Wild Horse Populations: Field Studies in Genetics and Fertility: Report to the Bureau of Land Management, U.S. Department of the Interior (1991)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Research and Results The Bureau of Land Management (BLM) awarded research grants in 1985 for the study of wild horse genetics and fertility control. The University of California at Davis conducted its 3-year study on wild horse population genetics and submitted a final report to BLM on January 15, 1988 (Bowling and Touchberry, 1988). The University of Minnesota began its study of wild horse fertility control on October 1, 1985, and will complete it in the fall of 1990. Although the results from one remaining field season in the fertility study are not included in this report, the data gathered to date have been extensively analyzed and permit the conclusions presented in Chapter 5. GENETICS STUDIES Research on wild horse parentage and population genetics included a genetic analysis of the inheritance of different proteins present in red blood cells and serum. From December 1985 through October 1986 researchers from the University of California at Davis collected blood samples from nearly 1,000 horses at seven trap sites in the Great Basin area of Oregon and Nevada. The genetics studies on the free-ranging feral horses had the following four objectives: 1. Assess average and individual heterozygosity in the populations to determine if there has been loss of heterozygosity or inbreeding through genetic drift, selection, removals, or management restrictions on animal movement; 2. Estimate the contributions of the original wild mustangs (descendants of animals released by the Spanish) and the current domestic lineages (13 breeds) to the present feral horse populations;

4 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY 3. Evaluate the several populations for possible divergence in gene fre- quencies and for the development of population substructure; and 4. Determine parentage and particularly paternity within bands to evaluate the proportion of foals sired by the dominant band stallion. An important feature of horse social organization and behavior is year- round bands. Bands generally consist of females, their young, and a single stallion. Bands with more than one male do exist, but are generally short lived, lasting from several hours to several months. Bachelor males also form bands, although older horses, more than 14 years old, spend signifi- cantly more time alone than younger males, particularly those 2 to 5 years of age. Bands can contain from one to eight (rarely more) females, with an average of between three and four mares per band. Females, however, do not band together for life, and will stray from one band to another. Interestingly, Berger, in his 5-year study of wild horses, reports never seeing a mare driven from a band by a stallion (Berger, 1986). Methods Used The assays used included red cell antigens (50 alloantigens at 7 loci), isoenzyme (red cell and serum isozymes), and serum protein electrophoresis (76 alleles at 12 loci) for a total of 19 polymorphic loci. Loci known to be polymorphic in domestic horses were chosen for analysis. Selection was based on the objective of measuring the occurrence and frequencies of alleles for a comparison with the data available on many domestic horse lineages. This approach does not evaluate polymorphisms or rare alleles at other loci. It detects new alleles at the studied loci. Observations on coat color and pattern variants at seven loci were also collected. Blood samples were collected from 975 horses in five populations. Two sites had two trap locations each, providing paired subpopulations; the other three sites had one trap site each (Flanigan, 175 animals in 30 bands; Wassuks, 119 in 21 bands; Beaty Butte, 112 in 17 bands; Stone Cabin, 239 in 30 bands at site 1 and 127 at site 2; and Clan Alpine, 104 in 17 bands at site 1 and 99 in 12 bands at site 2). Separation of the locations by distance and geography makes it unlikely that genetic exchange occurred between these populations. The band compositions and paternity assignments may have been compromised in two cases: (1) a roundup and removal of some horses was conducted by BLM at the Flanigan study area in September 1985 prior to the sample collection in December; and (2) samples were collected at the Beaty Butte area in February 1986 after a BLM reduction roundup in November 1984. Data on the domestic breeds were drawn from blood samples from a breed registry laboratory; they ranged from 79 to 14,517 samples per breed. Thirteen domestic breeds are thought to be associated with the horses of the Great Basin (Figure 2-1). They include the Arabian, Criollo, and Mangalarga breeds.

RESEARCH AND RESULTS FIGURE 2-1 The Great Basin of the western United States. Results Data were analyzed by calculating allele frequencies at individual loci, proportion of heterozygous loci in individuals and in populations, Nei's genetic distance, and dendrograms. Parentage was accepted on the basis of congruent allelic specificity at each locus for foal and dam pairs and with the band stallions. The numeric results are summarized in Tables 2-1, 2-2, and 2-3, and in Figure 2-2. They indicate the following: • There were no significant deviations from Hardy-Weinberg equilib- rium proportions indicating that the horses sampled at each site were mem- bers of an interbreeding population within the limits of the 19 loci examined. • The horses from the paired sites within the Stone Cabin and Clan Alpine sites could each have been members of a single randomly mating population. • The number of effective alleles for wild horses averaged 41.3 ± 2.8 (range 38.8 to 46.3) and for domestic breeds averaged 40.3 ± 4.0 (range 33.7 to 46.8). • The average heterozygosity was 0.402 ± 0.009 and 0.353 ± 0.011 for wild and domestic horses, respectively. • The differences between populations of Great Basin horses were less than between breeds of domestic horses, based upon Nei's population mea- sures. • Unique variants in the wild horses were observed only at the highly polymorphic Pi locus. • The above data and the dendrograms support the hypothesis that Great Basin horses originated from escaped or released domestic draft, saddle, and cavalry animals. • Paternity assignments included 121 foals from 69 intact harem bands. The data indicate that about one-third of the foals were not sired by the harem stallions. This exclusion rate did not change when the data from the sites disturbed by roundups were excluded.

6 WILD HORSES: FIELD STUDIES IN GENETICS AND FEXTJLITY TABLE 2-1 Allelic Variation in Horses at 19 Blood Type Loci Number of Alleles Number of Effective Alleles Average Heterozygosity (Standard Error) Trap Site or Domestic Breed Blood Group Protein Total Trap site, wild and free-roaming horses Flanigan 30 41 71 38.3 0.378 (±0.058) Wassulcs 24 33 57 38.8 0.380 (±0.062) Beaty Butte 25 36 61 39.7 0.442 (±0.045) Stone Cabin 1 30 48 78 46.3 0.423 (±0.062) Stone Cabin 2 28 38 66 43.0 0.416 (±0.059) Clan Alpine 1 31 5I 82 42.5 0.404 (±0.058) Clan Alpine 2 28 45 73 40.5 0.368 (±0.061) Domestic breed Arabian 29 37 68 37.1 0.346 (±0.062) Thoroughbred 24 34 58 33.7 0.295 (±0.060) Quarter horse 36 50 86 45.0 0.403 (±0.062) Standardbred 29 40 69 39.1 0.413 (±0.048) Morgan horse 34 51 85 42.9 0.410 (±0.056) American 34 47 81 40.7 0.386 (±0.059) saddlebred Tennessee 29 42 71 35.8 0.350 (±0.056) walking horse Belgian 31 42 73 46.8 0.443 (±0.059) Shire horse 28 43 71 38.3 0.381 (±0.058) Argentine 29 46 75 41.1 0.410 (±0.061) Criollo Chilean 30 44 74 41.8 0.428 (±0.052) Criollo Mangalarga 26 38 64 36.8 0.305 (±0.068) Mangalarga 30 48 78 45.4 0.412 (±0.066) marchador SOURCE: Bowling, A. T., and R. W. Touchberry. 1988. Wild horse parentage and population genetics. Final research report to the U.S. Department of Interior, Bu- reau of Land Management, January 15. University of California at Davis. Photo- copy.

RESEARCH AND RESULTS TABLE 2-2 Net's Population Measures for Horses Population Cluster" Wild Domestic Total Heterozygosity Total 0.429 0.443 0.441 Subpopulation 0.402 0.383 0.389 Diversity In terpopul ational 0.027 0.060 0.052 Subpopulations 0.063 0.136 0.117 relative to total population "Population number is 7 wild and 13 domestic horses. SOURCE: Bowling, A. T., and R. W. Touchbeny. 1988. Wild horse parentage and population genetics. Final research report to the U.S. Department of Interior, Bureau of Land Management, January 15. University of California at Davis. Photocopy. TABLE 2-3 Foal Paternity in Harem Bands Trap Site Per- Per- TypeofBand F W B SCI CA1 CA2 Total cent Total" cent Harem bands 12 10 12 20 8 7 69 45 Bands, 1 stallion 9 4 10 12 6 1 42 23 Excluded for all foals 3 0 4 4 1 0 12 29 5 22 Sire of all 6 4 3 6 4 0 23 56 14 61 Sire of some, not all 0 0 3 2 1 1 7 26 4 27 Bands, 2 or more stallions 3 6 2 8 2 6 27 22 Excluded for all foals 2 2 0 0 0 2 6 22 4 18 One is sire of all 1 3 2 4 2 2 14 52 11 50 One is sire of some 0 1 0 4 0 2 7 25 7 32 Foals in band with stallion 18 11 23 40 16 13 121 80 Without qualifying sire 7 3 9 I5 2 4 40 33 24 30 NOTE: Study area designations are F, Flanigan; W, Wassuks; B, Beaty Butte; SCI, Stone Cabin/site 1; CA1, Clan Alpine/site 1; CA2, Clan Alpine/site 2. Stone Cabin/ site 1 was located in west central Nevada, Nye County. Clan Alpine/site 1 was located in west central Nevada, Churchill County; site 2 was located in the same area, about 10 miles away. "Considering only W, SCI, CA1, and CA2 sites for which no recent Bureau of Land Management roundups had occurred.

8 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY S3 O CO 1 S-s *p VI o -g Q

RESEARCH AND RESULTS 9 In summary, the genetic data support the domestic origin of these Great Basin horse populations, the lack of differentiation within or between these populations, and their likely origin from a substantial number of founders, with no evidence of a bottleneck or loss of heterozygosity. Because of their substantial sample sizes and high levels of polymorphisms, these studies are very robust and allow substantial confidence in the interpretations. The interpretations, however, apply only to the populations sampled. It is pos- sible that there are other populations that have experienced a long history of isolation, small population sizes, or selective origins. The methodology and data base are now available to allow testing of any such proposed popula- tion. MARE FERTILITY CONTROL STUDIES AT LOVELOCK CORRALS The overall goal of the fertility control research has been to develop a method to block reproduction in wild and free-roaming horses that would be effective over several years. The research involved vasectomizing dominant band stallions or using steroid hormones to block pregnancy in mares. The research was initiated in the fall of 1985 with a corral study to develop methodology that could be used in the field in 1986 with wild and free- roaming animals. The objectives of the corral studies in Lovelock, Nevada, were to develop dosage levels and techniques for administering steroids to captive wild mares. Such dosages and procedures had not previously been developed in horses, and were needed before the research team could administer steroids to wild and free-roaming mares in the field. These studies at the Lovelock corrals were conducted by researchers from the University of Minnesota. The relative merits of surgical implants versus remote delivery of steroids were considered. The advantage of remote delivery is that, operationally, the steroids could be administered in the field with darts fired from helicopters without subjecting the mares to capture, anesthesia, and surgical implant. However, the decision to use implants was made for the following reasons: • The implantation procedure has been used with other animals, includ- ing humans, with success in multiyear fertility control. • At the time research began in the fall of 1985, no studies using remote delivery had achieved more than 1 year of fertility control, nor was any technology known that would allow multiyear dosage. • No dosage levels of remotely delivered female hormones were known. • It was necessary to capture the experimental animals in the field in order to determine their ages and to mark them for subsequent observation. Hence, capture and restraint were necessary for either implants or remote- delivery animals.

10 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY Methodology The first phase of the pen studies involved the development of silastic rods—silicone rods impregnated with progesterone or estradiol—to serve as the dosage delivery vehicle. The rate of release of the steroid was measured by testing this release periodically when the silastic rods were immersed in saline solution. The first set of steroid implants in corraled males was initiated in mid- November 1985 and involved administering six different treatments to 30 mares each. These treatments involved the subcutaneous insertion in the neck of rods containing one of the following dosages: • 8 g estradiol; • 24 g progesterone; • 8 g estradiol and 8 g progesterone; • 12 g progesterone and 4 g estradiol; • 12 g progesterone and 2 g estradiol; and • placebos containing no steroids. A seventh group of 30 mares was maintained as nonimplanted controls. A second experimental sequence on corralled horses was initiated in April 1986 involved the seven dosages listed below using one of the follow- ing methods: subcutaneous implants in the rump and flank areas, intermuscular implants, and intraperitoneal insertion. Ten mares were used in each of the following treatments: • 36 g progesterone and 24 g estradiol-l7p; • 36 g progesterone, 24 g estradiol-170 and glycerol; • 36 g progesterone and 12 g estradiol-17p; • 36 g progesterone and 12 g estradiol benzoate; • 36 g progesterone and 8 g ethinylestradiol; • 36 g progesterone and 8 g ethinylestradiol in methyl ether; and • 1 g levo norgestrel. The third series of treatments of corralled horses was run in January 1987 involving 50 mares for the following five treatments: • 10 pregnant mares given 8 g ethinylestradiol; • 10 pregnant mares given 3 g ethinylestradiol and 36 g progesterone; • 10 open mares given 1.5 g ethinylestradiol; • 10 open mares given 36 g progesterone and 8 g ethinylestradiol; and • 10 open mares given 8 g ethinylestradiol. Dr. Gerald M. Peck, the veterinarian under contract to the BLM during the Lovelock studies, was present during the surgical work to oversee the procedures.

RESEARCH AND RESULTS 11 Blood samples were taken at bi-weekly intervals from veins of the ex- perimental animals to assay the amount of hormone released by the implants. To indicate the length of time over which fertility was controlled, blood samples were taken from November 1985 through November 1988. Normally, ovulation in horses is followed by an elevation of serum progesterone greater than 1.5 ng/ml. This elevation persists for 10 to 14 days and can be de- tected through bi-weekly sampling. This correlation of ovulation and eleva- tion of serum progesterone has been validated by rectal palpations in domestic horses (Ginther, 1979) and, to a more limited extent, in feral horses (Wolfe et al., 1989). The failure to observe pregnancy despite elevated progesterone levels in some animals may reflect either the use of too rigorous a criterion to accept ovulation, or the fact that estrogens may exert a contraceptive effect at the level of the fallopian tube (delays migration of ovum or embryo into uterus) and of the uterus in antagonizing the progestational changes necessary for implantation. Results The first series of implants proved to be ineffective in preventing preg- nancy when the mares were exposed to stallions during their normal estrus periods. In addition, some of the mares rubbed the implants out of their necks. The second and third series proved more effective. • Implants containing 36 g of progesterone plus 8 g of ethinylestradiol or implants containing 36 g of progesterone plus 24 g of estradiol-l7p with glycerol as a cosolvent effectively blocked ovulation for two breeding sea- sons in more than 60 percent of the mares. • Implants containing 8 g of ethinylestradiol without progesterone were capable of blocking ovulation for at least one breeding season. • Implants containing 8 g of ethinylestradiol with or without 36 g of progesterone when implanted during mid-pregnancy allowed the successful completion of pregnancy and delivery of apparently normal foals. • Scarring around the implants may be a significant problem that could reduce the longevity and effectiveness of the implants. • Route of administration had no effect on the level of hormone achieved in the serum and consequently on the effectiveness of the implant. Pregnancy was blocked through two breeding seasons in more than 90 percent of the animals receiving 36 g progesterone plus 8 g ethinylestradiol, or those animals receiving 8 g ethinylestradiol alone. Pregnancy was blocked through the peak of the breeding season of the third year in more than 70 percent of the animals receiving 36 g progesterone plus 8 g ethinylestradiol or 36 g progesterone plus 24 g estradial-17(i and glycerol.

12 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY Over 80 percent contraceptive efficiency was achieved through the peak of the second breeding season in open mares that had been treated with 1.5 g ethinylestradiol alone or in pregnant mares that received 3 g ethinylestradiol plus 36 g progesterone. Moreover, the data show that ethinylestradiol at 1.5 g and 3.0 g blocks pregnancy without blocking ovulation, whereas 8.0 g of ethinylestradiol with or without 36 g of progesterone blocks ovulation as well as pregnancy. The data also suggest that 36 g of progesterone plus 24 g of estradiol-17p with glycerol blocks ovulation during the second year, but in the third year the treatment blocks pregnancy while allowing ovula- tion. More generally, the following conclusions were drawn: • Effective contraception can be achieved for at least 28 months. • The length of time that the contraceptive is effective beyond 28 months is not known. • The mechanism of contraceptive effect is not completely understood. Ovulation could be blocked at the higher concentrations of ethinylestradiol or estradiol with or without progesterone. Or, at lower concentrations of ethinylestradiol, the contraceptive effect might occur at some point after ovulation, such as fertilization or implantation. • The peritoneal cavity (intraperitoneal) is an efficient and effective location for routine implantation of hormonal contraceptives. The experimental animals were removed from the Lovelock corrals in 1989 and transported to a wild horse sanctuary in Oklahoma after the ste- roid levels in their blood had been monitored for about 2 years. They are a valuable resource for establishing the length of time over which the im- plants are effective. Modeling the cost-to-effectiveness ratio of contracep- tion versus roundup for horse population control depends on the number of years a treatment works. At present, the treatments are contraceptive up to 28 months, but it is not known how much longer they might remain effective. Hence, the committee recommends that annual monitoring of blood hor- monal levels continue until the levels are not significantly higher than the initial control levels. The cost of this monitoring would be low relative to the value of the results in determining the practical utility of contraception. The committee concludes that the pen studies at the Lovelock corrals achieved their objectives and were handled in a professional manner. These chcmosterility studies have led to the following publications in peer-reviewed literature: Plotka et al., 1987; Plotka et al., 1988a,b; and Plotka et al., 1989. At the time of its development, the intraperitoneal approach was on the cutting edge of a rapidly developing technology. MARE FERTILITY CONTROL STUDIES IN THE FIELD Steroid implants in the field constitute the second component of the mare fertility control experiments. The objective is to implant mares in selected

RESEARCH AND RESULTS 13 study areas with steroids and placebos, follow their reproductive histories over 2 to 3 years, and evaluate the effectiveness of the treatments. Research is expected to be completed by the fall of 1990. Methods Wassuks Study Area In January 1986, 41 mares, 3 years old and older, in the Wassuks area of west central Nevada were implanted with silastic placebo rods to serve as undosed comparison animals. These mares and those in the Clan Alpine area (described below) are not formal control animals. The horse populations in these areas are not large enough to support the formation of control groups and treatment groups. Therefore, separate or adjacent areas were used to compare treated and untreated mares. The 41 placebo-implanted mares in the Wassuks area were equipped with radio collars. An additional 62 mature male and female animals were fitted with marker collars. The marker collar animals at this and other sites were intended to be used to study band stability and structure. This portion of the study was eventually discontinued because of the lack of funds. Since many of these animals must have been pregnant from the 1985 breeding season, any possible blocking effect on future conception and pregnancy by the placebo would not be observable until the spring or summer of 1987. Stone Cabin Study Area In September 1986, 101 mature mares in the Stone Cabin area of south- central Nevada were implanted either with the 8 g ethinylestradiol plus 36 g progesterone (EP) combination (51 mares), or with 8 g ethinylestradiol (EE) only (50 mares). The 101 treated animals were equipped with radio collars. An additional 114 mature male and female animals were fitted with marker collars. Many of these animals were undoubtedly pregnant from the 1985 breed- ing season. Hence, any blockage of future pregnancies by the steroids would not occur until the 1987 breeding season and would not be observ- able until the spring and summer of 1988. Clan Alpine Study Area In September 1987, 100 mature mares in the Clan Alpine area of central Nevada were implanted either with the 8 g ethinylestradiol plus 36 g progesterone combination (50 mares), or with 8 g ethinylestradiol only (50 mares). Forty- nine mares were implanted with placebos to serve as untreated comparison animals. All 149 treated animals were equipped with radio collars. An

14 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY additional 152 adult male and female animals were fitted with marker col- lars. Many of these animals were undoubtedly pregnant from the 1987 breed- ing season. Hence, any blockage of pregnancy as a result of the steroids would not occur until the 1988 breeding season and only be observable by spring and summer 1989. Aerial Surveys Four aerial surveys by helicopter were made in each study area each spring between the months of April and June. The purpose of these flights was to observe the foaling rates of treated and placebo-implanted mares. Results The results of the 1988 and 1989 aerial surveys are summarized in Tables 2-4, 2-5, and 2-6. Table 2-4 presents the foaling rates in 1988 for each study area and Table 2-5 presents the foaling rates in 1989. Table 2-6 compares the data from each year. In the Clan Alpine area, where no treatment effects were expected until 1989, 97 of the 100 treated mares were observed in 1988 (Table 2-4). Of the 97 treated mares, 51 (53 percent) were seen with foals. Thirty-one of the 49 placebo implants were observed and 13 (42 percent) were seen with foals. Among an estimated 109 untreated mares with marker collars ob- served during the aerial observations, 51 (47 percent) were seen with foals. In the Stone Cabin area where mares had been treated in the fall of 1986 and any treatment effect would have been evident in 1988, 88 of 101 treated mares were observed (Table 2-4). Of the 45 EE-treated mares and 43 EP- treated that were seen during the flights, 5 (11 percent) and 3 (7 percent) were seen with foals, respectively. In the Wassuks, where 41 mares had received placebo implants, 15 (45 percent) of the 33 mares observed were seen with foals (Table 2-4). Clan Alpine and Wassuks percentages are remarkably close, and none is statistically different from the others. They indicate that neither the roundup and treatment of the previous September in the Clan Alpine area nor the implants (placebos or steroids) in the Wassuks and Clan Alpine areas caused any statistically detectable signs of abortion or early foal mortality. A proportion of the animals were pregnant at the time of implantation, and the percentages of the four groups seen with foals are approximately what would be ex- pected in a sample of undisturbed 3- to 12-year-old mares. For 1989 in Clan Alpine, 4 (9 percent) of the observed 45 EE mares and 3 (6 percent) of the observed 50 EP mares were seen with foals. Of the 31 observed placebo mares, 14 (45 percent) were seen with foals. In Stone

RESEARCH AND RESULTS 15 TABLE 2-4 Foaling Rates as Determined by Aerial Observation, 1988 Area and Treatment Number of Mares Number of Foals Foaling Rate 95 Percent Confidence Interval" Clan Alpine Treatment* 97 51 Placebo 31 13 Untreated, marker collarede 109 51 Stone Cabin Ethinylestradiol 45 5 Ethinylestradiol and 43 3 progesterone Wassuks Placebo 33 15 .53 .42 .47 .11 .07 .45 .43-.63 .25-.59 .38-.56 .02-.20 .00-. 15 .28-.62 The 95 percent confidence intervals were calculated by the formula p ± 1.96(pq/n)-5, where p is the proportion of mares with foals and q is the proportion of mares without foals. 'Treatments in the Clan Alpine mares were not effective until 1989. This estimate is based on an aerial survey (third census) of the Clan Alpine study area (May 21, 1988). In Clan Alpine, 413 adults (yearlings or older) were captured, 149 of which were females 3 to 12 years old and were implanted with treatment capsules or placebos. The proportion of mares in that sample was 149/413 = .36. On May 21, 1988, 301 adults were counted in bands with no marked horses. Assuming the same distribution of mares as in the collared sample (.36), 109 of the 301 adults were mares in the 3- to 12-year age class. In Clan Alpine, 66 foals were assigned to collared mares, 56 (.85) to 3- to 12- year-old mares and 10 (.15) to older mares. On May 21, 1988, 60 foals were counted in bands with no marked horses. Assuming the same distribution of foal production as in collared mares, 51 of the 60 foals were attributed to the 109 unmarked 3- to 12-year-old mares (Siniff et al., 1988a,b). Cabin, 1 (3 percent) of the observed 35 EE mares and 6 (16 percent) of the observed 37 EP mares were seen with foals. In the Wassuks, 21 (70 per- cent) of the observed 30 placebo-implants were seen with foals (see Table 2-5). The effects of the Clan Alpine steroid treatment were now evident in 1989 as were the Stone Cabin treatments for the second year. Meanwhile, the placebo implants bore foals in 1989 at about the same rate as in 1988. (In the Wassuks data the 95 percent confidence intervals around the foaling rate of 70 percent in 1989 overlap the 95 percent confidence intervals for

16 WILD HORSES: F/ELD STUDIES IN GENETICS AND FERTILITY TABLE 2-5 Foaling Rates as Determined by Aerial Observation, 1989 Area and Treatment Number of Mares Number of Foals Foaling Rate 95 Percent Confidence Interval" Clan Alpine Ethinylestradiol Ethinylestradiol and 45 50 4 .09 .06 .01-. 17 3 .00-. 12 progesterone Placebo 31 Stone Cabin Ethinylestradiol 35 Ethinylestradiol and 37 progesterone Wassuks Placebo 30 14 1 6 21 .45 .03 .16 .70 .27-.63 .00-.09 .04-.28 .53-.86 "The 95 percent confidence intervals were calculated by the formula/) ±l.96(pq/n)-5, where p is the proportion of mares with foals and q is the proportion of mares without foals. 1988, and thus the foaling rates for 1988 and 1989 at Wassuks are not statistically different.) For the combined areas and years (Table 2-6), the percentage of ob- served, treated mares seen with foals varied between 6 percent and 11 per- cent. If the 1988 and 1989 data for each of the two treatments are pooled, the effectiveness in curbing reproduction is essentially the same: 8 percent for 125 EE-treated mares, 9 percent for 130 EP-treated mares. The percentages compare with 49 percent and 57 percent for the 2 years in the placebo implants. Evidently both EE and EP implants effectively reduce pregnancy rates in mares for at least 2 years. By pooling all of the data, 22 (9 percent) of 255 observed, treated mares were seen with foals, while 114 (51 percent) of 222 observed, placebo- implanted mares were seen with foals. The obvious and consistent difference in the very low fertility of the hormone-treated horses compared with the placebo-treated or untreated horses cannot easily be attributed to any other cause but the hormone treatment. The horses in the Wassuks study area serve as comparisons to the hormone- treated animals in the study. The placebo-treated horses in the Clan Alpine study area, which were originally recruited from the Augusta Mountain area, are also consistent comparisons to the hormone-treated horses. The measure of fertility in the placebo-implanted and untreated horses is consistent

RESEARCH AND RESULTS 17 TABLE 2-6 Pooled Foaling Rates Year and Treatment Number of Mares Number of Foals Foaling Rate 95 Percent Confidence Interval" 1988 Ethinylestradiol (EE) Ethinylestradiol and 45 43 5 3 .11 .07 .02-.20 .00-. 15 progesterone (EP) Placebo* 161 79 1989 Ethinylestradiol (EE) 80 5 Ethinylestradiol and 87 9 progesterone (EP) Placebo 61 35 .49 .06 .10 .57 .41-.57 .Ol-.ll .04-. 16 .45-.69 "The 95 percent confidence intervals were calculated by the formula/? ±1.96(pq/n)-5, where p is the proportion of mares with foals and q is the proportion of mares without foals. 'Clan Alpine treatments combined with placebo implants in 1988. with generally observed foaling rates of wild horses, and the difference between them and the hormone-treated mares is clear and consistent in all test areas. VASECTOMY EXPERIMENTS The research on reproductive inhibition was designed to experiment with both stallion and mare fertility control. Because the objective is to achieve as many years of inhibition as possible, the decision was made to use vasectomy for stallion fertility control. In order to achieve treatment efficiency, the approach was to vasectomize dominant stallions in bands on the assumption that this would curtail reproduction in the bands. This assumption carries with it several often-unstated conditions if the procedure is to be effective: • The dominant stallion must do all or most of the breeding in the band. If a significant amount is done by the subdominant stallions, the effective- ness of sterilizing the dominant is diminished. • Bands must be relatively stable. If there is a significant exchange of mares or stallions between bands with vasectomized and intact stallions, the technique's effectiveness is reduced.

18 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY • Stallions must retain their dominance for several years. If dominance is highly transitory, then one-time sterilization of the dominant stallions of bands can be of only short-term effectiveness. Methods The Flanigan and Beaty Butte areas were selected for the vasectomy experiments. In the Flanigan area in northwestern Nevada, dominant stallions from 20 bands were vasectomized and fitted with radio collars; 5 dominant stallions from 5 other bands were radio-collared and left intact as compari- son animals; and 139 3-year-old or older animals from these 25 bands were given plastic marker collars. These treatments were administered in December 1985. In February 1986, 20 dominant stallions were vasectomized in the Beaty Butte area in the southeast corner of Oregon. An additional 67 animals, 3 years old and older, including 5 dominant stallions as comparison animals, were fitted with marker collars. In the spring of 1988, four aerial surveys were conducted by the research team in the Flanigan area and three in the Beaty Butte area. One survey of the Beaty Butte area was cancelled because of low clouds and high winds. Behavioral observations of horses continued during the spring and summer of 1988. The aerial surveys in the spring of 1988 were the last scheduled surveys for the stallion areas, and aerial data collection for these areas is now complete. Results Preliminary results of the stallion study are summarized in Table 2-7. Dominant vasectomized stallions were classified by whether they were (1) in stable bands where they remained dominant, (2) in unstable situations where they were switching bands, or (3) in stud bands. They were further classified as to whether any foals had been born into their bands that year. Examination of these data suggests that sterilizing dominant stallions may have been effective in diminishing reproduction in mountainous habitats, such as the Flanigan area. However, its effectiveness in the flat Beaty Butte area, where bands of horses regularly mingle, is less certain. Final recom- mendations on the effectiveness of sterilizing dominant stallions will be delayed until thorough analysis of aerial survey data and ground behavioral observations is complete. In the committee's view, vasectomy is an acceptable and effective method of sterilizing individual wild horses. However, several questions remain regarding its effectiveness as a long-term, population control procedure. Even though a dominant stallion may be vasectomized and sterile, will

RESEARCH AND RESULTS 19 TABLE 2-7 Status of Vasectomized Stallions as Determined by Aerial Observation, 1986 to 1988 Study Area and Year of Observation Cf 11; Mallion Code Number 1986 1987 1988 Flanigan 201 Stable, foals Stable, no foals Stable, no foals 202 Unstable Stable, no foals 204 Unstable Stable, no foals Stable, no foals 207 Stable, foals 209 Stable, foals 211 Unstable Stable, foals 216 Stable, foals Stable, no foals Stable, no foals 219 Stable, foals Stable, no foals 523 Stable, no foals Stable, no foals 529 Stable, no foals Stable, no foals Beaty Butte 520 Stable, foals Stable, foals 521 Unstable 532 Unstable 537 Unstable Stable, no foals Stable, no foals 696 Unstable Unstable Stable, foals 697 Unstable Stud band Stud band 701 Unstable Unstable Unstable 719 Stable, foals Unstable 732 Stable, no foals Unstable Stud band 734 Unstable Stable, no foals Stable, foals 735 Stable, foals 738 Stable, no foals 739 Unstable Stable, no foals Stable, no foals 744 Stable, foals Unstable 747 Stable, foals Stable, foals Stable, foals 750 Stable, foals Stud band Stud band SOURCE: Siniff D. B., J. R. Tester, T. C. Eagle, R. A. Garrott, and E. D. Plotka. 1988a. Fertility control in wild horses. Progress report to the U.S. Department of the Interior, Bureau of Land Management, November 30. Table 2. University of Minnesota and Marshfield Medical Foundation. Photocopy.

20 WILD HORSES: FIELD STUDIES IN GENETICS AND FERTILITY subordinate males in the band assume the duties of the dominant male and sire foals? How many foals are actually produced by mares in bands where the dominant stallion has been vasectomi/ed? Again, the data in Table 2-7 give ambiguous answers to these questions. In the Flanigan area, observations of the bands associated with a vasectomized dominant stallion showed six stable bands with no foals out of the six bands that were observed in 1987 and only one stable band out of the seven observed in 1988 that did have foals. But in the Beaty Butte area, in 1987 there were two bands with foals out of the five stable bands observed, and in 1988 there were three that contained foals out of the five stable bands that were observed that year. Other bands at Beaty Butte that were observed to include a vasectomized stallion during these 2 years were either unstable or stud bands. Moreover, in the genetics studies reviewed above, approximately one- third of the foals in the sample bands were not sired by the dominant stallion of that year. Either mares were moving between bands or the sub- dominant males were siring young. Because Berger (1986) has also reported that stallions of differing social status mate with mares, there is reason to withhold judgment on one-time stallion sterilization as a general fertility control procedure. Vasectomizing the dominant stallion is likely to extend the breeding season. Wild horse mares come into estrus every 21 days, from April through September. Once the mare becomes pregnant this cycle stops, and the stallion does not have to mate with this mare or defend her from other interested stallions. Therefore, another question arises. Will the vasectomized dominant stallion be undernourished as winter arrives because of the additional time and energy required to guard mares during estrus? The evidence so far indicates that this is not the case. It appears that vasectomized stallions graze sufficiently to maintain body weight and survive the winter satisfactorily. Does an extended breeding season give the subordinate males more op- portunities to breed the recycling females? And if so, do the mares become pregnant and produce foals? Berger (1986) found that stallion dominance was transitory, changing between years. This substantially diminishes the prospect that one-time vasectomy is a population control procedure that would remain effective over a period of years. These and possibly other factors must be weighed before conclusions can be reached on the efficacy of dominant-stallion vasectomy as a population control technique. On-site behavior observations also need to be integrated into the overall evaluation. At this point its efficacy looks doubtful.