Wild and Free-Roaming Horses and Burros: Current Knowledge and Recommended Research. (1980)

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.

EXECUTIVE SUMMARY BACKGROUND OF THE REPORT The Public Rangelands Improvement Act of l978 (PL 95-5l4) and contract AA 55l-CT9-l6 between the Bureau of Land Managment (BLM) and the National Academy of Sciences (NAS) direct NAS to impanel a committee to assess the state of knowledge on wild horses and burros, recommend research to fill gaps in knowledge, oversee the research during its conduct, and compile all relevant information at the end of a 2-year research effort. The state-of-knowledge assessment and the research design were designated Phase I of the total undertaking, and this document is the final report of Phase I. It reviews knowledge about a wide array of topics, recommends l8 research projects, and discusses information relative to policy questions without, itself, advocating policy. The Committee on Wild and Free-Roaming Horses and Burros was impaneled in June l979. It divided its task among three subcommittees with responsibility for horse and burro biology, effects on other ecosystem components, and sociopolitical and economic considerations. Following the introductory statement in Chapter l, the main body of this report is divided into four major sections. Chapters 2, 3, and 4 correspond to the subject matter investigated by the three subcommittees, and Chapter 5 is concerned with research and management methodology. There are also three appendixes. BIOLOGY OF HORSES AND BURROS History and Paleontology of Equids in North America The mainstream of equid evolution occurred in North America. Fossil evidence shows the presence of a large horse and an ass, structurally indistinguishable from the modern horse and donkey, as recently as ll,000 years ago. Their extinction occurred at that time along with the demise of a number of other species of large mammals. Modern wild horses and asses were reintroduced into North America by the Spaniards in the late l5th century. Some observers believe that the vegetation in the West was vulnerable to the introduction of domestic herbivores because it had experienced little grazing pressure since the late Pleistocene period. These observers consider equids to be particularly disruptive to the ecosystem because they are alien to

the region. However, the view may need to be tempered by a knowledge of the paleohistory of equids in North America. The possibility exists that there are vacant niches into which these animals could fit, Social Organization Two types of social organization have been reported in wild equids: (l) the harem or stable family group, with a dominant male; and (2) the territorial form, in which stable bonds occur only between mother and offspring. These may constitute the extremes of a continuum along which different species—and different populations within a species—occur, depending on environmental, social, and population factors. For example, feral asses (burros) in the arid southwestern United States have little social structure except for the mother-young relationship, exist at low densities, and display considerable aggressive behavior. In constrast, asses on humid Ossabaw Island, Georgia form stable groups and display little aggressive behavior. Under conditions of dry-season water stress in arid areas, asses concentrate within 3 km of water sources, and lactating females commonly threaten and reject their own young when they attempt to nurse. Concentration of the animals around watering areas has a heavy impact on local vegetation. Arid-land burros are browsers and may spend up to half their time feeding. On Ossabaw Island, however, burros are grazers and spend only about a third of their time feeding. Male asses in the Southwest display greeting behavior among themselves, but rarely are social grooming or social play seen among the young. Ossabaw Island animals exhibit the reverse of these patterns. The basic social organization among wild horses is that of a family group with a dominant male, subdominant males, and females and their young, but some variations on this pattern occur, as do exchanges between groups. In arid areas, distribution of horses is oriented around water during the dry season, but in areas more to the north, distribution seems to be oriented around availability of forage. Equid Demography Horses Although confined domestic fillies begin ovulating and breeding at l year of age, only one 2-year-old mare has been observed to bear a foal in seven wild horse studies spanning l to 5 years' duration. A small percentage (mean of l3 in the studied herds) breed at 2 and foal at 3 each year (gestation period is about ll months). Evidence suggests an increasing percentage of mares foaling in each older age class, as occurs in domestic horses, with around two-thirds of 5-year-old and older animals bearing young. Whether the percentage declines after ages l0 to l2, as in domestics, is not known. Wild

horse breeding is highly seasonal, with most foals born from April to June. Only crude approximations exist of first-year survival rates in wild horses; the available values range from 50 to 86 percent. Mean annual adult survival rates are also poorly known, but most estimates fall between 75 to 95 percent. Age compositions of 8,764 animals rounded up during herd reductions show that the greatest numbers are in the youngest age classes (40 to 45 percent of all animals are in the foal through 2-year-old classes), with progressively fewer in each older age group. Males slightly outnumber females at birth, decline to 39 percent of animals at 4 to 6 years of age, and then may increase again slightly in the older age classes. Total herd sex ratios approximate 55 percent female. Population increase rates calculated from BLM and U.S. Forest Service (USFS) census data average l5 to 20 percent annually for western U.S. horse herds, rates similar to those quoted by these agencies and cited in a number of earlier publications. In some cases, these may be magnified by (a) increasing commitment to and proficiency at censusing, (b) increasing visibility as herd sizes increase, and (c) change from fixed-wing to helicopter censuses in the l970s. But in others, the experience of observers, low-stature vegetation and moderate topography, and fencing which prevents ingress or egress would seem to preclude these biases. In contrast, two authors have projected increase rates with population models that incorporate birth and death rates similar to those published for several herds, and concluded that annual herd increase rates well below l0 percent are probable. Similar calculations with life tables in this report indicate that l5 to 20 percent increase rates can only occur in populations with geometric age distributions with (a) very high reproductive rates, and (b) virtually no mortality. Such demographic conservatism is produced in populations with half their numbers in pre-breeding or low-breeding (3-year) age classes, only about two-thirds of older mares foaling each year on the average, and some mortality. The question of increase rates is central to horse management, and the disagreement cannot be resolved with presently available information. Research is needed to settle the question. Burros A small percentage of 2-year-old burros foal l year earlier than horses. The percentage of 2-year-old and older jennies foaling exceeds 60 percent per year, on average, with the 2-, 3-, and 4-year-old percentages probably exceeding those for horses of the same ages. Some populations breed year-round, albeit with spring-summer emphasis in some. Survival rates are less well known in burros than in horses, but some evidence suggests high first-year loss in some areas and years, low in others. Age compositions are roughly similar in the two species, but some burro populations have higher percentages

of foals. The earlier breeding and higher fertility rates potentially enable burro populations to increase faster than horses, but reported rates of 20 percent per year and higher press the biotic potential of the species, given geometric age distributions. Some populations have been reported to increase very slowly or not at all, as in the case of several Death Valley populations. In general, burro demography appears more variable than that of horses, suggesting some sensitivity to density and the plasticity of a species adapted to the desert. Fecundity rates of females rounded up during herd reduction could be determined readily through rectal palpation. Genetic Polymorphism A knowledge of genetic polymorphism in horses and burros could give some idea of the minimum herd size needed to survive through periods of environmental change, and could delineate the racial lineage of wild horses, including their relationship to Spanish mustangs. While some work has been done on the genetics of domestic horses, none has been done on wild animals. Modern techniques of blood-group genetics provide a powerful tool for addressing these two biological questions. Nutrition While burros apparently prefer green grasses and forbs, they are highly opportunistic, broad-spectrum feeders, and are capable of surviving on high-fiber, low-nitrogen diets, including coarse shrub branches, yucca, and cholla cacti. Studies conducted so far show grasses ranging from 0 to 79.6 percent, forbs from 8.0 to 77.4 percent, and browse from 5.7 to 83.8 percent of burro diets at different seasons and in different areas. Horses are much more selective feeders. Some use of forbs and browse has been reported, but in 29 published diet analyses, consumption of grasses ranged from 36 to l00 percent of total diet, averaged 89.4, and made up 85 percent or more in 24 of the studies. This dietary preference coincides closely with that of cattle, and overlaps to some degree and in some seasons with those of elk, bighorn sheep, bison, and pronghorn antelope. Most dietary studies have not related animal data to vegetation composition, nor have they described spatial and habitat overlap with sympatric ungulates or lack thereof. There is some reason to believe that equids have higher forage intake rates per unit of body weight than ruminants because food can pass more rapidly through the equid's cecal digestive system. The ruminant is limited in its throughput rate by the capacity of the rumen and the fermentation rate that occurs there. As a result, the equid may have an advantage when only high-fiber forages are available, since it can compensate for the low nutrient content by increasing its intake.

Essentially no data exist on the nutritional responses of free-ranging equids in western North America to the well-studied and well-documented seasonal changes in nutritional content of vegetation. In the Southwest, forage quality is highest in late winter and early spring in the Mojave Desert, with its winter rainfall season; in late summer and fall in the Chihuahuan Desert, with its late-summer season; and at both times in the Sonoran Desert, with its bimodal rainfall pattern. In the Great-Basin-Intermountain region, forage quality is highest in spring and early summer. If equids can compensate for low-quality forage by increasing intake, then quantity rather than quality may be the factor that limits food; thus equids may be less subject to seasonal nutritional stress than are ruminants. Habitat Preferences Understanding habitat preferences and uses is important to detecting competition between equids and other herbivores, wild or domestic; to making forage-allocation decisions; and to establishing site-suitability criteria for equids, domestic animals, and wildlife. Competition occurs when two species use a common resource and reduce it to the point where the numbers of one or both species are limited. If the resource is not reduced to this point, the two species can both use it without competing. It is conceivable that two or more species of herbivores (a) may choose and occupy different habitats and thus not compete; (b) may have overlapping habitat preferences but segregate through behavioral interaction, thus competing only if food becomes limiting; (c) may occur in the same habitat but eat different foods, in which case they will not compete; and (d) may co-occur and eat similar foods, competing only when food becomes limiting. Recommended Research Seven research projects on the biology of horses and burros are recommended: • Project l: Habitat Preference and Use • Project 2: Food Consumption Rates and Nutrition • Project 3: Nutritional Plane, Condition Measures, and Reproductive Performance • Project 4: Blood Assays • Project 5: Demography • Project 6: Social Structure, Feeding Ecology, and Population Dynamics

• Project 7: Genetic Polymorphism. The projected 2-year span for the research is unrealistic. Because of the extreme year-to-year variability of environmental conditions and equid performance, no comprehensive picture can be developed in less then 6 to l0 years. Project l should be conducted in areas not less than 5 to 6 square miles per experimental treatment. Projects 2—and 8 and 9 to be listed later—can be carried out in paddocks of l00 to 300 acres. All experiments should be conducted in treatments involving horses only, cattle only, and horses and cattle, each at moderate and heavy grazing intensities. Horses and cattle are emphasized here because the possibility of their competition, both for space and for food, seems to be greatest. If funds permit, the research could be repeated with burros and with domestic sheep. Projects 2 and 9 should contain control areas without grazing. EFFECTS OF EQUIDS ON OTHER ECOSYSTEM COMPONENTS Impacts on Rangeland y Although it is widely alleged that horses and burros have severe grazing impacts on western rangelands, there are few published studies about the nature and extent of these impacts. Most of the existing studies are on grazing effects of burros. Studies along the lower Colorado River and in Death Valley National Monument showed heavy impacts on vegetation from grazing burros within a radius of 2 to 2.5 km from water areas. Studies in the Grand Canyon National Park showed heavy impacts at the Colorado River elevation and moderate to light effects at progressively higher elevations. Range in Bandelier National Monument was degraded over 4,000 ha by l07 to l20 burros. A study in the Lake Mead National Recreation Area, however, revealed no major impacts. Little controlled research has been done on impacts of grazing horses; the extensive management of horse range apparently proceeds largely from management-level inventories, experience, and j udgment. The range-ecology conceptual framework used in livestock management can at least be used as a starting hypothesis for, if it cannot be applied directly to, equid management. In this scheme, plant-community successional trends are roughly proportional to grazing intensity. Properly managed grazing—which takes into account the species, number of animals, season, and distribution of grazing—can be harmonious with most resource needs and values. The specifics of managing range vegetation vary geographically and seasonally with climate and vegetational type, year-to-year variation in precipitation can be a more influential factor in altering plant-community composition than season and intensity of grazing. Annual forage production is strongly correlated with that same variation, and herbivore numbers properly should be adjusted to the changes.

Interspecies Competition Because competition only exists where a population is limited to some degree, it is best demonstrated experimentally by manipulating the numbers of one suspected competitor and observing whether or not the other responds. If the population cannot be manipulated, a preliminary indication can be gained by calculating the resource need of each species, measuring the amount of resource available, and determining whether the need exceeds that available. Ideally, such calculations should be combined with population-limitation experiments. Burros are widely claimed to compete with desert bighorn sheep for water, forage, and space. Reports on water are conflicting and may depend on abundance. Competition for forage could occur near water holes. Two authors indicate that sheep avoid areas occupied by burros. While all of this evidence is equivocal, several authors point to negative correlations between burro and bighorn distribution in space and time. The possibility of burro competition with mule deer has been reported for Bandelier National Monument, and there is evidence of competition with small mammals in Grand Canyon, Death Valley, and Bandelier. Less work has been done on horse competition. Dietary overlap has been reported for some seasons and some areas between horses, cattle, elk, mule deer, pronghorn antelope, and bighorn sheep, with joint occupation of the same habitat in some cases. Effects of Equids on Soils There are numerous anecdotal or localized reports of equids, mostly burros, compacting soil surface, forming trails in steep terrain that accelerate erosion, and polluting water holes. Equids are potentially capable of the same types of impacts as are created by livestock. The latter have been thoroughly studied. Overgrazing (a) reduces protective cover and increases the impact of raindrops, (b) reduces soil organic matter and soil aggregates, (c) increases surface vesicular crusts, (d) reduces infiltration rates, and (e) increases erosion. Overgrazing reduces vegetation mulch, increases the proportion of bare ground and rock cover, increases soil bulk density, and reduces moisture infiltration rates. Heavy grazing increases the sediment load of watershed runoff, an effect caused mostly by vegetative reduction, but also partly by trampling. Serious problems of sediment production in the riparian zone are often associated with bank instability. Total and fecal coliform counts generally increase with the presence of livestock, especially during runoffs. In some cases, bacteria are stored in the bottoms and banks of streams.

8 Recommended Research The following research projects are recommended: • Project 8: Grazing Impacts on Range-Plant Communities • Project 9: Hydrologic Impacts • Project l0: Riparian-Zone Impacts Horse-cattle studies are again accorded priority because horses are more widespread than burros, potentially more serious competitors with livestock, and more likely to compete with cattle than with sheep. Horse-sheep studies should be initiated if resources permit. Studies of equids in relation to wildlife are not recommended at this time because the possible combinations (horse-elk, horse-deer, horse-antelope, horse-bighorn, burro-desert bighorn) are so numerous, and because controlled experiments with wild ungulates are so difficult. But we urge that federal and state agencies watch for opportunities to take before-and-after censuses of wildlife populations in areas slated for horse or burro herd reductions. Censusing l or 2 years before and several years after herd reductions could give clues to the existence of competition, especially if censusing were replicated in several areas. If nearby populations in areas with no equid reductions could also be censused in the same years, the results could be compared to create a roughly controlled experiment. SOCIOECONOMIC AND POLITICAL ISSUES In the Committee's opinion, several kinds of socioeconomic and political information are needed to facilitate decision making in horse and burro management. While there is abundant information on range and ranch economics in the western United States, there is little economic literature specific to wild, free-roaming horses and burros, and development of market and nonmarket valuation techniques is limited. Areas in which inquiry is needed include: (a) the value of and demand for wild horses and burros; (b) evaluation of adoption procedures; (c) evaluation of control and managment techniques; (d) analysis of optimal numbers for wild equids and management alternatives; and (e) evaluation of the costs of existing legal regulations and restrictions. The legal-political literature on wild horse and burro matters is extensive, particularly in terms of providing a perspective on the public agencies' overall land-management responsibilities—the context in which policies concerning wild horses and burros should be considered. Review of civil cases under the Wild and Free-roaming Horse and Burro Act of l97l shows that most lawsuits fall into two categories: (l) those challenging the need for roundups, and (2) those questioning the adequacy of the environmental impact statements

relied upon both by the government and by those challenging federal—as opposed to state—government authority over the animals. Concern has been expressed over protection of the animals, preservation of state control, impacts on rangeland, and the validity of information and views on population characteristics and impacts on other wildlife as well as the range resources. There are almost no data on sociological aspects of the wild horse and burro issue. Recommended Research Six research projects, one of which is designed at three levels of intensity, are recommended. They will provide a base of socioeconomic and political data that will facilitate decision making in equid management. The projects are organized into three groups in descending priority in terms of importance of information and urgency of funding: Group l includes: • Project llA: Taxonomy of Values and Benefits • Project l3: Management Costs of Alternatives • Project l4: Economic Considerations for Management Alternatives Drawn from Proposed Research Programs Group 2 includes: Project llB: Public Preferences for Alternative Management and Control Strategies • Project l2: Analysis and Evaluation of Demands for Excess Wild Equids • Project l5: Nonmarket Values Group 3 includes Groups l and 2 and adds the following investigations to provide socioeconomic data necessary to a systems-level understanding of wild-equid management: • Project llC: Public Attitudes, Preferences, and Knowledge • Project l6: Conceptual Development of Public Rangeland Management Models

l0 RESEARCH AND MANAGEMENT METHODOLOGY Methodology for censusing animal populations falls into three basic categories: (l) indices, (2) complete counts, and (3) various kinds of estimates based on sampling. Indices do not appear to have much potential in equid census, because they do not provide the estimates of actual numbers needed for forage allocation unless calibrated to total numbers. Current agency census efforts attempt complete counts from the air. The completeness of these—as well as the effects of such factors as vegetation type, topography, airspeed, altitude, type of aircraft, and observer experience—remain largely unstudied. One study showed experienced observers to be more efficient at spotting horses than inexperienced ones. The accuracy of existing censuses must be tested and correction factors devised for deviations from total accuracy. Several approaches can be taken. Complete counts are most likely to err on the conservative side, but the Committee's impression is that current horse censuses, especially in open terrain, are reasonably accurate. On the other hand, one test of accuracy of a burro census in Arizona showed that only about a third of the burros had been counted. Some estimation techniques—especially mark-resight methods—may be useful with burros, and plot sampling may be possible for horses. These methods should be coupled with others, preferably complete counts, so that accuracy can be checked. Accuracy of an equid census can be affected by relative visibility, which may increase as group size increases; by observers' experience, as mentioned above; and by certain approaches to random sampling. Preliminary analysis of BLM and USFS census data showed: (a) a failure to standardize the season of census, which raised the problem of a seasonal change in numbers due to foaling; (b) an abrupt 88 percent mean increase in horse numbers in the years when helicopter census replaced fixed-wing-aircraft census; and (c) less variability in the helicopter counts. The "Soil-Vegetation Inventory Method" is commonly used in contemporary range-survey work and for a number of other purposes, including compliance with the wild horse and burro mandates of recent legislation. The Committee reviewed l0 BLM and joint BLM/USFS wild horse capture plans with their accompanying environmental analysis reports (EARs). Eight reductions were proposed because of problems perceived in range conditions. However, few provided much information on range condition and the techniques used to determine it, or on which herbivores (horses, cattle, wildlife) caused the problem. The most recent EAR provided detailed supporting data. The Committee concluded that, while range studies have not always been properly used to support adjustments in numbers of wild equids, the technology exists and appears adequate. Fecal analysis, the most widely used technique for analyzing diets, is currently subject to question in ungulate studies. Not only do some consumed plant species fail to appear in feces, but the proportions of food items consumed and those showing up in fecal remains differ. The equid digestive tract may be less subject to

11 these problems, a possibility that is supported by studies on zebra diets. However, conclusions based on equid fecal analysis should be drawn with caution until the method has undergone further study, preferably with the use of fistulation. A more intractable problem is that of the time lag between consumption of forage and fecal deposition in highly mobile species such as equids. Defecation may not occur until 37 hours after ingestion, making it difficult to relate diets to the vegetation and habitat from which they were taken. Statistical problems and lack of microhistological reference material may pose other difficulties. The in-vitro techniques widely used for studying ruminant nutrition should not be relied upon until they have been proven for equids. In-vivo comparisons, the use of indicators, and regression procedures should all be tried. Assays of a number of chemical constituents in the blood may have potential for (a) evaluating nutritional condition of individual animals, and (b) using an animal's condition to indicate the nutritional adequacy of the range it occupies. Blood samples could be taken easily from horses and burros brought in from herd roundups, and from animals used in the research projects. A number of the research projects outlined in this report can use confined animals, domestic ones, or both. Questions will arise as to the degree to which the results from the two categories can be extrapolated to wild and free-roaming animals. Observations of the behavior of the two former groups and of wild and free-roaming animals can be used to assess the comparability of results and to facilitate extrapolation from one group to another. A set of observations of behavior is set forth to assist in cross comparisons. The set includes considerations in selecting the animals to be observed, statistical aspects, behaviors to be recorded, and schedules of observation. In addition, recommendations are set forth for observations of behavior to be made within the specific research projects outlined in this report, including an extensive repertoire of social and maintenance behavior. The rationale for each recommendation is included. If fertility control is deemed a desirable method for limiting population, a range of contraceptive agents is available that could be implanted and might be effective for up to 5 years. Considerations of population and behavior point to attempts at reducing fertility in mares rather than stallions. The technique needs to be researched, however, initially in captive animals. Chemical immobilization is not deemed an efficient primary capture technique for wild horses, but it can be used to quiet captured animals for purposes of research and handling. The preferred drug for this use is etorphine (also known as M99 or Immobilon). Two methodological research projects are recommended: • Project l7: Census Methods Project l8: Contraception Studies

l2 Project l7 should investigate the validity of two or three alternative census techniques, including "complete" counts. The project should begin with a pilot effort on horses, later extended to burros. Project l8 should evaluate contraceptive methods.