered in 1967 with no critical habitat designation. The original recovery plan for the Sonora topminnow was approved on March 15, 1984; this is a revision of that plan, but only includes the Gila topminnow. The Yaqui topminnow Recovery Plan is currently being prepared under a separate document.

The Gila topminnow is restricted to only 9 natural populations in Arizona. In the United States, the species occurs in the Gila River drainage, Arizona, particularly in the upper reaches of the Santa Cruz River, Sonoita and Cienega creeks, and middle Gila River. In Mexico, the species occupies the upper Rio Sonora, and Rio de la Concepci6n, Sonora. It is also believed that the Gila topminnow may occur in the Mexican portion of the upper Santa Cruz and San Pedro rivers, Sonora.

Habitat Requirements and Limiting Factors: The fish inhabit a variety of water types: springs, cienegas, permanent and intermittent streams, and margins of large rivers. Habitat alteration and destruction, and introduction of nonnative fish, principally , Gambusia affinis, that prey on or compete with the species, are the main reason for the decline of the Gila topminnow.

Recovery Objectives: Delisting of the subspecies is not considered feasible in the foreseeable future. Downlisting of the Gila topminnow in the United States may be possible.

Recovery Criteria: Protect and monitor existing natural populations. Protect habitat and remove nonnative species from it. Replicate natural populations within historic range. Restore genetic diversity to historic/natural levels. Replicated population will not be considered established until they have persisted a minimum of 10 years.

Actions Needed: 1. Protect remaining populations and their habitats. 2. Re-establish populations throughout historic range. 3. Monitor natural and reintroduced populations and their habitats. 4. Develop and implement genetic protocol. 5. Maintain captive populations. 6. Study life-history and interactions with nonnative fishes. 7. Inform and educate the public and resource managers. April-1993 Draft page v

Costs ($0001: Year Need 1 Need 2 Need 3 Need 4 Nedd $ Need 6 Need 7 1 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 2 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 3 XX.0 XX.0 XX.0 XX.0 XX.0 30C.0 XX.0 4 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 5 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 6-10 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0

Total XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 XX.0 Cost April-1993 Draft page vi

TABLE OF CONTENTS

DISCLAIMER ...... jj

ACKNOWLEDGEMENTS ...... LII

EXECUTIVE SUMMARY ...... iv

I. INTRODUCTION ...... 1 Description ...... 1 Historic and present distribution ...... 2 Ecology and life history ...... 5 Reasons for decline ...... 9 Conservation measures ...... 10

II. RECOVERY ...... 12 Objective ...... 12 Narrative Outline ...... 14 Literature Cited ...... 30

III. IMPLEMENTATION SCHEDULE ...... 38 Definition of Priorities ...... 38 Abbreviations Used ...... 38

IV. APPENDIX Table 1. Historic records of Gila topminnow ...... 42 Table 2. Status of natural populations in Arizona...... 46 Table 3. Summary of succesful introduced populations ...... 47 Table 4. Summary of all known reintroductions of Gila topminnow in Arizona 49

V. COMMENTS RECEIVED ...... 55 April-1993 Draft page 1

I - INTRODUCTION

The Gila topminnow, Poeciliopsis occidentalis occidentalis, is one of two subspecies of Sonoran topminnows found in Arizona and northern Mexico. Endemic to the Gila River Basin of New Mexico, Arizona, and northern Sonora Mexico, Gila topminnows were once considered among the most abundant fishes in the lower Colorado River watershed (Hubbs and Miller 1941). However, habitat loss and predation by exotic fishes, particularly mosquitofish, Gambusia affinis, caused large scale reductions in its distribution within the United States and its subsequent listing as an endangered species (USDI 1967, Schoenherr 1974, Hendrickson et al. 1981, Meffe et al. 1983, Meffe 1985, Minckley 1985). Gila topminnows are still widespread in Sonora Mexico (Vrijenhoek et al. 1985, Varela-Romero et al. 1990a 1990b, Minckley et al. 1991), however, increases in exotic fishes and human development may impact the species in Mexico as well (Hendrickson 1983, Meffe and Vrijenhoek 1988, Gomez-Alvarez et al. 1990). The Gila topminnow is listed as threatened by the State of Arizona (Arizona Game and Fish Department [AGFD] 1988).

Since being federally listed in 1967, Gila topminnows have been reintroduced into more habitats than any native fish in the southwest (Hendrickson and Brooks 1991). Gila topminnow populations, native as well as reintroduced, continue to decline. This recovery plan details the Gila topminnow recovery effort, acquaints the reader with the species, its status, the threats it faces, and provides a plan for its recovery.

Recovery efforts for Gila and Yaqui topminnows have been incorporated into a previous recovery plan (USFWS 1984). However, since recovery actions for Yaqui topminnows parallel actions required for other listed species from the Rio Yaqui drainage, Yaqui topminnow recovery is included in a separate recovery plan for the endangered and threatened fishes of the Rio Yaqui.

Description

Gila topminnows are small live-bearers of the family which comprises 18 known species of the genus Poeciliopsis (Wetherington et al. 1989, Morizot et al. 1990). Males seldom exceed 25 millimeters (mm) standard length and females average 30 to 45 mm. They are tan to olive bodied and usually white on the belly. The scales of the dorsum are darkly outlined and the fm rays are outlined with melanophores, although lacking in dark spots. Breeding males are blackened, with some gold on the pre-dorsal midline, orange at the base of the gonopodium, and bright yellow pelvic, pectoral, and caudal fins (Minckley 1973).

The species was originally described in 1853 (Baird and Girard) from a specimen collected in 1851 from the Santa Cruz River near Tucson. It was named Heterandria occidentalis, but was redescribed in 1941 by Hubbs and Miller as P. occidentalis. Both forms of Sonoran topminnows were recognized as separate subspecies by Minckley (1969a), who gave their distinguishing April-1993 Draft page 2 features.

The two subspecies can be distinguished by several characteristics. In P. o. occidentalis the snout is short, the mouth subsuperior and the dark lateral band of the female extends from the opercle to the base of the caudal fm. In P. o. sonoriensis the snout is longer, the mouth superior and the lateral band of the female rarely begins before the base of the pelvic fins (Minckley 1973). Recent molecular genetic data found greater mitochondrial DNA diversity in Yaqui topminnows than Gila topminnows which suggests they may more appropriately be considered separate species (Quattro et al. 1992). For the purposes of this plan we will continue to refer to them as subspecies.

P. occidentalis is the only member of the family Poeciliidae that is native to the Gila River drainage. Other members of the family, including mosquitofish, Gambusia affinis; guppies, Poecilia reticulata; sailfm mollies, P. latipinna; Mexican mollies, P. mexicana; green swordtail, Xiphophorus helleri; and variable platyfish, X. variatus, have been purposefully introduced into waters within the Gila River drainage as vector controls, or accidentally through the tropical fish trade.

Mosquitofish can be distinguished from Gila topminnows by the presence of a dark, sub-orbital bar (tear drop) and black spots on the dorsal and caudal fin of mosquitofish. Mosquitofish males do not turn black like breeding male topminnows. The gonopodium is relatively longer in topminnows, reaching beyond the snout when in the copulatory position, whereas in mosquitofish it does not reach past the tip of the snout (Minckley 1973).

In Mexico Gila topminnows hybridize with an all-female form, - occidentalis (Moore et al. 1970, Moore and McKay 1971, Lanza 1983. Although male Gila topminnows have a genetic contribution in the first generation, the paternal genome is discarded during oogenesis and is replaced by mating with another male. This process, known as "hybridogenesis", produces offspring with only the maternal P. monacha genome being passed down (Angus 1980, Schenck and Vrijenhoek 1986, Morizot et al. 1990).

The Gila topminnow has a recovery priority of 2, which means it has a high degree of threats to its existence, but has a high probability for recovery.

Historic and Present Distribution

Gila topminnows represent the northernmost extension of their tropical genus, which extends from the north Andes in Colombia, along the Pacific coast of Central America and Mexico to the Gila. Two members of the Genus also occur in some Atlantic streams of southern Mexico, Guatemala and Honduras (Rosen and Bailey 1963). Gila topminnows were historically widespread in the Gila River drainage below 1800 meters (M) elevation (Minckley et al. 1977)(Table 1). They were found from the San Francisco River at Frisco Hot Springs, New Mexico west to the mainstem Gila River near Yuma, Arizona, and possibly even into the lower Colorado River itself (Minckley and Deacon 1968). They thrived in the Salt River as far April-1993 Draft page 3 upstream as the present site of Roosevelt Lake. Although there are no museum specimens from the Verde or San Simon Rivers, it is likely that they did occur there. Two collections were made from the San Pedro. In 1943 J. R. Simon collected topminnows near Feldman, Arizona and in 1978 a population was discovered in a spring 13 kilometers (km) southeast of Mammoth (McNatt 1979). Records of topminnows from the Santa Cruz system are abundant and include the headwater area above Lochiel, Arizona (Minckley et al. 1977); that part of the river that flows through Sonora, Mexico, before returning to the United States; and various tributary streams and springs, most notably Sonoita Creek. Gila topminnows were found throughout the Rios de la Concepcion and Sonora in northern Sonora, Mexico (Vrijenhoek et al. 1985, Hendrickson and Juarez-Romero 1990, Minckley et al. 1991).

Historically, Gila topminnow must have formed an almost continuous population at low elevations throughout the Gila. During times of environmental extremes, such as droughts and floods, it disappeared from marginal habitats only to reinvade as conditions improved. During wet weather years populations would disperse and this presumably lead to widespread contact between geographically separated populations (Deacon and Minckley 1991).

Gila topminnows are still extant in several of the above localities (Figure 1, Table 2). The most recent status report for populations in the United States (Brown and Abarca 1992) lists eight remaining native populations of Gila topminnows in the United States. An additional natural population was discovered in 1992 (David Weedman per. comm.). However, two native populations believed to be extirpated, North Fork of Ash Creek and Sheehy Spring, may still support topminnows. Thus, there are 9 native populations of Gila topminnows in the United States. In 1984, the original recovery plan for Gila topminnows listed 11 extant native populations (USFWS 1984). However, since then an additional population was discovered, North Fork of Ash Creek (Jennings 1987), and two populations were lost, Salt Creek and Middle Spring (Marsh and Minckley 1990). These populations were eliminated by mosquitofish and lack of water, respectively.

Gila topminnows are still widespread throughout northern Sonora Mexico in the Rios de la Concepcion and Sonora (Minckley et al. 1991). These drainages also contain the unisexual hybrid P. monacha-occidentalis (Schultz 1960, Angus and Schultz 1979, Hendrickson et al. 1981, Schultz 1989, Campoy-Favela et al. 1989, Hendrickson and Juarez-Romero 1990). In the Rio de la Concepcion the unisexual hybrid comprises 0-3% of all Poeciliids (Moore et al. 1970). It is believed that populations of Gila topminnows may also be in the Mexican portions of the San Pedro and Santa Cruz rivers.

Seven of the remaining native populations of Gila topminnows are in the Santa Cruz River system: Redrock Canyon, Cottonwood Spring, Monkey Spring, Sonoita Creek, Cienega Creek, Sharp Spring, Sheehy Spring and the upper Santa Cruz River. Bylas Springs and the North Fork of Ash Creek populations are located on the Gila and San Carlos Rivers, respectively, on the San Carlos Apache Indian Reservation. Bylas Springs, S-I (Marsh and Minckley 1990), has been unsuccessfully poisoned for mosquitofish twice (Meffe 1983, Brooks 1985, Marsh and Minckley 1990). The North Fork of Ash Creek population has been severely reduced or eliminated by April-1993 Dre0 page 4

Figure 1. Natural Gila topminnow populations in the United States. Solid circles indicate extant populations. Open circles are extirpated populations. April-1993 Draft page 5 mosquitofish. Monkey Spring, Cottonwood Spring and Cienega Creek are the only natural populations not contaminated with mosquitofish.

As part of the recovery actions, over 208 Gila topminnow reintroductions have taken place at 178 wild locations (Figure 2). For the purpose of this count a wild location refers to a habitat which does not have a mailing address. As of 1992, 20 wild populations remain, 19 of which are in historic range and count towards recovery (Brown and Abarca 1992, Figure 2, Table 3). Recent surveys indicate loss of one population, Sycamore Spring, and reappearance of another, Cave Creek. According to Brown and Abarca (1992), thirteen of these populations are secure enough that they should persist into the foreseeable future.

Gila topminnows have also been stocked into more than 100 hundred captive locations for the purpose of propagation and conservation (Table 4). Eleven captive populations persisted in 1992. The following publicly maintained populations are large enough to provide individuals for introductions: Arizona-Sonora Desert Museum, Boyce-Thompson Arboretum, Dexter National Fish Hatchery, and Roper Lake State Park. The Boyce-Thompson Arboretum population is not recommended for use in reintroductions since it is derived from individuals of several different native populations and is thus of unknown genetic integrity (See Conservation Measures in this plan). Simons (1987), Bagley et al. (1991) and Brown and Abarca (1992) describe, in detail, the plight of captive and reintroduced populations of Gila topminnows.

Ecology and Life History

A survey of genetic diversity of four native Gila topminnow populations in Arizona revealed that three populations (Monkey and Bylas Springs and Cienega Creek) were homozygous for all 25 loci tested. The fourth population, Sharp Spring, had a mean heterozygosity of 3.7% (Vrijenhoek et al. 1985). Quattro and Vrijenhoek (1989) compared life history characteristics of Gila topminnows from Monkey and Sharp Springs. It was suggested by the authors that the greater survival, growth, fecundity, and developmental stability of topminnows from Sharp Spring may be attributed to their genetic diversity.

Habitat

Habitat requirements of P. occidentalis are fairly broad; it prefers shallow, warm, fairly quiet waters, but can adjust to a rather wide range, living in quiet to moderate currents, depths up to 2.0 M (Meffe et al. 1982) and temperatures from near freezing under ice to 37°C (Heath 1962, Meffe et al. 1983). Topminnows can live in a fairly wide range of water chemistries, with recorded pH's in existing habitats from 6.6 to 8.9, dissolved oxygen readings from 2.2 to 11 milligrams/liter (mg/1) (Meffe et al. 1983), and salinities from tap water to sea water (Schoenherr 1974). Since topminnows have these broad tolerances, they can live in a wide variety of water types; springs, cienegas, marshes, permanent streams, intermittent streams, and formerly along the edges of large rivers. Reintroduced populations, such as Mud Spring, have been known to survive for over 11 years in habitats as simple as cement watering troughs. Historically, their distribution would expand during wet years, while retreating to springs and April-1993 Draft page 7 other permanent water sources during periods of drought. It has been reported by Meffe et al. (1983) that topminnows can also tolerate almost total loss of water by burrowing into the mud for 1-2 days. Preferred habitat contains dense mats of algae and debris, usually along stream margins or below riffles, with sandy substrates sometimes covered with organic muds and debris (Minckley 1973). Topminnows are usually found in the upper 1/3 of the water column and young show a preference for the warmest and shallowest areas (Forrest 1992). Simms and Simms (1992) noted topminnows occupying pools, glides and backwaters more frequently than marshes or areas of fast flow. They also recorded topminnow densities in Cienega Creek as high as 566 topminnows/M.

Mosquitofish can also tolerate similar environmental extremes and occupy similar habitats (Meffe et al. 1983). However, topminnows are more tolerant of moderate floods. Meffe (1984) noted that flooding events removed a larger percentage of mosquitofish. In Sharp Spring, prior to moderate flooding, mosquitofish comprised 11.5% of the fish fauna. After flooding mosquitofish comprised 0.7% of the fish fauna. Controlled experiments using artificial streams showed that as flow increased, topminnows oriented to the flow and moved to the edge where current was reduced. In contrast, mosquitofish tried to maintain their midchannel position and were swept downstream. In areas that are not prone to flooding, coexistence rarely exceeds 3 years. However, in habitats that do flood, such as the Santa Cruz River, topminnow have survived in the presence of mosquitofish for over 29 years.

Not all flooding is beneficial for Gila topminnows. Extreme flooding has removed several reintroduced populations; Camp and Cave Creeks (Minckley 1969a), Tule Creek (Collins et al. 1981), and Seven Springs (USFWS 1984).

Reproduction

Topminnows have internal fertilization, whereby spermatophores (packets of sperm) are transferred from the male's gonopodium to the females genital pore. Males stop growing once they are sexually mature Large males (14-20 mm standard length) turn black, defend territories and court females. Small males do not turn black or defend territories. Instead they take on a "sneaking" mating strategy where they attempt to mate with uncooperative females. Subordinate males have a relatively longer gonopodium which may have an adaptive benefit for its mating strategy (Constantz 1989). However, if the larger, territorial males are removed, smaller males will move in, take on breeding coloration, and defend territories (Constantz 1975, Schoenherr 1977).

Sexual maturity occurs from 2-11 months, depending on the season of birth (Constantz 1976, 1979; Schoenherr 1974). Breeding occurs primarily during January through August, but in thermally constant springs young are produced throughout the year (Heath 1962, Minckley 1973, Schoenherr 1974). During the peak of the breeding season, up to 98% of the mature females are pregnant (Minckley 1973). Females usually have two and sometimes three broods developing simultaneously (Schoenherr 1974, 1977; Constantz 1980). Brood size ranges from 1-90, with and average of 5-15 (Constantz 1974; Schoenherr 1974, 1977). Brood interval is 24-28 days April-1993 Draft page 6

Figure 2. Reintroductions of Gila toprninnows into wild habitats within the United States. Solid circles indicate extant populations. Open circles are extirpated populations. April-1993 Draft page 8

(Minckley 1973). Females average 2-5 broods per year (Constantz 1974, Schoenherr 1974). Females can store sperm in the folds lining the ovary and gonoduct which may be used to fertilize future broods (Constantz 1974, 1989; Schoenherr 1974).

Mosquitofish generally have fewer broods, with a brood size range of 1-315 young. Thus, they have a greater reproductive potential (Moyle 1976).

Brood size and the onset of breeding is influenced by food abundance, photoperiod, temperature, predation and female size. Increased food supply and female size are believed to contribute to the 440% greater fecundity seen in topminnows from Monkey Spring canal compared to topminnows from Monkey Spring headspring (Constantz 1974, 1979; Schoenherr 1974, 1977).

Gila topminnows support an all female hybrid P. monacha-occidentalis throughout its range except in the Gila. Since territorial male topminnows have been shown to prefer to mate with conspecifics, it appears that subordinate males are responsible for proliferation of the hybrid form (Moore et al. 1970, Schoenherr 1974, Vrijenhoek et al. 1977, Keegan-Rogers and Schultz 1988, Schultz 1989).

Growth

Growth rate of Gila topminnows is dependant on age, sexual maturity, habitat and available resources (Constantz 1974, Schoenherr 1974). Males stop growing once they reach sexual maturity. Females continue to grow throughout their lives. Males rarely exceed 30 mm SL while females can attain 50 mm. Females usually outlive males, which can live over one year (Schoenherr 1974).

Mosquitofish, although smaller at birth, have a faster growth rate than Gila topminnows. Female mosquitofish over 50 mm standard length are not uncommon. Males do not grow as large as Gila topminnow males.

Diet

Gila topminnow are opportunistic feeders. They have an omnivorous dentition. Primary food items include detritus, vegetation, amphipods, insect larvae, and very rarely fish (Minckley 1973; Constantz 1976, 1980; Gerldng and Plantz 1980; Meffe et al. 1983; Meffe 1984). Gerking and Plantz (1980) noted that Gila topminnows prefer larger prey than available if choosing randomly but, prey sizes are limited by mouth size. This prey selectivity may allow Gila topminnows to invest more energy into reproduction.

In contrast, mosquitofish have a carnivorous dentition, possessing strong conical teeth and a short gut. They feed primarily on rotifers, snails, spiders, insect larvae, , algae, detritus, and fish fry, including cannibalism (Minckley 1973, Meffe and Crump 1987). Since there is little diet overlap between mosquitofish and topminnows, competition for food is not considered a factor in the decline of Gila topminnows (Schoenherr 1974, 1981). April-1993 Draft page 9

Reasons for Decline

Habitat destruction and introduction of exotic species have caused severe reductions of Gila topminnow populations, and are the main reasons for its listing as an endangered species (USFWS 1984; Williams et al. 1985, 1989; Simons et al. 1989). These two variables are involved in the reduction of 98% of the North American fishes listed as endangered, threatened or of special concern (Miller 1972, Deacon 1979, Deacon et al. 1979, Ono et al. 1983, Williams et al. 1989, Williams and Miller 1990).

During the late 1800s and early 1900s, several factors caused widespread habitat changes throughout the southwest. Heavy overgrazing and wood cutting combined with a drought during 1891-1893 caused widespread loss of vegetation and a 50-75% loss of cattle (Hastings and Turner 1965, Deacon and Minckley 1974, Hendrickson and Kubly 1984). This lack of vegetation made the area vulnerable to erosion when the drought ended. Floods, unbuffered by vegetation, scoured watercourses, deeply incised marshy cienega habitats, lowered the water table, desiccated watersheds and turned permanent flowing waters into occasionally flooding arroyos. Marshes dried, springs failed and streamside backwater and inlets disappeared (Miller 1961, Rea 1983, Fradkin 1984, Hendrickson and Minckley 1985, Bahre 1991). In only 10 years the San Pedro River was "incised from its mouth for 125 miles upstream" (Bryan 1925). During this time the water pump came into use and caused additional lowering of the water table (Rogers 1980). Habitats were further impacted by construction of water diversions and dams which created artificial habitats favoring exotic fish species (Minckley et al. 1991).

Large scale reductions of the Gila topminnow corresponds strongly with the spread of mosquitofish which were first collected from Arizona in 1926 (Miller and Lowe 1964). Mosquitofish, like many fish, have an alarm reaction to chemicals released from the injured skin of conspecifics; Gila topminnow do not (Vives 1991). Gila topminnows evolved with a naturally depauperate fish fauna lacking many predators. The predators which were present, Colorado River squawfish, Piychocheilus lucius, and species of the Genus Gila, occupied different habitats and had little impact on Gila topminnows (Miller 1961, Minckley et al. 1991).

In contrast, the predacious mosquitofish has very similar habitat preferences (Meffe et al. 1983). Mosquitofish prey directly on young topminnows and cause early death of adults due to infection of shredded fins (Schoenherr 1974, Meffe 1985). Elimination of topminnows by mosquitofish can occur rapidly: <2 years for elimination of a reintroduced topminnow population in Aravaca Creek (Miller 1961), 3 years or less to elimination a native population from artesian ponds near Safford (Minckley and Deacon 1968). Schoenherr (1974, 1981); Minckley et al. (1977) and Meffe (1984) reported on over 20 populations which were severely reduced or eliminated by mosquitofish in less than 3 years. Long-term coexistence appears to be related to the frequency of flooding which removes a larger percentage of mosquitofish (Meffe 1984, Minckley and Meffe 1987). Since mosquitofish have attained a cosmopolitan distribution, it is unlikely that this threat can be removed throughout the Gila topminnow's historic range. April-1993 Draft page 10 Conservation Measures

In September of 1981 a memorandum of understanding between the U.S. Fish and Wildlife Service, the U.S. Forest Service, and the Arizona State Game and Fish Commission provided a catalyst for large-scale introductions of topminnows. Of the 178 reintroductions into wild habitats, more than 132 wild locations have been stocked in Arizona and New Mexico since 1981 (Brooks 1985, Bagley et al. 1991, Table 4). Fish were placed into habitats ranging from cement cattle troughs to ponds and streams. This introduction program has had limited success (Brooks 1985, 1986; Simons 1987; Bagley et al. 1991; Brown and Abarca 1992).

Currently, 19 reintroduced populations, within historic range, persist in the wild (Table 3). From 1987 through 1989, the downlisting criteria outlined in the 1984 Sonoran Topminnow Recovery Plan, 20 populations surviving in the wild for three years, was met. However, downlisting was delayed since many of the populations appeared tenuous. In 1991, the number of successful reintroduced populations fell below the 20 required for downlisting. Of the populations which have failed since 1985; 51% were lost to desiccation, 20% to flooding, 20% to unknown causes, 2% to mosquitofish and the remaining 7% to cattle overuse, dredging or low oxygen (Brown and Abarca 1992). Delisting criteria appearing in the 1984 Recovery Plan were set at 50 wild populations surviving three years, or thirty wild populations surviving five years. Requirements for natural populations were omitted after 1987 (USFWS 1984).

The majority of the reintroductions initiated since 1981 have used topminnows from Boyce- Thompson Arboretum. However, this introduced population is apparently made up of individuals from Bylas Springs, Cocio Wash and Monkey Spring (Bagley et al. 1991). Constantz (1979) noted life history differences between various topminnow populations. Since then, researchers have determined that there are genetic as well as fitness differences between some natural Gila topminnow populations (Vrijenhoek et al. 1985, Quattro and Vrijenhoek 1989). Based on these studies, it was suggested that topminnows at Sharp Spring were more fit than those at Monkey Spring. Thus, the Dexter National Fish Hatchery population of Gila topminnows from Monkey Spring fish was replaced with Gila topminnows from Sharp Spring. Roper Lake State Park supports a large population of topminnows derived from Middle Spring and the Arizona-Sonora Desert Museum is a refugium for topminnows from Monkey Spring. Recent reintroductions have utilized different natural populations in order to preserve the genetic variation which currently exists. Successful wild reintroductions represent three different native populations. Monkey Spring topminnows are found in Big Spring, Cold Springs and Mescal Warm Spring. Middle Spring fish are found at Salt Creek. Heron Spring contains fish from Sharp Spring. The remaining populations were created with fish from Boyce-Thompson Arboretum and may be of mixed origin.

Recovery has also taken the form of reclaiming historic habitats by removing exotic fish species (Meffe 1983). Physical and chemical renovations have taken place at Bylas Spring, Salt Creek, Hassayampa River Preserve, Roper Lake State Park, and Boyce-Thompson Arboretum. These efforts have had limited success (Meffe 1983, Bagley et. al. 1991). Renovations were temporarily successful at Bylas Spring, Salt Creek, Roper Lake State Park, and Boyce-Thompson April-1993 Draft page 11

Arboretum. However, all of them currently support topminnow populations coexisting with exotics.

Since the large introduction program of the early 1980s, the recovery effort has primarily focused on habitat management and population monitoring. In addition, management activities include maintenance of stocks at universities, museums, and hatcheries, and an ongoing program of basic research (Minckley et al. 1991). April-1993 Draft page 12

IL RECOVERY

Objective and Criteria

This plan describes specific recovery actions determined necessary to secure the continued existence of the species through the protection of current existing habitats, establishment of successful additional populations within historic range, and elimination of threats.

Delisting of the species is not considered feasible in the foreseeable future. Downlisting from endangered to threatened will only be achieved if recovery actions, as delineated below, prove to be successful. Reclassification criteria are preliminary and they need to be periodically revised and remain dynamic so as to continually reflect the most recent knowledge of topminnow genetics and habitat relationship in management actions.

An important aspect in the recovery of the Gila topminnow will be the restoration of its genetic diversity to historic/natural levels. This plan provides an approach to enhance and maintain the genetic integrity of the species, as well as delineates a protocol to allow exchange of genetic material among natural and re-established populations.

Downlisting of the Gila topminnow will be determined by the success of four major activities. First, our ability to secure natural populations and their habitats. Second, implementation of a large-scale reintroduction program based on genetic modelling and using experiments under laboratory conditionsm, semi-natural, and natural habitats. Third, our ability to effectively perform population and genetic monitoring of natural and reintroduced populations, and fourth, the willingness of private and public parties to stock a large number of populations.

Successful recovery of the Gila topminnow will require substantial efforts from the following agencies and organizations: U.S. Fish and Wildlife Service, Region 1; U.S. Forest Service, Region 3; National Park Service; Bureau of Land Management; Arizona Game and Fish Department; Arizona State Land Department; New Mexico Department of Game and Fish; The Nature Conservancy; Rail-X Ranch, Arizona; San Rafael Cattle Company, Arizona; San Carlos Indian Reservation; The Nature Conservancy; Centro Ecologic° de Sonora, Mexico; Secretarfa de Recursos Hidratilicos (SARH), Mexico; and Secretarfa de Desarrollo Social (SEDESOL), Mexico. APRIL-1993 DRAFT PAGE 13

DOWNLISTING CRITERIA

Gila topminnow, Poeciliopsis o. occidentalis, will be considered for downlisting when:

1. Remaining natural populations and their habitats are secured. These include four metapopulations at nine locations: a) SAN RAFAEL VALLEY (Sharp Spring + uppermost Rio Santa Cruz); b) SONOITA CREEK (Sonoita Creek + Redrock Canyon + Fresno Canyon + Monkey Spring + Cottonwood Spring) C) CIENEGA CREEK (single continuous population on BLM property) d) BYLAS SPRINGS COMPLEX (Medicine/Boss springs); and

2. Should jw6'additional metapopulations (SAN CARLOS RIVER DRAINAGE [Ash Creek], RIO SAN PEDRO [headwaters]) and two additional population (Sheehy Spring [SAN RAFAEL VALLEY], Middle Spring [B'YLAS SPRINGS COMPLEX]) prove to be extant, these natural populations and habitats are secured as well. )

3. Natural populations are replicated and secured within historic range;

4. Genetic diversity is improved and maintained through a genetic protocol, which allows exchange of genetic material among populations, as delineated below;

5. Plans for monitoring of populations and their habitats, and periodic assessment of genetic integrity are developed and implemented.

A metapopulation is herein defined as all individuals naturally occurring in a single sub-basin (e.g., Sonoita Creek, Cienega Creek, etc.) with some probability of gene flow among them, but isolated from other gene pools (other subasins). a- AAA 044 ' ILEALFRIL- A secured population is herein defined as that located in an area under the control of an agency or organization mandated to or dedicated to legal protection against detrimental land and water practices which may threaten the continued existence of Gila topminnow.

A viable population is herein defined as that containing at least 500 overwintering adults or existing numbers, whichever is greater, and possesses an adequate representation of all age classes and evidence of reproductive activity. April-1993 Draft page 14

Step-down Outline

1. Protect remaining populations and their habitats.

1.1 Identify extent of geographic distribution of natural populations. 1.2 Acquire habitats currently occupied by remnant populations. 1.3 Protect natural populations from invasion by nonnative fishes. 1.4 Protect habitats currently occupied by natural populations from detrimental land and water use practices.

2. Re-establish populations throughout historic range.

3. Monitor natural and reintroduced populations and their habitats.

3.1 Develop standardized monitoring protocol. 3.2 Determine criteria for extirpated population.

4. Develop and implement genetic protocol.

5. Maintain captive populations.

6. Study life-history and interactions with nonnatives.

7. Inform and educate the public and resource managers.

Narrative Outline

TASK 1. PROTECT REMAINING POPULATIONS AND THEIR HABITATS.

Before the introduction of mosquitofish in the 1920s (Hubbs and Miller 1941, Miller 1961), the Gila topminnow was considered one of the most common fishes in the southern part of the Colorado River Basin. At present, only nine naturally occurring populations exist. These populations should receive the highest priority for protection since they represent the only original genetic material for the survival of the species. Currently, natural populations occupy headwaters and middle reaches of relatively small basins within a mosaic of private, State and Federal owned land. A thorough description of monitoring and management actions for natural populations is summarized by Brooks (1985 1986), Marsh and Minckley (1990), Bagley et al. (1991), and Brown and Abarca (1992). April-1993 Draft page 15

1.1 Identify extent of geographic distribution of natural populations.

Several natural populations were discovered in the late 1970s. Since then, annual monitoring of these populations has been concentrated to single locations, or reduced number of river miles, but no serious attempts have been made to determine the extent of their geographic distribution. Once this task has been accomplished, land ownership identification and habitat assessment should follow to determine protective measures. These measures should be incorporated into individual management plans for each site.

Similarly, there is an urgent need to survey portions of the Gila River basin for undiscovered populations of topminnow. Because this species had a wide distribution and has the ability to exist in a variety of habitat types including small springs and seeps, there is a high probability that some undiscovered small remnant populations still persist. In addition, some known topminnow populations have declined beyond detection, but may obtain detectable population levels again in the near future. The two drainages of greatest concern are the Rio San Pedro, Arizona and Mexico, and the San Carlos drainage, Arizona. Populations located at Middle Spring near Bylas, Arizona and Sheehy Spring of the San Rafael Valley may still be extant.

It is of primary importance to locate and verify the existence of any extant populations in the above locations or elsewhere in the Gila Basin.

Rediscovered or new populations and habitats should be secured, replicated, and monitored in the same manner as extant remnant populations already discovered.

1.2 Acquire habitats currently occupied by remnant populations.

Identify land ownership of habitat critical for the survival of remnant natural populations. Agencies and organizations which can supply legal protection from adverse land and water management practices need to acquire adequate amounts of land including water rights necessary to maintain and control habitat integrity for the near and distant future. In cases where a land owner is reluctant to sell, attempts will be made to purchase conservation easement for proactive management activities that favor topminnow habitat security.

Five of the nine existing topminnow locations are within private lands. Since the early 1980s, private land owners have been extremely cooperative by allowing continuous monitoring of those locations. Appropriate mechanisms must be used to protect these populations. A 25 yr cooperative agreement with the land owner should be pursued for monitoring, habitat enhancement and protection, eradication of nonnatives, and relocation of fishes if necessary. Currently, the AGFD is initiating conversations with Mr. Robert Sharp of the San Rafael Cattle Company for habitat improvement of Sheehy and Sharp springs, and the upper Santa Cruz River populations. April-1993 Draft page 16

Topminnow habitat in Mexico (Santa Cruz, San Pedro, de la Concepcion, and Sonora rivers) is under State, Federal, private and ejido ownership. Ejido is a communal farm in Mexico. Formal protection of land and water should be pursued similar to that in the United States.

1.3 Protect natural populations from invasion by nonnative fishes.

Where possible, removal of nonnatives should be attempted (Bylas, Sharp and Sheehy springs and Redrock Canyon) following erection of appropriate barriers to reinvasion. In those sites where exotics have not yet invaded (e.g. Cottonwood Spring), improved bathers to invasion should be erected. Periodical thorough surveys of upstream habitats must be conducted in selected areas (Mexican portion of San Pedro, San Carlos Reservation) to locate and remove nonnatives.

Mosquitofish still remain a primary threat to the continued existence of the Gila Topminnow. Several factors should be taken into account when prioritizing habitats for renovation including population origin (natural vs. reintroduced), immediacy of threat, status of replicate populations of the same lineage, and probability of both short-term and long-term success. Some factors affecting success include poorly organized and executed renovation, potential recontamination by public or nearby populations on the watershed, habitat complexity and size, and lack of barriers to fish migration (Marsh and Minckley 1990, Rinne and Turner 1991).

Bullfrogs may also prey heavily on topminnow. Like mosquitofish, bullfrogs may need to be controlled in order to provide adequate security for affected topminnow populations.

Topminnow habitat(s) at risk of contamination or recontamination with nonnative fishes will require preventative measures. Some other measures that may reduce the risk of contamination include inventory of nonnative fish distribution on watershed, reclamation of all sources of contamination with potential for dispersal, for example, by immigration during flood (close or distant proximity) or human transport (close proximity).

1.4 Protect habitats currently occupied by natural populations from detrimental land and water use practices.

Once sufficient land and water or conservation easements have been acquired, several tasks must be accomplished before topminnow populations can be considered secure. These include special designation which confer regulations that will provide long term protection and management (e.g. Area of Critical Environmental Concern, Research Natural Area, Wild and Scenic River, etc.), assurance of water quality and quantity, protection against habitat degradation, control or removal of detrimental nonnative plants and , prevention of invasion by nonnative fishes, and modification of land management practices either directly or indirectly detrimental to aquatic habitats. April-1993 Draft page 17 Management plans that cover single or multiple populations must be drafted and properly implemented before a topminnow population(s) will be considered secure. Impacts of activities such as livestock grazing or watering, mining, timber harvest, phreatophyte control, recreation, agricultural or residential development, etc., must be assessed and factored into each plan. Aquatic vegetation generally adds to habitat diversity. However, dense growths not checked by occasional disturbance (e.g. floods, herbivorous animals, etc.) can crowd surface water to the point that topminnow carrying capacity is severely diminished (e.g, Bylas Springs I and II). Habitat features need to be monitored in order to recognize and avoid such subtle shifts in habitat quality. Following identification of vegetative overgrowth problem, manipulation of vegetation may be required to enhance habitat features for Gila topminnow survival.

1.5 Incorporate Gila topminnow recovery plan into site specific management plans.

Management plans that cover single or multiple populations must be drafted and properly implemented before a topminnow population(s) will be considered secure. Cooperative planning that involves all major entities within the watershed where a natural population(s) occurs or where recovery related activities are needed should be established when possible. Pertinent tenets of this recovery plan need to be incorporated into management decisions as they develop. Government (federal, state, local) and private entities should be encouraged to participate in such "ecosystem level" planning. This planning and subsequent full implementation of such plans is crucial to long-term persistence of Gila topminnow. This is especially true for natural populations affected by several resource management entities. Impacts of activities such as livestock grazing or watering, mining, timber harvest, phreatophyte control, control, recreation, agricultural or residential development, etc., must be assessed and factored into each plan.

TASK 2. RE-ESTABLISH POPULATIONS THROUGHOUT HISTORIC RANGE.

Stocking of topminnow started in 1936 (Minckley 1969c) and was intensified in 1982 thanks to a Memorandum of Understanding (MOU) between the USFWS, USFS, and AGFD. Since then, one of the most aggressive reintroduction efforts for an endangered species, has been implemented, with more than 340 documented stockings of Gila topminnow (Minckley and Brooks 1985, Brooks 1986, Simmons 1987, Bagley et al. 1991, Brown and Abarca 1992). Among short-lived fishes in North American deserts, no other fish has been transplanted so many times as the Gila topminnow (Hendrickson and Brooks 1991). Prior to 1982, Gila topminnow were stocked into more than 56 wild sites (Minckley and Brooks 1985). In 1982, 90 wild sites were stocked, followed by 24 in 1983 (Brooks 1985; Appendix A). An additional 22 wild sites have been stocked or populated by dispersal from stocked populations since 1983. Thus, Gila topminnows have been introduced into a total of 178 wild habitats (Simmons 1987, Bagley et al. 1991, Brown and Abarca 1992). Introductions have also occurred into more than April-1993 Draft page 18 102 captive sites. Despite this large-scale reintroduction effort, successful established population percentage remains low ( < 15 %) (Bagley et al. 1991, Brown and Abarca 1992). Reasons for failure are attributable to negative impacts by mosquitofish, desiccation, summer flood, low levels of oxygen, and habitat destruction by cattle, but declines might also be related to the overall fitness of the introduced lineage. Bagley et al. (1991) indicated that several introductions received Gila topmimiows from more than one population of that subspecies. As an example, Boyce-Thompson Arboretum received Gila toprninnows in 1971 from Page Springs Hatchery (Minckley and Brooks 1985). These fish originally came from Monkey Spring. However, around 1973 fish from Cocio Wash, a now extirpated natural population, were also stocked into the Arboretum (AGFD files). AGFD files also report Gila topminnows from Bylas Spring being stocked into the Arboretum prior to 1978. With some few exceptions, all of the introductions in 1982 and 1983 used fish from Boyce-Thompson Arboretum. These actions clearly call for a re-design of the reintroduction program.

In 1984, the original Recovery Plan indicated that downlisting of the Gila topminnow would occur when 20 populations had been successfully reestablished in the wild, and had survived for at least three years. Delisting of the subspecies, after 1987, would occur when 50 populations were successfully reestablished in the wild, within historic range, and had survived for at least three years, or 30 populations were successfully reestablished and had survived for at least five years. Just a year later (1985), criteria stipulated for downlisting were met (Brooks 1985 1986) and continued to exceed at least until 1987 (Simons 1987, Simons et al. 1989). Simons et al. (1989) and Hendrickson and Brooks (1991) discussed management implications of these actions and recommended comprehensive experiments with topminnow transplants.

Almost ten years after the original recovery plan was written, less than 15% of the reintroduced populations remain extant. It seems that a more innovative and aggressive reintroduction program should be implemented to assure the long-term survival of the Gila topminnow. This program should reflect current available genetic data which supports the notion that the Gila basin historically harbored what was a single, essentially basin-wide, panmictic population (Joe Quattro pers. comm.) (See also Task 4).

The new reintroduction protocol is based on two major considerations:

A) Manage reintroductions to assure that the program does not adversely impact any natural populations, and to assure no additional losses of genetic material in Arizona. The most secure way to achieve this is by assuring that each remaining native population is replicated in at least two separated refugia and through reintroductions to wild sites. Future reintroductions should then use refugia stocks derived from the hydrographically nearest natural population whenever stockings are made to waters from which they might readily escape and invade natural populations.

Populations of Gila topminnow are to be re-established according to the following specifications. April-1993 Draft page 19

Stock to be used I Geographic Area Cienega Creek Cienega Creek above interstate 10. Sharp Spring San Rafael Valley drainage. Redrock Canyon Redrock Canyon drainage. Cottonwood Spring or Remaining Santa Cruz drainage (excepting San Rafael Valley Sonoyta Creek and Cienega Creek drainage above interstate 10). Various remaining Bylas Gila River drainage above San Carlos reservoir (except San Springs stock(s) Simon drainage). Monkey Spring Because the stock is now widely distributed, it need not to be used in future stockings until other stocks are pqually secured, or until it becomes endangered once again through loss of currently extant reintroduced populations.

B) Manage the reintroduction program as a large-scale field experiment. Introductions to sites from which it is deemed highly unlikely that individuals might invade natural populations (i.e. entire San Simon basin; Cienega Creek and other Santa Cruz tributaries below interstate 10 and interstate 19 [Tucson-Nogales highway] including Avra-Altar drainage; Gila drainage below San Carlos dam, including middle and lower San Pedro, Salt/Verde, Agua Fria, Hassayampa, and lower Gila) should utilize all native Arizona stocks and mixes all of them in an experimental design which will contribute to testing of hypotheses of the relationship of heterozygosity to fitness. If impossible to test directly for correlations of genetic diversity and fitness, at a minimum this experiment could test for differences in fitness among pure native stocks and hybrids.

In addition, it is recommended that future reintroductions should consider the following guidelines:

C) Reintroduction sites which have maintained populations for extended periods and are thus of proven stability, should be given as much protection as possible, and should not receive new stockings unless future studies clearly demonstrate that such action would be advantageous.

D) Stockings of fish from the Rio de la Concepcion basin to Arizona habitats in that basin (Sycamore Creek drainage or natural or artificial sites in that drainage, e.g. on Buenos Aires National Wildlife Refuge) should be carried out to further protect those stocks against the threat of mosquitofish or other impacts in Mexico.

E) Reintroduce populations into a diversity of habitats (springheads, cienegas, streams, margins of rivers) if such habitats are available. These habitats should reflect, as much as possible, historic conditions prior to arroyo cutting. Large numbers ( 500) of topminnows should not April-1993 Draft page 20 be concentrated into a single habitat type, but rather be distributed among suitable habitats. Reintroduction of large numbers is extremely important, since small populations of small species, such as the topminnow, are more prone to extinction than are similar-sized populations of large (long-lived) species (Hendrickson and Brooks 1991). In addition, large population numbers may also prevent genetic bottlenecks which reduce genetic diversity (Echelle 1991).

F) No reintroductions should take place in the upper San Pedro, and San Carlos rivers and their tributaries, until it is proved that no natural populations occur.

G) Supplemental stockings in a single location must be evaluated in a case-by-case basis and should be carried out only if future studies clearly demonstrate that such action would be advantageous.

H) Considering present situation of stream and riparian habitats within the historic range of the Gila topminnow, this plan acknowledges that many reintroduction efforts will require habitat restoration and/or improvement of selected locations prior to actual fish stockings.

I) Classification system proposed by Simons (1987) for reintroduced populations should continue to be used. The system identifies the quality of a site, including its potential to persist in the future. This system is far better than simply counting the total number of successful sites and considering them all of equal quality (Brown and Abarca 1992). The categories were defined as follows:

Category 1 populations have moderate to high numbers of individuals in habitats with natural surface water, and are expected to persist indefinitely without anthropogenic support.

Category 2 populations have moderate to high numbers of individuals in habitats expected to persist indefinitely, but only with anthropogenic support.

Category 3 populations have such low numbers, or habitat of such limited quantity or quality, that populations are not expected to persist for many years.

Category 4 populations are threatened by mosquitofish (or other exotic fish), regardless of population size or habitat quality.

Efforts should be made to enhance habitat conditions and improve population numbers for categories 2 to 4, so they become category 1.

J) Control and documentation of stockings. To avoid duplication of efforts and records, the proposing agency should coordinate all activities with the AGFD, before and after stocking of fish. All stocking records must be stored at AGFD for proper distribution to pertinent agencies and individuals. April-1993 Draft page 21

K) Stocks of Gila topmimiow should be obtained from Dexter National Fish Hatchery or selected refugia populations as they become available. Direct use of wild populations for stockings should be discouraged since the mosquitofish has already invaded some natural populations and the probability of contamination might be high.

Detailed habitat assessment must be conducted prior to any reintroduction as recommended by Williams et al. (1988), and be sanctioned by the pertinent agencies. High quality sites for potential reintroduction will be those that include permanent water; absence of mosquitofish and other predatory species; adequate combination of run, pool and backwater areas (habitat complexity); preferably springheads with high carbonate output (presumably avoided by mosquitofish); and lack of detrimental human activities. Specifically, the following reintroduction site criteria (modified from Brooks 1985) are recommended:

Criterion Comments

Drainage area < 1.0 km' Elevation < 1,600m Stream flow Perennial, lotic, sheltered areas with <0.1 m3/sec Stream gradient < 3% Stream geomorphology Stream channel classification of C-2 through C-5 (Rosgen 1980) Pond surface area <2 ha Pond depth < 2 m Channelization Little or none Habitat composition Complex, heterogeneous, protected form major reoccurring flash flooding Cover Present. Moderate to abundant aquatic vegetation Other species Only native, nonpredatory fishes. A variety of insect life. Presence of Physella, Planorbella and or Hydrobiids Water quality ADHS Surface Water Quality Standards Developmental potential Low or none Location Gila River drainage (see 2.1 above) April-1993 Draft page 22 A proposed stocking does not necessarily have to meet all these criteria. They should not be interpreted as absolute values, but rather as a basic approach for proposed reintroduction sites. Further information of habitat preferences and quantitative analyses on failure/success of reintroduced populations should prompt revision of this protocol.

Efforts should be made to continuously survey for potential reintroduction sites within each subasin, and within those areas not currently occupied by topminnow throughout the Gila River system.

Discovery of new natural populations should prompt revision of these plans to accommodate reintroductions using them.

TASK 3. MONITORING NATURAL AND REINTRODUCED POPULATIONS AND THEIR. HABITATS.

3.1 Develop standardized monitoring protocol.

Success in meeting and measurement of progress toward goals and objectives of this recovery plan will depend on reliable data accumulated in a systematic way to assess population and habitat changes over time. Frequent monitoring of natural populations will allow early detection of destructive nonnative organisms and habitat degradation. Monitoring of natural populations should be done at least once a year between June and November. Preferably, natural populations will be monitored twice a year in order to document overwintering population minima and late summer population maxima (needed to evaluate limiting factors and genetic bottlenecking). Semiannual sampling should be conducted once during March or April and once during September or October.

Because regular, well structured monitoring is the only reliable means for evaluating the health of natural populations and evaluating/updating reintroduction methods, it is imperative to develop a comprehensive population and habitat monitoring protocol. This protocol must be sufficient to detect changes in habitat quality and to explain reasons for success/failure of natural as well as reintroduced populations. Any protocol used should fit with a well planned reintroduction study design aimed at determining habitat and population requirements for survival (see also Task 6).

Several natural resource agencies are involved in Gila topminnow monitoring of these populations. Therefore, a standardized monitoring protocol must be developed and implemented by responsible agencies. Comparable methodology (collecting gear, effort, season, location, etc.) should be carried out every year in order to provide an accurate assessment of population characteristics. Habitat data should be collected with population data. Each visit to a particular April-1993 Draft page 23 site should occur at approximately the same time of year in order to minimize seasonal variation occurrences. Permanent habitat photopoints and stream cross-section endpoints will aid in interpretation of other habitat data collected. After a broad inventory data set has been gathered on associated aquatic biota, topminnow distribution, physical habitat, water quality and quantity, watershed condition, etc., monitoring data tailored to determining trend should include the following categories at a minimum: date, time, location, weather, sampling technique, # fish captured, capture per unit of effort (cpue), size class distribution (adult vs juv.), temperature, water transparency, discharge, dominant aquatic (including algae) and semiaquatic (riparian) plants, dominant substrate types, dimensions of dominant macrohabitat types, percent instream cover, overstory, etc. Other site specific data may be necessary for site specific needs. Voucher specimens of fish should accompany any collection where doubt concerning identification exists. If practicable, underwater inspection for qualitative estimates of habitat and population may be incorporated into the above data set. Surface and underwater fish counts need verification of species identity since mosquitofish and topminnow are difficult to distinguish at a distance.

In the past, a large number of reintroduced populations precluded resource managers from evaluating them in a single field season. New reintroductions will be monitored twice during the first year. If the population has become established after the first year, it will be incorporated into a pool of reintroduced populations for monitoring at less frequent intervals. A priority system to visit these populations should be established every year. This priority system will determine which populations will be monitored during any given year and will depend on the number of extant populations. At a minimum, all introduced populations should be monitored within a period of two years. Monitoring should occur between March and October and include data elements listed above.

AGFD should be designated as the depository agency for the habitat and population monitoring data. Annual reports should be generated and distributed to other interested parties directly involved in the management of the Gila topminnow.

3.2 Determine criteria for extirpated populations.

Because collecting complete population census data on Gila topminnow is not feasible, it is difficult to ascertain when a population occupying a discrete location is extirpated. A reasonable approach is to sample a location repeatedly until a high degree of certainty is reached concerning the presence or absence of topminnow. Several populations have been mistakenly reported extirpated, only to reappear in the next monitoring period (Brown and Abarca 1992).

A Gila topminnow population will be considered extirpated when the following criteria are met:

1) Qualified biologists, from natural resource agencies or the academia, failed to collect topminnow in at least 5 documented consecutive samplings for natural populations and 3 documented consecutive samplings for reintroduced populations; and 2) Sampling covered the majority of the potential habitat and was not limited to

April-1993 Draft page 24 small segments of habitat; and 3) Sampling was conducted at times of the year with the highest potential of finding topinhinow (usually June through November); and 4) If mosquitofish were present, preserve all collected specimens at each sampling, and place them in an appropriate fish collection at a museum for further verification; and 5) If the site is observed to be completely dry by qualified biologist, efforts must be made to detect and sample surface waters above and below dry site as described in steps 1 to 4. If no other waters are present, then the population may be declared extirpated.

After a population is provisionally considered as extirpated, the following actions must be taken:

1) A report is prepared describing site history, sampling methodology and data, and probable causes for extirpation; and 3) The report is subjected to review by the Desert Fishes Recovery Team and other qualified biologists; and 3) The Recovery Team, in coordination with FWS and AGFD, will make a final decision regarding declaration of extirpation and will send a notification to interested parties.

TASK 4. DEVELOP AND IMPLEMENT GENETIC PROTOCOL.

A successful recovery program for an endangered species such as the Gila toprninnow, must take into account an evolutionary perspective that addresses the need for continued adaptive change in all populations (Meffe and Vrijenhoek 1988, Meffe 1990, Leberg 1990, Hendrickson and Brooks 1991). The optimal strategy for preserving both management options and evolutionary flexibility of taxa is to maintain as many populations as possible while retaining natural patterns of genetic flow within and between populations (Echelle 1991). Maintenance of the genetic diversity within species and populations has become a necessary approach for many threatened and endangered species (Frankel and Soule 1981, Templeton 1991, Templeton et al. 1991, Hedrick and Miller 1992).

Comprehensive genetic analyses for the Gila topinimiow started after massive reintroduction efforts began (Meffe and Vrijenhoek 1988). Initial studies on genetic geographic variation (allozyme) indicated the existence of three distinctive groups of natural populations from Arizona and Sonora (Vrijenhoek et al. 1985). The first group included all populations from the Gila River basin, Arizona and Rio Sonora and Rio de la Concepci6n, Sonora. The second group was formed by the entire Rio Yaqui, the Rio Matape, and the lower Rio Mayo. A third distinctive group comprised the upper Rio Mayo. April-1993 Draft page 25

Allelic diversity examined in 25 loci of 16 natural populations of Sonora topminnow, indicated that the populations at Cienega Creek, Monkey and Bylas springs, had no detectable genetic diversity (Vrijenhoek et al. 1985, Meffe and Vrijenhoek 1988). Hatchery produced Monkey Spring topinimiows were extensively used for the initial stocking program (Brooks 1985 1986). Since this population presented low fecundity (Schoenherr 1977, Constantz 1979), low growth rate, and survival (Quattro and Vrijenhoek 1989), it was recommended to replace Monkey Spring topminnow for Sharp Spring fishes, which presented higher levels of heterozygosity and higher rates for those traits. A more genetically diverse stock (Sharp Spring) might have a better chance to survive and adapt to new and changing environments, which is essential for the recovery of the species. Therefore, this recommendation was followed and since 1986 Sharp Spring constitutes the hatchery stock at Dexter National Fish Hatchery.

The hypothesis of positive correlation between high heterozygosity levels and fitness tested by Quattro and Vrijenhoek (1989), prompted recommendations of conducting experimental mixing of topminnow populations within the same subdivision (groups), but not among subdivisions. They also cautioned that existing natural localities should always remain genetically pure. This plan acknowledges both statements.

Remnant natural Gila topminnow populations in Arizona present genetic differences that were thought to be the result of local adaptations due to historical isolation (unique evolutionary lineages). However, further genetic studies utilizing mitochondrial DNA (mtDNA) (Joe Quattro pers. comm.) support the notion that the Gila basin historically harbored what was a single, essentially basin-wide, panmictic population, and that geographic differences are the result of recent bottlenecks probably caused by man-made actions.

Based on this rationale, recovery actions proposed in this plan should attempt to restore genetic flow within subasins of the Gila River, as well as between subasins, first under controlled habitats, and later (if prove to be adequate) under field experiments. In addition, the use of Mexican populations (high heterozygosity levels) as part of the reintroduction effort in the Gila River basin (low heterozygosity levels) should be considered as a possibility to restore and improve genetic diversity of the Arizona populations.

As a first step for the genetic protocol, a species-wide computer model of topminnow population genetics should be developed using existing information of genetic composition, basic biology, rough estimates of population size, etc. The model should be especially detailed for the Gila River basin, incorporating not only management of remaining natural populations as isolated demes, but also incorporate human-controlled gene flow among introduced populations to be established with stocks from all the remaining natural populations (and hybrids of these stocks).

The second step would be to implement highly controlled laboratory and/or large-scale outdoor experiments using semi-natural mesocosms, in addition, and in coordination with, the field introductions of Task 2 (item B). In addition to using remnant Arizona stocks, these experiments should also include stocks from the Rio de la Concepcion and (perhaps) Rio Sonora, and progeny from crosses of these Sonoran stocks with Arizona stocks. Such crosses should provide April-1993 Draft page 26 increases in heterozygosity which will be easily quantifiable using alleles which have already been surveyed.

The theoretical model described above should be utilized along with a comprehensive accounting of all sites available for reintroductions in the Gila basin to design a field experiment which utilizes the model, available sites, and native Arizona stocks (only) to test the same hypotheses tested in the mesocosm experiments relating fitness in the field. The exact design for this field experiment should be developed by competent population geneticists, but attempts should also be made to incorporate habitat quality effects in the model. A graphic representation of the reintroduction program is provided in Figure 3. Essentially, each natural population would be stocked to reintroduction sites near the parental population. Other sites, from which there would be little or no chance of gene flow to natural populations (outer ring), would receive other pure stocks of remaining natural Arizona populations or hybrids of them. The bottom line intent of the design would be to test for fitness differences among hybrids and pure stocks, so adequate numbers of each would be needed. Specific areas for implementing experimental stockings should follow descriptions in Task 2 (item B).

Along with population and habitat information, a single agency database should be developed in conjunction with those conducting field experiments, as described above, which will assure inter-agency coordination for full compliance or reintroductions with the experimental design, and detail tracking of stockings.

Concurrent with activities in Task 2 and model testing, additional genetic studies must be conducted using new and more powerful techniques to detect inter-population differences among naturally occurring populations in the Gila River basin, including extinct stocks as preserved in museum collections. In addition, allozyme and mtDNA analyses should be completed for those natural populations lacking that information.

Should experiments described above indicate increased fitness of hybrids carrying Mexican alleles (as would be hypothesized), and if the recovery effort using mixed stocks of Arizona populations is not appearing to show greater signs of success than did earlier reintroduction efforts, then field introductions to areas described in Task 2 (item B) should begin to incorporate alleles from these Mexican populations.

Periodic genetic monitoring of both existing and reintroduced populations should be implemented. Allozyme diversity monitoring at five-year intervals is recommended. Scientifically accepted protocols should be followed to collect, freeze, store and analyze adequate population subsamples. April-1993 Draft page 27

Figure 3. Graphic representation of the Gila topminnow reintroduction plan proposal. Patterned disks depict discrete and pure stocks of natural populations. Solid gray disks depict hybrid stocks containing genetic material from >1 (or all) natural populations. Large patterned disks depict natural populations in their natural habitats. Small disks depict introduced populations. Outer ring of introduced populations are in sites in parts of Gila drainage isolated from natural populations. April-1993 Draft page 28

TASK 5. MAINTAIN CAPTIVE POPULATIONS.

Artificial refugia are an important component of the effort to preserve several endangered or nearly endangered fish species, especially the highly endemic and severely threatened fish fauna of the North American deserts (Pister 1981, Johnson and Jensen 1991). These refugia should preserve a large fraction of the genetic variability originally present in their progenitors (Turner 1984).

Captive populations of each extant demes in Arizona should be maintained at separated suitable facilities under the control of an agency or organization mandated to or dedicated to legal protection of the Gila topminnow in United States. A captive population should contain a minimum of 500 overwintering adults or existing numbers, whichever is greater, and possess an adequate representation of all age classes and evidence of reproductive activity.

Patterns of genetic variation in artificial populations may vary from those in natural populations (Templeton 1991, Jason Dunham pers. comm.). Each captive population should be assessed for genetic diversity every 5 years and be periodically supplemented with fish from the correspondent natural location if such action proves to be advantageous. Similar plans should be developed in Mexico to secure populations from the Rio de la Concepcion and Rio Sonora.

Dexter National Fish Hatchery has played a major role in the reintroduction program for the Gila topminnow. Literally thousands of topminnows (from Monkey and Sharp springs) have been produced by the hatchery and stocked into 178 sites in Arizona since 1981 (Johnson and Jensen 1991). Other captive populations include populations held at zoos, museums, universities and backyards (Bagley et al. 1991, Brown and Abarca 1992). Since these populations present high fluctuations in size and structure, periodic reviews of currently maintained captive populations must be implemented to determine the importance of such populations in the overall recovery program.

TASK 6. STUDY LIFE-HISTORY AND INTERACTIONS WITH NONNATIVE FTSHFS.

Because of the large number of sites to be visited, most of the natural and reintroduced populations have only been evaluated for the presence and abundance of topminnow and habitat type and quality. A more quantitative approach and more rigorous statistical analyses need to be explored to further our understanding of topminnow habitat and population dynamics. Studies should be focussed on quantifying minimum viable population size, and survival and reproduction success. Thorough understanding of these aspects, coupled with successful April-1993 Draft page 29

experimental genetic manipulations, may be used as a basis to develop delisting criteria.

Further studies on Gila topminnow may include, but not be limited to, minimum temperature thresholds, hardening and related mortality; temperature preference and preference breadth; minimum oxygen requirements; emergent plant density as a limiting factor; resistance to flooding under different channel configurations and temperatures; holding and transportation stress and associated mortality; niche partitioning and shift in carrying capacity when syntopic with historic native fishes, especially desert pupfish; differences in water quality; local dietary composition at different natural sites; Interactions between topminnow and nonnative fishes at various life stages; cause and incidence of diseases at existing populations; and movement patterns of adult and juvenile topminnow.

Future protection actions against upstream invasion by mosquitofish will certainly include fish bathers in those subasins currently occupied by the Gila topminnow. Close population and genetic monitoring will be necessary to document effects of this additional "fragmentation."

TASK 7. INFORM AND EDUCATE THE PUBLIC AND RESOURCE MANAGERS.

As part of the recovery actions for the Gila topminnow, a public information and education program should be developed to inform interested parties of the objectives and needs of the recovery program. An informed and caring public will provide strong support to the conservation of endangered species, in particular to the Gila topminnow. The desert pupfish (Cyprinodon macularius), has gained popularity among young scholars at high school level (Randy Babb pers. comm.) thanks to a successful education and display program.

Information and education materials must be developed in formats that are appropriate for the target audience, and must be prepared in both English and Spanish. These materials may take the form of brochures, newspaper and magazine articles, videotape or slide presentations, displays of live topminnows, television presentations, seminars, and workshops. When possible, the media and environmental groups must be encouraged to disseminate information.

All participating agencies and groups should participate in periodic meetings to update and exchange information pertinent to the recovery program of the Gila topminnow. Training seminars, particularly on proper sampling methodology and identification of the Gila topminnow and mosquitofish, should be implemented as needed, especially when new resource managers start to participate in management activities. April-1993 Draft page 30 LITERATURE CITED

Angus, R.A. 1980. Geographic dispersal and clonal diversity in unisexual fish populations. Am.Nat. 115(4):531-550.

Angus, R.A., and R.J. Schultz. 1979. Clonal diversity in the unisexual fish Poeciliopsis monacha-lucida: A tissue graft analysis. Evolution 33:27-40.

Arizona Game and Fish Department. 1988. Threatened Native Wildlife in Arizona. Arizona Game and Fish Department. Phoenix, Arizona.

Baird, S.F., and C. Girard. 1853. Description of new species of fishes collected by Mr. John H. Clarck, on the U.S.-Mexican boundary survey, under Lt. Col. Jas. D. Graham. Proc. Acad. Nat. Sci. Philadelphia 6:387-390.

Bahre, C.J. 1991. A legacy of change historic human impacts on vegetation in the Arizona borderlands. University of Arizona Press, Tucson.

Bagley, B.E., D.A. Hendrickson, F.J. Abarca, and S. Hart. 1991. Status of the Sonoran Topminnow (Poeciliopsis occidentalis) and Desert Pupfish (Cyprinodon macularius) in Arizona. Arizona Game and Fish Department, Phoenix, Arizona.

Brooks, J.E. 1985. Factors affecting the success of Gila topminnow introductions on four Arizona National Forests. Report to U.S. Fish and Wildlife Service, Office of Endangered Species, Albuquerque, New Mexico. Arizona Game and Fish Department, Phoenix, Arizona.

Brooks, J.E. 1986. Status of natural and introduced Sonoran topminnow (Poeciliopsis o. occidentalis) populations in Arizona through 1985. Report to U.S. Fish and Wildlife Service, Office of Endangered Species, Albuquerque, New Mexico. Arizona Game and Fish Department, Phoenix, Arizona.

Brown, M. and F.J Abarca. 1992. An update status report of the Sonoran topminnow, Poeciliopsis occidentalis, and desert pupfish, Cyprinodon macularius, in Arizona. Phoenix, AZ: Arizona Game and Fish Dept.

Bryan, K. 1925. Date of channel trenching (arroyo cutting) in the arid southwest. Science 62(1607):338-344.

Campoy-Favela, J., A. Varela-Romero, and L. Juarez-Romero. 1989. Observaciones sobre la ictiofauna nativa de la cuenca del Rio Yaqui, Sonora, Mdxico. Ecologica 1(1):1-29. April-1993 Draft page 31

Collins, J.P., C. Young, J. Howell and W.L. Minckley. 1981. Impact of flooding in a Sonoran Desert stream, including elimination of an endangered fish population (Poeciliopsis o. occidentalis, Poeciliidae). Southw. Nat. 26(4):415-423.

Constantz, G.D. 1974. Reproductive effort in Poeciliopsis occidentalis (Poeciliidae). Southw. Nat. 19:47-52.

Constantz, G.D. 1975. Behavioral ecology of mating in the male Gila topminnow, Poeciliopsis occidentalis (: Poeciliidae). Ecology 56:966-973.

Constantz, G.D. 1976. Life history strategy of the Gila topminnow, Poeciliopsis occidentalis: A field evaluation of theory on the evolution of life histories. Unpub. Ph.D. Dissert. Arizona State University. Tempe, Arizona.

Constantz, G.D. 1980. Energetics of viviparity in the Gila topminnow (Pisces: Poeciliidae). Copeia 4:876-878.

Constantz, G.D. 1989. Reproductive biology of Poeciliid fishes. In: Ecology and Evolution of livebearing fishes (Poeciliidae), edited by Meffe, G.K. and F.F. Snelson Jr. Englewood Cliffs, New Jersey: Prentice Hall, p. 33-50.

Deacon, J.E. 1979. Endangered and threatened fishes of the west. Great Bas. Nat. Mem. 3:41-64.

Deacon, J.E. and W.L. Minckley. 1974. Desert Fishes. In: Desert Biology, Volume II, edited by Brown, G.W. New York: Academic Press, p. 385-488.

Deacon, I.E., and W.L. Minckley. 1991. Western fishes and the real world: The enigma of "Endangered species" revisited. In: Battle Against Extinction - Native Fish Management in the American West, edited by Minckley, W.L. and J.E. Deacon. University of Arizona Press. Tucson, Arizona. pp. 405-413.

Deacon, J.E., G. Kobetich, J.D. Williams and S. Contreras-l3alderas. 1979. Fishes of North America Endangered, Threatened, or of Special Concern. Fisheries 4:29-44.

Echelle, A.A. 1991. Conservation genetics and genic diversity in freshwater fishes of western North America. In: Battle Against Extinction - Native Fish Management in the American West, edited by Minckley, W.L. and J.E. Deacon. University of Arizona Press. Tucson, Arizona. pp. 141-153.

Forrest, R.E. 1992. Habitat use and preference of Gila topminnow. Unpub. M.S. Thesis. University of Arizona. Tucson, Arizona. April-1993 Draft page 32

Fradkin, P.L. 1981. A River No More: The Colorado River and the West. University of Arizona Press, Tucson.

Frankel, 0.H., and M.E. Soule. 1981. Conservation and evolution. Cambridge University Press.

Gerking, S.D. and D.V.J. Plantz. 1980. Size-biased predation by the Gila Topnthmow, Poeciliopsis occidentalis (Baird and Girard). Hydrobiologia 72:179-191.

Gomez-Alvarez, A., M.T. de J. Yocupicio-Anaya and P. Ortega-Romero. 1990. Niveles y distribucion de metales pesados en el Rio Sonora y su afluente el Rio Bacanuchi, Sonora, Mexico. Ecologica 1(2):10-20.

Hastings, J.R. and R.M. Turner. 1966. The changing mile: an ecological study of vegetation change with time in the lower mile of an arid and semiarid region, Tucson, Arizona:University of Arizona Press, pp. 1-317.

Heath, W.G.. 1962. Maximum temperature tolerance as a function of constant temperature acclimation in the Gila topminnow, Poeciliopsis occidentalis. Unpub. Ph.D. Dissert. University of Arizona, Tucson, Arizona.

Hedrick, P.W., and P.S. Miller. 1992. Conservation genetics: techniques and fundamentals. Ecol. Applications 2(1):30-46.

Hendrickson, D.A. 1983. Distribution records of native and exotic fishes in Pacific drainages of northern Mexico. J. Ariz. Nev. Acad. Sci. 18(2):33-38.

Hendrickson, D.A. and D.M. Kubly. 1984. Desert Waters: Past, Present, and Future. Nature Conservancy News 34(5):6-12.

Hendrickson, D.A. and L. Juarez-Romero. 1990. Los Peces de la cuenca del Rio de la Concepci6n, Sonora, Mexico, y el estatus del charalito Sonorense, Gila ditaenia, una especie en amenaza de extincion. Southw. Nat. 35(2):177-187.

Hendrickson, D.A. and J.E. Brooks. 1991. Transplanting short-lived fishes in North American deserts: Review, Assessment and Recommendations. In: Battle Against Extinction - Native Fish Management in the American West, edited by Minckley, W.L. and J.E. Deacon. University of Arizona Press. Tucson, Arizona. pp. 283-298.

Hendrickson, D.A., W.L. Minckley, R.R. Miller, D.J. Siebert and P.H. Minckley. 1981. Fishes of the Rio Yaqui basin, Mexico and United States. J. Ariz. Nev. Acad. Sci. 15(3):65-106. April-1993 Draft page 33

Keegan-Rogers, V. and R.J. Schultz. 1988. Sexual selection among clones of unisexual fish Poeciliopsis:Poeciliidae, genetic factors and rare-female advantage. Am. Nat. 132:846-868.

Lanza, J. 1983. Population control of an all female fish Poeciliopsis-Monacha-Latidens. Biotropica 15:100-107.

Leberg, P.L. 1990. Genetic considerations in the design of introduction programs. Trans. N.A. Wild!. and Nat. Res. Conf. 55:609-619.

Marsh, P.C. and W.L. Minckley. 1990. Management of endangered Sonoran topminnow at Bylas Springs, Arizona. Description, critique and recommendations. Great Basin Naturalist 50 (3):265-272.

McNatt, R.M. 1979. Discovery and loss of a new locality for the Gila Topminnow, Poeciliopsis occidentalis occidentalis (Poeciliidae). Southw. Nat. 24(2):555-556.

Meffe, G.K. 1983. Attempted chemical renovation for an Arizona spring brook for management of the endangered Sonoran topminnow. N. Amer. J. Fish. Mgmt. 3:315-321.

Meffe, G.K. 1984. Effects of abiotic disturbance on coexistence of predator-prey fish species. Ecology 65(5):1525-1534.

Meffe, G.K. 1985. Predation and species replacement in American southwestern fishes: A case study. Southw. Nat. 30:173-187.

Meffe, G.K. 1990. Genetic approaches to conservation of rare fishes: examples from North American desert species. J. Fish Biol. 37:105-112.

Meffe, G.K. and M.L. Crump. 1987. Possible growth and reproductive benefits of cannibalism in the Mosquitofish. Am. Nat. 129(2):203-212.

Meffe, G.K. and R.C. Vrijenhoek. 1988. Conservation genetics in the management of desert fish. Conserv. Biol. 2(2):157-169.

Meffe, G.K., D.A. Hendrickson and J.N. Rinne. 1982. Description of a new topminnow population in Arizona, with observations on topminnow/mosquitofish co-occurence. Southw. Nat. 27(2):226-228.

Meffe, G.K., D.A. Hendrickson and W.L. Minckley. 1983. Factors resulting in decline of the endangered sonoran topminnow, Poeciliopsis occidentalis (Antheriniformes: Poeciliidae) in the United States. Biol. Conserv. 25:135-159. April-1993 Draft page 34

Miller, R.R. 1961. Man and the changing fish fauna of the American Southwest. Pap. Michigan Acad. Sci., Arts, Lett. 46:365-404.

Miller, R.R. 1972. Threatened freshwater fishes of the United States. Trans. Amer. Fish. Soc. 2:239-252.

Miller, R.R. and C.H. Lowe. 1964. Fishes of Arizona. In: The vertebrates of Arizona, edited by Lowe, C.H. Tucson, Arizona: University of Arizona Press, pp. 133-151.

Minckley, W.L. 1969a. Native Arizona fishes, Part 1-Livebearers. Wild!. Views. 16:6-8.

Minckley, W.L. 1969b. Attempted reestablishment of the Gila topminnow within its former range. Copeia 1969:193-194.

Minckley, W.L. 1973. Fishes of Arizona. Arizona Game and Fish Department. Phoenix, Arizona. pp. 1-293.

Minckley, W.L. 1985. Native fishes and natural aquatic habitats in U.S. Fish and Wildlife Service Region II, west of the continental divide. Albuquerque, New Mexico. U.S.D.I. Fish and Wildlife Service, Div. of Endan. Species, pp. 1-158.

Minckley, W.L. and J.E. Deacon. 1968. Southwestern fishes and the enigma of "Endangered Species". Science 159:1424-1432.

Minckley, W.L. and J.E. Brooks. 1985. Transplantations of native Arizona fishes: records through 1980. J. Ariz. Nev. Acad. Sci. 20(2):73-90.

Minckley, W.L. and G.K. Meffe. 1987. Differential selection by flooding in stream fish communities of the arid american southwest. In: Community and evolutionary ecology of north american stream fishes, edited by Matthews, W.J. and D.C. Heins. Norman, Oklahoma: University of Oklahoma Press p. 93-104.

Minckley, W.L., J.N. Rinne and J.E. Johnson. 1977. Status of the Gila topminnow and its co- occurrence with mosquitofish. USDA Forest Service Research Paper RM-198:1-8.

Minckley, W.L., G.F. Meffe, and D.L. Soltz. 1991. Conservation and management of short- lived fishes: the Cyprinodontoids. In: Battle Against Extinction - Native Fish Management in the American West, edited by Minckley, W.L. and I.E. Deacon. University of Arizona Press. Tucson, Arizona. pp. 247-282.

Moore, W.S. and F.E. McKay. 1971. Coexistence in unisexual - bisexual species complexes of Poeciliopsis (Pisces:Poeciliidae). Ecology 52:791-799. April-1993 Draft page 35

Moore, W.S., R.R. Miller and R.J. Schultz. 1970. Distribution, adaptation and probable organ of an all-female form of Poeciliopsis (Pisces: Poeciliidae) in Northwestern Mexico. Evolution 24(4):789-795.

Morizot, D.C., R.J. Schultz and R.S. Wells. 1990. Assignment of six enzyme loci to multipoint linkage groups in fishes of the genus Poeciliopsis (Poeciliidae): Designation of linkage groups III-V. Biochem. Genet. 28(1-2):83-96.

Ono, R.D., J.D. Williams and A. Wagner. 1983. Vanishing fishes of North America, Washington, D.C. Stone Wall Press.

Pister, E.P. 1981. The conservation of desert fishes. In: Fishes in North American Deserts, edited by Naiman, R.J. and d.l. Soltz. New York, New York: John Wiley & Sons p. 411-446.

Quattro, J.M. and R.C. Vrijenhoek. 1989. Fitness differences among remnant populations of the endangered Sonoran topminnow. Science 245:976-978.

Quattro, J.M., P.L. Leberg, and R.C. Vrijenhoek. 1992. Mitochondrial DNA diversity, within and among populations of the endangered Sonoran topminnow (Poeciliopsis occidentalis). Abstract. Proc. Desert Fishes Council 22-23:67.

Rea, A. 1983. Once a River: Bird life and habitat changes in the middle Gila. University of Arizona Press, Tucson.

Rinne, J.N., and P.R. Turner. 1991. Reclamation and alteration as management techniques, and a review of methodology in stream renovation. In: Battle Against Extinction - Native Fish Management in the American West, edited by Minckley, W.L. and J.E. Deacon. University of Arizona Press. Tucson, Arizona. pp. 219-244.

Rogers, J.A. 1980. The geographic and historic importance of cienega environments in the upper Santa Cruz River basin. Presented at AZ-NEV Academy of Sciences and Southwest and Rocky Mountain Regional Annual Meeting of AAAS.

Rosen, D.E. and R.M. Bailey. 1963. The Poeciliid fishes (Cyprinodontiformes), their structure, zoogeography, and systematics. Bull. Amer. Mus. Nat. Hist. 126(1):1-176.

Rosgen, D.L. 1980. A stream classification system. Presented at the Symposium - Riparian ecosystems and their management: Reconciling conflicting uses. Tucson, Arizona.

Schenck, R.A. and R.C. Vrijenhoek. 1986. Spatial and temporal factors affecting coexistence among sexual and clonal forms of Poeciliopsis. Evolution 40:1060-1070. April-1993 Draft page 36

Schoenherr, A.A. 1974. Life history of the topminnow, Poeciliopsis occidentalis (Baird and Girard) in Arizona, and an analysis of its interaction with the mosquitofish Gambusia affins (Baird and Girard) Unpub. Ph.D. Dissert. Arizona State University. Tempe, Arizona.

Schoenherr, A.A. 1977. Density dependent and density independent regulation of reproduction in the Gila topminnow, Poeciliopsis occidentalis (Baird and Girard). Ecology 58:438-444.

Schoenherr, A.A. 1981. The role of competition in the replacement of native fishes by introduced species. In: Fishes in North American Deserts, edited by Nairnan, R.J. and D.L. Soltz. New York, New York: John Wiley & Sons, p. 173-204.

Schultz, R.J. 1961. Reproductive mechanism of unisexual and bisexual strains of the viviparous fish Poeciliopsis. Evolution 15:302-325.

Schultz, R.J. 1989. Origins and relationships of unisexual Poeciliids. In: Ecology and Evolution of livebearing fishes (Poeciliidae), edited by Meffe, G.K. and F.F. Snelson Jr. Englewood Cliffs, New Jersey: Prentice Hall, p. 69-87.

Simms, J.K., and K.M. Simms. 1992. What constitutes quality habitat for Gila topminnow (Poeciliopsis occidentalis)? An overview of habitat parameters supporting a robust population at Cienega Creek, Pima, Co., Arizona. Abstract. Proc. Desert Fishes Council 22-23:82.

Simons, L.H. 1987. Status of the Gila topminnow (Poeciliopsis o. occidentalis) in the United States. Special Report on Project E-1, Title VI of the Endangered Species Act of 1973, Phoenix, Arizona:Arizona Game and Fish Department.

Simons, L.H., D.A. Hendrickson and D. Paponlias 1989. Recovery of the Gila topminnow: a success story? Conserv. Biol. 3(1):11-15.

Templeton, A.R. 1990. The role of genetics in captive breeding and reintroduction for species conservation. Endangered Species Update 8(1):14-17.

Templeton, A.R., K. Shaw, E. Routman and S.K. Davis. 1990. The genetic consequences of habitat fragmentation. Annals of the Missouri Botanical Garden 77:13-27.

Turner, B.J. 1984. Evolutionary genetics of artificial refugium populations of an endangered species, the desert pupfish. Copeia 1984(2):364-369.

U.S. Fish and Wildlife Service. 1984. Gila and Yaqui Topminnow Recovery Plan. U.S. Fish and Wildlife Service, Albuquerque, New Mexico. April-1993 Draft page 37

Varela-Romero, A., L. Juarez-Romero and J. Campoy-Favela. 1990a. Los peces dulceacuicolas de Sonora. Proc. Desert Fish. Counc. 22:68.

Varela-Romero, A., L. Juarez-Romero and J. Campoy-Favela. 1990b. Los Peces Dulceacuicolas de Sonora, Mexico. Centro Ecologico de Sonora. pp. 1-236.

Vives, S.P. 1991. Alarm reactions of some southwestern fishes. Abstract. Proc. Desert Fishes Council 20-21:227.

Vrijenhoek, R.C., R.A. Angus and R.I. Schultz. 1977. Variation and heterozygosity in sexually vs. clonally reproducing populations of Poeciliopsis. Evolution 31(4):767-781.

Vrijenhoek, R.C., M.E. Douglas and G.K. Meffe. 1985. Conservation genetics of endangered fish populations in Arizona. Science 229:400-402.

Wetherington, J.D., S.C. Weeks, K.E. Kotora and R.C. Vrijenhoek. 1989. Genotypic and environmental components of variation in growth and reproduction of fish hemiclones Poeciliopsis:Poeciliidae. Evolution 43:635-645.

Williams, J.E. and R.R. Miller. 1990. Conservation status of the North American fish fauna in fresh water. J. Fish Biol. 37A:79-85.

Williams, J.E., D.B. Bowman, J.E. Brooks, A.A. Echelle, R.J. Edwards, D.A. Hendrickson and J.L. Landye. 1985. Endangered aquatic ecosystems in North American deserts with a list of vanishing fishes of the region. J. Ariz. Nev. Acad. Sci. 20:1-62.

Williams, J.E., D.W. Sada, C.D. Williams, J.R. Bennett, I.E. Johnson, P.C. Marsh, D.E. McAllister, E.P. Pister, R.D. Radant, J.N. Rinne, M.D. Stone, L. Ulmer and D.L. Withers. 1988. American Fisheries Society guidelines for introductions of threatened and endangered fishes. Fisheries 13(5):5-11.

Williams, J.E., J.E. Johnson, D.A. Hendrickson, S. Contreras-Balderas, J.D. Williams, M. Navarro-Mendoza, D.E. McAllister and J.E. Deacon. 1989. Fishes of North America endangered, threatened, or of special concern: 1989. Fisheries 14(6):2-20. April-1993 Draft page 38

IMPLEMENTATION SCHEDULE

Definition of Priorities

Priority 1 - All actions that are absolutely essential to prevent the extinction of the species in the foreseeable future.

Priority 2 - All actions necessary to prevent a significant decline in species population/habitat quality or some other significant negative impact short of extinction.

Priority 3 - All other actions necessary to provide for full recovery of the species.

Abbreviations Used

AGFD = Arizona Game and Fish Department NMDGF = New Mexico Department of Game and Fish FWS = Fish and Wildlife Service FS = Forest Service BLM = Bureau of Land Management SCAIR = San Carlos Apache Indian Reservation ASPR = Arizona State Parks and Recreation TNC = The Nature Conservancy CES = Centro Ecologic° de Sonora

FR = Fish and Wildlife Service, Fisheries Resources Program HR = Fish and Wildlife Service, Habitat Resources Program PAO = Fish and Wildlife Service, Public Affairs Office April-1993 Draft page 39

PART III - IMPLEMENTATION SCHEDULE

Responsible Agency Cost Estimates ($000's) Priority Task Plan Task ' N N Task Duration FWS Conunenta Other FY FY FY FY FY (Yrs) Region I Program 1 2 3 4 5 1 1 1.1 Identify extent of geographic X 2 FWE,RE distribution of natural AGFD I populations. FS I BLM I SCAIR ASPR TNC CES 1 1.2 Acquire habitats currently x 2 FVVE,RE occupied by remanent AGFD populations. FS BLM TNC CBS

1 1.3 Protect natural populations from X 2 FWE invasion by nonnative fishes. AGFD FS BLM SCAIR ASPR TNC CBS

1 1.4 Protect habitats currently X 2 FWE,FR occupied by natural populations AGFD from detrimental land and water FS use practices. BLM SCAIR ASPR TNC CBS April-1993 Draft page 40

PART III - IMPLEMENTATION SCHEDULE

Responsible Agency Cost Estimates ($000's) Priority Task Plan Task # # Task Duration FWS Comments Other Y Y Y Y Y (yrs) Region I Progam

2 2.0 Re-establish populations X 2 FWE,FR throughout historic range AGFD FS BLM SCAIR ASPR TNC CES

1 3.1 Develop standardized X 2 FWE monitoring protocol. AGFD NMDGF FS BLM CES

2 3.2 Determine criteria for X 2 FWE extirpated population. AGFD FS BLM SCAIR TNC CES

2 4.0 Develop and implement genetic X 2 FWE protocol. AGFD NMDGF FS BLM CES April-1993 Draft page 41

PART III - IMPLEMENTATION SCHEDULE

Responsible Agency Cost Estimates ($000's) Priority Task Plan Task # # Task Duration FWS Cominents Other FY FY FY FY FY (Yrs) Region I Program 1 2 3 4 5

2 5.0 Maintain captive populations. X 2 FVVE AGFD FS BLM

3 6.0 Study life-history and X 2 FWE interactions with nonnatives. AGFD FS BLM

3 7.0 Inform and educate the public. X 2 FWE,PAO AGFD FS BLM SCAIR ASPR TNC CES - Total I II I April-1993 Draft page 42

Table 1. Gila topmismow historic records from the United States. Specimens are housed at Arizona State University (ASU), United States National Museum (USNM), University of New Mexico (UofNM), and University of Michigan (UMMZ).

LOCATION COLLECTORS DATE MUSEUM

Gila River

Bylas Springs Johnson, J. E. 1968 ASU

Simons, L. and Brooks, J. 1986 ASU

Frisco Hot Springs Koster, W. J. 1948 UoINM

Gila River - near Adonde Siding Mearns, E. A. 1894 USNM

Gila River -2 mi. below Dome Hubbs & Schultz 1926 UMMZ

Gila River - near Gila Mearns, E. A. 1894 USNM

Gila River - just below Gillespie Dam Knutzthor, G. M. 1929 USNM Myers, G. S.

Gila River - San Carlos Reservoir 1 mi. NE Johnson, J. E. 1968 ASU of Bylas

Gila River -1 mi. below Winkleman Simon, J. R. 1943 UMMZ

North Fork of Ash Creek Jennings, M 1985 ASU

Artesian spring fed ditch and reservoir 7 mi. Miller, R. R. 1950 UMMZ SE of Safford Winn, H. E.

Farm pond 6.5 mi. SE of Safford Minckley, W. L. 1964 ASU Koehn, R. K.

Tributary of Gila River near Phoenix Arizona Fish and Game Comm. 1934 UMMZ

Salt River

Salt River - between Phoenix and Tempe Hubbs and Schultz 1926 USNM

Salt River - near Roosevelt Chamberlain, F. M. 1904 USNM

Tempe Gilbert, C. H. USNM

Tonto Creek - near Roosevelt Chamberlain, F. M. 1904 USNM April-1993 Draft page 43 Table 1. Continuation.

LOCATION COLLECTORS DATE MUSEUM

Tonto Creek - midway between Roosevelt Hubbs and Schultz 1926 UMMZ Dam and Payson

Tonto Creek -14 mi. above Roosevelt Lake Gee, M. A. 1936 UMMZ

Lower Tonto Creek Gee, M. A. 1936 UofNM

Tonto Creek - 10 mi. above Roosevelt Lake Hubbs, L. G. 1941 UMMZ

San Carlos River

3 mi. above San Carlos Lake Hubbs, L. G. 1941 UMMZ

San Pedro River

San Pedro River - 4 mi. N of Feldman Simon, J. R. 1943 UMMZ

Artesian Spring 13 km SE of Mammoth McNatt, R. 1978

Santa Cruz River

Cienega Creek Various 1974 ASU

Demarais, B. and Meffe, G. 1983 ASU

Hendrickson, D. & Simons, L. 1986 ASU

Cocio Wash Hanks, K. 1969 UMMZ

McNatt, R. and Constantz, G. 1972 ASU

Cocio Wash Constantz, G. 1973 ASU

Constantz, G. 1974 ASU

Constantz, G. 1975 ASU

Cottonwood Spring Hubbs & Family 1938 UMMZ

Simon, J. R. 1943 UMMZ

Minckley, W. L. 1965 ASU

Various 1967 ASU

Hendrickson, D. & Simons, L. 1986 ASU

Fresno Canyon Weedman, D. A. 1992 April-1993 Draft page 44 Table 1. Continuation.

LOCATION COLLECTORS DATE MUSEUM

Lower Lampshire Canyon Simons, L. 1987 ASU

Monkey Spring Chamberlain, F. M. 1904 USNM

Hubbs & Family 1938 UMMZ

Minckley, W. L. & Koehn, R. K. 1964 ASU

Minckley, W. L. 1965 ASU

Barber, W. E. 1966 ASU

Constantz, G. 1973 ASU

Constantz, G. 1974 ASU

Constantz, G. 1975 ASU

1980 ASU

Potrero Creek Simon, J. R. 1943 UMMZ

Redrock Canyon Hendrickson, D. & Simons, L. 1986 ASU

Simons, L. and Dean, G. 1987 ASU

Sabino Creek Kranzthor, G. M. Myers, G. S. 1929 USNM

Simon, J. R. 1943 UMMZ

Santa Cruz River - near Gage Gorsuch & Ashburn 1939 UMMZ

Minckley, W. L. 1978 ASU

Weser, J. and Simons, L. 1986 ASU

Various 1987 ASU

Santa Cruz River - 2 mi. NE of L,ochiel Ashburn, M. F. 1940 UMMZ

Santa Cruz River -7 mi. NNE of Lochiel Vorchies et al. 1943 UMMZ

Santa Cruz River - near Nogales Chamberlain, F. M. 1904 USNM

Santa Cruz River - near San Xavier Chamberlain, F. M. 1904 USNM

Santa Cruz River - near Tucson Brown, H. 1893 USNM Chamberlain, F. M. 1904 USNM

Santa Cruz River - 7 mi. S of Tucson Simon, J. R. 1943 UMMZ April-1993 Draft page 45

Table 1. Continuation.

LOCATION COLLECTORS DATE MUSEUM

Sharp Spring Hendrickson, D. & Simons, L. 1986 ASU

Various 1987 ASU

Sheehy Spring Ashburn & Gorsuch 1939 UMMZ

Ashburn, M. F. 1940 UMMZ

Hendrickson, D. & Simons, L. 1986 ASU

Sonoita Creek - near Cottonwood Spring Simon, J. R. 1943 UMMZ

Minckley, W. L. & Rime, J. 1967 ASU

Sonoita Creek - near Patagonia Chamberlain, F. M. 1904 USNM

Minckley, W. L. & Rinne, J. 1967 ASU

Simons, L. 1986 ASU

Sonoita Creek - 2.6 mi. SW of Patagonia, Minckley, W. L. 1967 ASU pool off Creek

Sonoita Creek -3 mi. SW of Patagonia Burt, C. E. 1928 UMMZ

Sonoita Creek -3 mi. SW of Patagonia Minckley, W. L. & Johnson, J. E. 1967 ASU

Sonoita Creek - 3.5 mi. below Patagonia Hubbs & Family 1938 UMMZ

Sonoita Creek - below Patagonia Lake Cindy, H. 1973 ASU

Frantz, B. and Silvey, B. 1976 ASU

Ginelly, H. et al. 1977 ASU

Bagley, B. & Dean, G. 1987 ASU

Spring 50 ft. W of Tanque Verde Creek Simon, J. et al. 1943 UMMZ

Spring 200 ft. E of Tanque Verde Creek Simon, J. et al. 1943 UMMZ April-1993 Draft page 46

Table 2. Status of natural populations of Gila topminnow, Poeciliopsis occidentalis occidentalis, in the United States. Site number corresponds with Simons' (1987) System.

Location Site # Ownership Comments

Cottonwood Spring 1 Private Small but stable population of topminnows contained in 40 m long stream. No exotic fish. Monkey Spring 2 Private Topminnows found in springhead and cement canal. Large population, no exotic fish. Redrock Canyon 11 USFS Topminnows coexist with mosquitofish, longfin dace and largemouth bass. Sonoita Creek 9 State Parks Topminnows are found in three locations: 1) Near Cottonwood Spring - small population, and mosquitofish were found 200 m downstream in 1991; 2) Near Patagonia- small population, one Recreation, found in 1986, 1987, & 1990 (Simons 1987; Brown & Abarca 1992); 3) Below Patagonia Lake- Private topminnows have coexisted with mosquitofish since 1969. Also present are longfin dace, red shiner, and fathead Sheehy Spring 3 Private Mosquitofish first recorded in 1979. No topminnow have been collected here since 1987. Santa Cruz River 10 Private Intermittent stream near the gaging station NE of Lochiel contains topminnows, mosquitofish, green sunfish, fathead minnow, largemouth bass, longfin dace, Sonora sucker, desert sucker, bluegill, yellow & black bullheads. Sharp Spring 4 Private Topminnow population has coexisted with mosquitofish in various pools since discovered in 1979 (Meffe et al. 1982). Mosquitofish dominance ranged from 76-100% in 1990 (Brown and Abarca 1992).

Fresno Canyon 164 Private Discovered in 1992. Topminnow out number mosquitofish 11 to 1 (n=96). Also present are green sunfish, largemouth bass and fathead minnow Cienega Creek 5 BLM, Private Topminnows are found in over 8 miles of creek, representing the largest natural topminnow habitat. No exotic fish are present. North Fork Ash 126 San Carlos Topminnows were found at "North Fork of Ash Creek approximately 3/4 mile south of Ash Creek Apache Indian Creek Ranch" in July of 1985 (Jennings 1987). No topminnows have been collected since. Reservation Mosquitofish are present. Bylas Spring 6/7 San Carlos Discovered in 1968, invaded by mosquitofish in 1978-79. Renovated for mosquitofish in 1982 & Apache Indian 1984. 93% mosquitofish dominance in 1991 (Bagley et al. 1991; Brown & Abarca 1992). Reservation

• April-1993 Draft page 47

Table 3. Summary of successful, reintroduced populations of Gila topminnow, Poeciliopsis occidentalis occidentalis, in the United States, as of April 1993. Site number corresponds with Simons' (1987) System.

Location Site # Year stocked Comments AD Wash 242 1993 Desert pupfishes were also stocked. Big Spring 84 1985 Small population of topminnows. Cave Creek 49b pre 1989 Topminnows were stocked here prior to 1965 but, were eliminated by flooding. This population may be the result of topminnow dispersing from Seven Springs.

Cold Springs 85 1985 Only one of the two ponds has topminnow. Desert pupfishes are also present. Cow Creek 72 1981 Few topminnows were found following floods in 1991. Longfin dace and green sunfish are present.

Dutchman Grave 19 1983 Large topminnow population. Spring Heron Spring 76 1981, 1987 Small population of topminnows. Hidden Water Spring 48 1976, 1981 Topminnows persist with longfin dace and leopard frogs. Johnson Wash Spring 35 1982 Small population limited by habitat size and encroaching vegetation. ICayler Spring 42 1982 Small population exists with longfin dace, red shiner, green sunfish, and crayfish. Mosquitofish wiped out by 1991 flooding.

Lower Mine Spring 12 1983 Large population with encroaching vegetation. Mescal Warm Spring 82 1985 Large topminnow population.

Mud Spring 18 1982 Entire population is in small cement watering trough. Salt Creek 8 1986 Once supporting a natural topminnow population, the introduced population is also threatened by No topminnows have been collected since 1989. mosquitofish. - April-1993 Draft page 48

Table 3. Continuation.

Location Site # Year stocked Comments Seven Spring 49 1964, 1975, Topminnows coexist with longfin dace. The former were restocked following severe flooding in 1980 1978. Tule Creek 75 1968, 1981 Topminnows had to be restocked following flooding in 1978. Unnamed drainage 68b 1986 Topminnows washed down from Mesquite Tank #2 and a small population was present in 1991. Walnut Spring 20 1982 Large population of topminnows. Watson Wash 134 ? An undocumented stocking resulted in this topminnow population which coexists with red shiner and guppies. BLM is seeking to have the population removed. Yerba Mansa 44 1984, 1985, Out of historic range. Site does not count for recovery (Bagley et al. 1991, Brown and Abarca 1988 1992). April-1993 Draft. page 49

Table 4. Summary of all known introductions of Gila topminnow (Poeciliopsis occidentalis occidentalis) in Arizona, as of April, 1993 (updated and modified from Bagley et al. 1991). This information is taken from the AGFD Native Fish Database and listed here in alphabetical order by site name. Fields are defined as follows: Site Name and Location = commonly accepted name which refers to a particular site. Site EQ, = an arbitrary number unique to each site (following Simons 1987, Bagley et al. 1991, and Brown and Abarca 1992). Date Stocked = date (in format of YR+MO+DY) site was stocked. E = number of fish stocked, U indicates an unreported number of fish. Source of Fish Stocked = place where the stocked fish came from. Origin of Fish Stocked = indicates which natural population the stocked fish originally came from. Township-Range- Section = legal description for a site. Latitude-Longitude = latitudinal and longitudinal coordinates for a site. Extant = indicates current status, if known, of the site: Y=yes, N=no, U=unknown. Authority = source of the stocking information.

RITE DATE POPULATION SOUICE OF ORIGIN OF TOWNSHIP. RANGE LATITUDE SITE NAME ARID LOCATION NA. STOCKED TYPE ASH STOCKED FISH STOCKED AND SECTION LONGITUDE EXTANT AUTHORITY

AD WASH 242 930306 WILD 600 DEXTER SHARP SPRING 874 2W 38 335908 1122530 I' STOCKING SLIP AGFD PONDS PHOENIX 108 6904XX CAPTIVE MONKEY SPRING MONKEY SPRING N MINCKLEY AI BROOKS 1966 AOFD MESA 101 1370202 CAPTIVE 10 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AGFD FILES AGFO PHOENIX 161 1300201 CAPTIVE BOYCE-THOMPSON MONKEY COCK/ BYLAS SPRINGS N ACIFD FILES A0F0 PHOENIX 161 690710 CAPTIVE so ROPER LX ST PS MIDDLE SPRING N STOCKING SUP MOO PHOENIX 161 900624 CAPTIVE ENGEL-WILSON MIDDLE SPRING N STOCKING SUP WFD PHOENIX 161 111)0IXX CAPTIVE 2 AGM PHOENIX MIDDLE SPRING N AGFD FILES ALAMBRE TANK oo 8201114 WILD 200 DEXTER MONKEY SPRING 136 17E a 3211308 1103820 N BROOKS 1996 ANTELOPE POND WATER CATCHMENT 263 920621 WILD BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N A0F0 FILES APACHE CANYON EAST FORK 178 780114 WILD 200 VAUGHT POND MONKEY COCK/ BYLAS SPRINGS 203 11E 36 313318 1111139 N STOCKING OUP MWCKLEY & BROOKS 1986 ARAVA/PA CREEK 177 87XXXX WILD MONKEY SPRING MONKEY SPRING N MINCKLEY & BROOKS 1906 ARAVAIPA CREEK 177 87XXXX WILD MONKEY SPRING MONKEY SPRING N MINCKLEY B BROOKS 1086 ARAVAIPA CREEK 177 773000( WILD BOYCE-THOMPSON MONKEY COCK) BOAS SPRINGS N STOCKING SUP. MINCKLEY • BROOKS 1066 AREA R SPRING 264 82XXXX WILD BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AOKI FILES AREA W WATER CATCHMENT 266 9230001 WILD BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AGFD FILES AREA V WATER CATCHMENT 2613 82XXXX WILD BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AOFD FILES ARIVACA CREEK 272 38)000C WILD 313422 1111930 N MINCKLEY AND BROOKS 1966 ARIZONA HISTORICAL SOCIETY TUCSON 138 1370818 CAPTIVE 20 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AGFD FILES ARIZONA MUSEUM OF SCIENCE & TECHNOLOGY 162 800723 CAPTIVE 16 ROPER LX ST PS MIDDLE SPRING N STOCKING SUP ARIZONA STATE UNIVERSITY TEMPE 102 920327 CAPTIVE 167 BYLAS SPRING BYLAS SPRING U MEFFE 11183 ARIZONA STATE UNIVERSITY TEMPE 102 8404XX CAPTIVE 400 BYLAS SPRING BYLAS SPRING U BROOKS 19811 ARIZONA STATE UNIVERSITY TEMPE 102 8501302 CAPTIVE 160 TULE CREEK MONKEY COCK) BYLAS SPRINGS N STOCKING SUP. BROOKS 1966 ARIZONA STATE UNIVERSITY TEMPE 102 960802 CAPTIVE so SHARP SPRING SHARP SPRING N STOCKING SUP. BROOKS 1985 ARIZONA STATE UNIVERSITY TEMPE 102 900422 CAPTIVE so BYLAS SPRING BTUS SPRING N STOCKING SUP ARIZONA-SONORA DESERT MUSEUM TUCSON 137 861112 CAPTIVE YELLOWSTONE TIC MONKEY SPRING AGFD FILES ARTESIAN WELL 13 40 1320808 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 914 1 I E 336252 1111610 N BROOKS 1986. 44250 FILES ARTESIAN WELL F 4 70 820806 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON 1 1E 336260 1111606 N BROOKS 1606. SASOCOMARI RIVER 273 88XXXX WILD 313933 1103125 U AOFD FILES BADGER SPRINGS 1130 750816 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ION 2524 U STOCKING SUP MINCKLEY B BROOKS 1986 BAIN SPRING 20 030002 WILD 600 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ION 2W 6 341439 1123012 N BROOKS 19136. AGA) FILES SEAR CANYON 102 920917 WILD 2000 BOYCE-THOMPSON MONKEY COCK) BOAS SPRINGS 235 17E36 312260 1102146 N BROOKS 1985 BENCH WELL 87 830828 WILD 100 BOYCE-THOMPSON MONKEY COCIOBYLAS SPRINGS ION IE 23 341114 1121323 N STOCKING SUP. BROOKS 1986 BIG SPRING 84 850722 WILD 600 DEXTER MONKEY SPRING SS 25E 5 If STOCKING SUP. BROOKS 1985 BLACKTAIL POND 293 820404 WILD 76 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AGRI FILES SLACKTAIL POND 293 8201318 WILD BOYCE-THOMPSON MONKEY COCK) WILAS SPRINGS N AGFD FILES ELM KWOMAN 149 1E171104 CAPTIVE 106 DEXTER SHARP SPRING N STOCKING REPORT WA SAFFORD 160 831228 CAPTIVE BOYCE-THOMPSON MONKEY COCK/ BYLAS SPRINGS N AGFO FILES SLUE 14171. SPRING 244 1320610 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS WI SE 19 336044 1114222 N BROOKS 19136. BOSTON WATER CATCIWPIT 267 920619 WILD BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AGED FILES - BOYCE-THOMPSON moons rum ED 7108XX CAPTIVE 4000 PAGE SPRINGS MONKEY SPRING 2S 12E M1NCKLEY • BROOKS 1996. AGRI FILES BOYCE-THOMPSON ARBORETUM BO 7201173 CAPTIVE COCK) WASH COCK/ WASH 28 1250 Y AGFO FILES SOYCE-1140MPSON ARBORETUM ED 70 PRE CAPTIVE BYLAS SPRING BYLAS SPRING 2S 1256 AGFD FILES e ! s a g m 1! ! ! !! I !f ! H! f ! e .- 2 a as s. ggggal e 5 0 1 1 i5 ;;;;12 I /w Itgm 1 4 i1 g -.1 x2111 0 ix msmOmi 110/1 glx0

>>>Z ZZZZZZZZ 2Z2M7 ZZZZZZZ ZZZZUZ>ZZZ=7.Z>Z2 ZZZZZZZZ = ZZZZZ Z223232

"8 1 I I ! --= ! - 1;7.= liE ill IR E s °R *RR . X2 1 RRs R R

ft. R. • ..! X zzz g... w N X.R g ! gR WAA w:- A N A WARA 44:2 A5x 5 IKRA2t 1! 14

IH WI 1 Hill 33 333 3 333! ! 313333 3 33! 33331 3 35 ! 1 3 ;;!;;; 3;;;; ;v;2;;;; ;1 12;;;;;;;;2 1;;Ig2 ; 22; §§g§§§ g§§§§ §g§§§§§§ §g §z§g§ 1 22222 hAAW V2222222 We'n 11111i 111111111ifilliiiii Mill hill

111 11;11 1 41111 . 111.11 11 ..1 111111 ; !!!!! P!!!!!!1 ! 1.!!!!!!°! "-;11!

9 ttt lg ttttttt M95999U99599999999g9999994959999959MM&2959595552555 uuu MiA3zzXXXX333333RuA333333333333333Auuuuuuu333333333333u3

FeWINITIS ! ZZZ ssssIgs::--g§sgUER!!;,A gg?2sissgg;EgsggigsgggsselEilEgIU; iss

ia i 4 x u g g

!IV 111 1 ''' ;g g g ' 1111 iii 2 4 Ow g II EE 4, g iililiggl Ogg. 2.0 ; 1680 11 8° gi in I.. gip ime88 iiiii mligiola:ggvAlamili i ilg ,dvilloAl - iii iiiilhIlm.glghs"Wu l'ilil" .. gsii Piigl' 8 888ggil 111111 "'"irliilMaid§ 1 8 gg g 22 I I I i igi i I :22 111 !RR ;; ! 1 1 11 1 1 01! ll 2 m i l" i i • dilifi• 4 a a : 5 22 g:R i II4ii iii i i M101111;151 li iii M il 2 1 mliglg :Ili 1 :g2:142 :m gl gg: 51 WWW:11 1 :g. 11 ii iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii

Z ZZZZZ Z>ZZ73.P.P.I.ZZZZZ ZZZZZZ 32 ZZZZZZ 2>=ZZZ ZZZZZZZZ 22>>Z322222=2

R f5/ 11 IFf 11 1 =v. gr. IS i 335g 1 5 01 U. •5 1E 2A E ERii g liE II ZRE g5gaEiRRig Xi

11 g x.R g *4:gx0X 1 g::;!--42g .5;;w 1i 5 W:WA0 ;01501 M 2 12101 ! 5 ii§!/ 5111 W 5-47.Eth 2 2i

40) iiiiiiiii 3 3 3 33331 133333 3 33S 33 133 333333333 33

00 flili Illiiiiiii111111111 111111111111111 &111 55555 lig5 8 555555555555 5 555g 55 55StIvarltivdt550555ff555 -1111i istlii1111111111111 -1.1111-11,11111111111110,11111111

, 1 re 1 1.111 11 111/ IfIL If 111 11114111 .011011 !!!!!d!! g 1! S WI!! U QUO 00 si 222 2 2228 2222 222

1 Iv 1ww 5333m959MOssgssses9s959.5sesgss599s9ss559sessss5s9sessM 2' 3333o0ooRXXXXXXAXXXXXXXXX33XXXXXXXXXXXX33333333305333ouuou

S kizag 8 !V 3 m E was SLAGGA xgaii , 1310,4 222EI gag $ gas IGE!R agaRgIR 21/

thig ARR 10 111 g Wig 4C§„ 1111 ;2- .. ggill 1 IMmem4 . 4, AffpflOggg 411 1 1 It gli Ax i2 34 o Llu x lug X;g;;IPiiii ig2/1111 "il /11. 111 WW21

MHOMMIHffidLttlinHaridiliniihninh" • 111

a

12 2 2 I. 12 11 222 . 1.2/112 1 1111 001101 iliii121q15551 11111111111111111111 ifflml; iiiii i ZZZZZZ M2222=22=2=22>02==2=272==>======7=M722=>22220270Z2

554 Iiiiiiii ii !gg!ssg! !!!!! 1 113 3 4 44 ;2 ; 11111111 11111 § 6§§§t 11 11 1 1 111 11 § "1 "SI S I SSS 1 S 'I S MS P iS S S SSS SS S 11111ili Mil i 1iiii .. ILA i i iii ii .. i

11111 111.1 111 w l I I § 3 22 22 2 20 8 .G10.01?TF .G..01.R1....11111GYS 0..0§.GAGSGISSG s OGGSAR. :111

tWV 999 M99 333 0UWi3UON NNNNNNNN N NNNNNNNNNNNNNNNN NN NNNNNNNNN NNNNNNNNNNNNNN

x VRO IPLGG.Tc .M.TIGG§G050 53GT!GIGG4ITALIIGG:I'M .T WXV:A§1 550 5

SIS

x li 1 Flamer n 1. 41055 Irt „ A2iM T Weh..51 i122 1111;1111g111111giiiiiii i 1"mg :dli!Mil5115/1U .1422ii2222222i1f2iiii Minglig1iIiiffgfIgffffffff ffflIffffIfIfffIffifffl Ifff ff misgE. is wgwgggwwwgww wwwwgwwwggggwwgwgggrogwwggoggwgggwg HIM g I 1 1 1 ernes s i i 1 i I mii I 1 1 1 1 lititit i ff. ! I - t 10;laael !Mai 5i Milli' 1 1 0000000 v . 12 ...... 2 22 inl 2 ^2 mii2 i igi tttIttrit..tt mt tt IF 1 111 11 1 1 1 /11 iM555555P1 /:1 1iigli:i111111i055111111111142

=232*=7=2222M=7==2332=Z=7Z2=3ZZZZZ2222).).= 222 2222.22=22223222

lb r. Hill Al Z ZZZ f f! IllsA A XE Ei

A AI R - XR R AA A N% ;WW 5 la as 5 i! 5 xxF

lb 111 1 NI 1 II

C 1 11C. 1CC1 ! llmI 22 9 29 111ili ttt 111> l ill h,S,SS SSS S 40 ssi t* S MSr.ASSnn L H šs1.111111111 11111.11 HHHH 111

1 111"41 I f 1 " "1/p1111, 1/1111/1111 111 If II! 11111141 A I I!

g ol a 23 lo 3 32§83 3 3 3 333 ool= m3=Fo 3 ooloo 3 $$4 1o 3 pol oolo.ol.11

WW Vt“ 99gg99599999g M9S9999999999999999 330c233333333ouuuuuXXXXXAXXXX33333XXX

reqsp R zsOgs R . ggUl v LI pirrn OLGIV sA—N— ASG p ;ERM:4 7.11

U VE224R2aRaaREIRWWWS03011x2201 MF ELL

m 0 1 q Piri2 iiqgill rinli50 I? ii i 111 iiilloolliiiiguisseggali§ " 111111111111111111111111 01 1 li 1°1111 M gggggggggggggggggggggWirtgb 5111,1 081221 9 1i April-1993 Draft page 54 SITE DATE POPULATION SOURCE OF MEMO OF TOWIMHIP, IIHME LATITUDE SITE NAME AND LOCATION No. STOCKED TYPE N FISH STOCKED FISH STOCKED AND SECTION LONGITUDE EXTANT AUTHORITY

TANK RANGE SPRMG AREA 10 285 820421 WILD 50 BOYCE-THOMPSON MONKEY COCIO BYLAS SPRINGS N AGFD FILES TANK RANGE SPRING AREA 10 286 920428 WILD 50 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AOFD FILES THE LAKE 121 820814 WILD 200 DEXTER MONKEY SPRING 138 1 7E 321936 1103715 N BROOKS 1905 DOCKET SPRING 15 930603 WILD 1000 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 1014 6E 36 341148 1114820 N BROOKS 1996 10110110 CHUL PARK TUCSON 138 970721 CAPTIVE U AZ-SON DES MUS MONKEY SPRING AOFD FILES TRES ALAMOS 121 38 840710 WILD 1000 TULE CREEK MONKEY COCK) BYLAS SPRINGS 1014 OW 13 N BROOKS 1998 TUCKER BOX 78 020810 WILD 000 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON 13E 20 334006 1110236 N BROOKS 1906. 1ULE CREEK 76 88XXXX WILD 1000 MONKEY & B-TH MONKEY SPRING • ON 1( 26 340020 1121910 N MINCKLEY • BROOKS 1995 TULE CREEK 75 810930 YAW U BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON 1E 28 340020 1121818 Y AGFD FILES YULE CREEK SEEP 12(2 73 82300C( WILD U TULE CREEK MONKEY COCK) BYLAS SPRINGS SN 1( 26 340009 1121548 N BROOKS 1098. BROOKS PERS. COARA. 1989. THE CREEK UNN. SPRING 11E1 74 82)000( WILD U TULE CREEK MONKEY COCK) BYLAS SPRINGS ON IE 28 340013 1121565 N BROOKS 1998 TURKEY CREEK 97 98XXXX WILD U ? 215 ISE 33 U AOFD FILES TWO MILE SPRING 13 930803 WILD 200 BOYCE-THOMPSON MONKEY/COCIMBYLAS SPRINGS ON SE 28 340517 1114413 N BROOKS 1996. UNIVERSITY OF ARIZONA TUCSON 174 020813 CAPTIVE U BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS N AGM) FILES UNIVERSITY OF ARIZONA TUCSON 174 010822 CAPTIVE U CIENEGA CREEK CIENEOA CREEK Y AGFD FILES UNNAMED ITT.) SPRING 145 820900 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS MEN BE 24 341126 1114730 N moms 1086. UNNAMED (LT.) SPRING 14A 83XXXX WILD U T.T. SPRING MONKEY COCK) BYLAS SPRINGS SAN 56 24 341125 1114730 N BROOKS 19116. UNNAMED DRAINAGE /80 MS 921095 WILD U MESQUITE TX #2 MONKEY COCK) BYLAS SPRINGS MI OE 1 V BROOKS 1990 UNNAMED SPRING AI 1 34 820804 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON OE 21 335120 1112858 N BROOKS 1985. UNNAMED SPRING It 175 820804 WILD MO BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON OE 21 336120 1112846 N BROOKS 1986. UNNAMED SPRING #2 29 030801 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 16N 3E 18 344122 1120209 N BROOKS 1986. AGFD FILES UNNAMED SPRING #3 37 820803 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 88 11E 2 334249 1111223 N BROOKS 1085. UNNAMED SPRING *4 50 820809 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS PAI N 24 334406 1113620 N BROOKS 1986. UNNAMED SPRING #5 06 930802 WILD 500 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS SAN 6E 32 340945 1115216 N BROOKS 1986. UNNAMED SPRING #0 ea 820809 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 10N 6534 341160 1114846 N BROOKS 1986. UNNAMED SP1I40 #7 100 030803 WED 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 7I1 10E 4 336900 1112046 N BROOKS 11/85. UNNAMED SPRING O IN 1E SEC. 23 148 830801 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 11N 1E 2 341851 1121263 N BROOKS 1996. UNNAMED SPRING FED TANK • 400 17C 920004 WILD 300 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS SN SE 21 335120 1112835 N BROOKS 1995. UNNAMED SPRING TANK • 498 39 820008 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON 106 2 334840 1111830 N BROOKS 1906. UPPER HORNELL SPRING 32 830803 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 211 I 2E 12 333127 1110516 N BROOKS 1996 MEWS OFFICE ALBUOUEROUE 144 890800 CAPTIVE 100 DEXTER SHARP SPRING N MVO FILES DEXTER FILES UBFWEI OFFICE SAN CARLOS 9109)01 CAPTIVE 13 SALT CREEK MIDDLE SPRING Y AOFD FILES VAUOHT POND 1,1 188 760811 CAPTIVE 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS U STOCKING SUP. AGFD FILES. VERDE LAKES 287 730603 WILD SOO ? U AGFD FILES VERDE mu. NEAR PERKINSVILLE 289 77 PRE WILD U r N AGM) FILES WALNUT SPRING zo 820804 WILD 1000 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS ON BE 3 335361 1113110 Y moon 1905 WARM SPRINGS, UTTLE BLUE RIVER 2119 (18)000( WILD U ? U NOD FILES WNW SPRINGS, SAN CARLOS RIVER 290 MIXXXX WILD U ? U AOFD FILES WATER CATCHMENT AREA 11 288 920611 WILD 250 BOYCE-THOMPSON MONKEY COCK) MAO SPRINGS N AGM) FILES WATER CATCHMENT AREA 14 287 720915 WILD 500 BOYCE-THOMPSON MONKEY COCK) SPRINGS N AGFD FILES MINCKLEY & BROOKS 1986 WATER CATCHMENT AREA 2 206 720916 WILD 250 BOYCE-THOMPSON MONKEY COCK) SPRINGS N AGFD FILES MINCKLEY 6 BROOKS 1905 WATER CATCHMENT AREA 8 281/ 720015 WILD 1600 BOYCE-THOMPSON MONKEY COCK) SPRINGS N AOFD FILES MINCKLEY & BROOKS 1986 WATER CATCHMENT AREA 9 284 720916 WILD 250 BOYCE-THOMPSON MONKEY COCK) SPRINGS N AGFD FILES MINCKLEY 6 BROOKS 1986 WATSON LAKE 291 700828 WILD 1380 PAGE SPRINGS MONKEY SPRING N HINCKLEY IND BROOKS 1995; AOFD FILES WATSON WASH 134 941089 WILD U ? 08 26€ 23 Y AGED FILES WAYNE STATE UNIVERSITY DETROIT 78 PRE CAPTIVE U UNKNOWN UNKNOWN U AOFD FILES WHITE ROCK SPRING 249 820009 WILD 200 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS IN BE 12 340765 1114720 N BROOKS 1086. WHITE TANK 230 920814 WILD 200 DEXTER MONKEY SPRING 135 1 7E 14 321802 1103435 N BROOKS 1986 WHIM TANK #I 89 830801 WILD 1000 BOYCE-THOMPSON MONKEY COCK) BYLAS SPRINGS 1 IN 1( 11 341021 1121303 N BROOKS 1986. WILLOW CREEK RESERVOIR 292 700828 WILD 3090 PAGE SPRINGS MONKEY SPRMO N MINCKLEY AND BROOKS 1996; AOFD FILES WINDMILL POND #1 270 720915 WILD SOO BOYCE-THOMPSON MONKEY COCK) SPRINGS N AOFD FILES MINCKLEY • BROOKS 1996 WINDMIU. POND f2 271 720915 WILD SOO BOYCE-THOMPSON MONKEY COCK) SPRINGS N AOFD FILES MINCKLEY 6 BROOKS 1985 YELLOWSTONE TANK 23 1120814 WILD 200 DEXTER MONKEY SPRING 135 17E 20 321722 1103600 N BROOKS 1985. AGED FIEB VERSA MANSA 0 44 841220 WILD 260 TULE CREEK MONKEY COCK) BYLAS SPRINGS 1IN 11W 21 V BROOKS 1988. AGED FILES VERSA MAMBA (21 44 850629 WILD SOO RILE CREEK MONKEY COCK) BYLAS SPRINGS 11N 11W 21 Y BROOKS 1980. STOCKING SUP VERSA MANSA fa 44 880009 WILD 260 DEXTER SHARP SPRING 11N 11W 21 Y AGFD FILES ZM ZAO SPRING 148 83 PRE WILD U BOYCE-THOMPSON MONKEY COCK) OVUM SIMMS 9.5N BE 25 341041 1114731 N AOFD FILES

NOTES: I. BYLAS SPRING HAS BEEN RENOVATED FOR THE REMOVAL OF MOSIMITOFISH TWICE AND RESTOCKED WITH BYLAS SPRING TOPMINNOW, THEREFORE IS A NATURAL POPULATION. 2. 8111 10 OUT OF HISTORIC RANGE FOR GILA TOPMEMOW AND DOES NOT COUNT TOWARDS RECOVERY.