Evaluation of Sheep Losses on a Range Lambing Operation in Southeastern New Mexico

Evaluation of Sheep Losses on a Range Lambing Operation in Southeastern New Mexico f» i AGRICULTURAL EXPERIMENT STATION RESEARCH REPORT 34 VEB^

SUMMARY CONTENTS Study area Procedures 2 Results 3 Flock size and composition 3 Sheep mortalities 3 Frequency of predation 8 Location of wounds 9 Parts of animals consumed 9 Effort of search Discussion 2 No control area 2 Radio telemetry 2 Predation 2 Losses to other causes 2 Literature cited 3 ACKNOWLEDGMENTS The study was funded by the U.S. Fish and Wildlife Service, Denver Research Center, through a contract with the Agricultural Experiment Station, New Mexico State University [Project No. 4-37]. We wish to thank E.W. Pearson, U.S. Fish and Wildlife Service, Denver Research Center for assisting with the harness design and radio telemetry work during 974. Several students including E.D. Hulsey, LD. Humphreys, R.R. Cudney, A. J. Hobday, A. C. Porter, and B. A. Locke assisted with field searches and necropsies. Mortality transmitters and receivers used in this study were provided by the U.S. Fish and Wildlife Service, Denver Research Center. This report presents an assessment of a predator damage on domestic sheep in southeastern New Mexico. The study area was a 3,84-acre (4 ha) range lambing operation without predator control but adjacent to lands with intensive control. The field studies were conducted during 974 and 97. In 974, there were 22 rams, 633 ewes, and 339 lambs in the flock. That year, between March 3 and October 9, 44 dead sheep were located ram, 27 ewes, and 6 lambs. These losses represented 4., 4.3, and 34.2 percent of each category of the flock, respectively. No losses of adult sheep were attributed to predation. Of the 6 lamb mortalities, 3 (46%) were attributed to predation and 63 (4%) were attributed to causes other than predation. The 3 lambs lost to predators represented.6 percent of the lamb crop. The breakdown of predation losses was 4 (77.4%) to coyotes, (9.4%) to bobcats, 2 (3.8%) to black bears, and (9.4%) to undetermined predator species. In 97, there were 8 rams, 29, ewes, and 43 lambs in the flock. Between December 7, 974 and October 3, 97, a total of 4 dead sheep, 36 ewes and 78 lambs, were located. These losses represented 6.8 and 8. percent of the ewes and lambs, respectively. Of the 36 ewe losses, (4%) were attributed to predation and 3 (86%) were attributed to other causes. The ewes killed by predators represented.9 percent of the ewes. Of the 78 lamb mortalities, 2 (67%) were attributed to predation and 26 (33%) to other causes. Predators killed 2. percent of the lamb crop. All 97 predation was attributed to coyotes. During 974, 3 of 44 losses (36.8%) were attributed to predators; in 97, 7 of 4 losses (%) were attributed to predators. New Mexico State Univ or national origin. sity is an equal opportunity employer. All programs are available to everyone regardless of race, color, Las Cruces, New Mexico

Evaluation of Sheep Losses on a Range Lambing Operation in Southeastern New Mexico Donald G. DeLorenzo and V.W. Howard, Jr. Organizations and groups of people representing various economic, aesthetic, and ecological interests have varied opinions of the influence of predators and their impact on the livestock industry. In New Mexico, sheep ranchers have reported for years that predators take varying percentages of lamb crops. The U.S. Fish and Wildlife Service has concentrated its efforts in predator control in the southeastern portion of the state in recent years (pers. com. Vernon Cunningham, U.S. Fish and Wildlife Service, Albuquerque). After the presidential ban on use of poisons on federal lands, many ranchers expressed concern that sheep losses to predators, particularly coyotes [Canis latransl, would increase. They believed that the remaining control measures would not be adequate to prevent predators from increasing. A need for scientific data concerning livestock losses to predators was highlighted in the "Cain Report" [Cain et al. 972]. Part II, Recommendation. The study reported here was initiated to determine the extent of sheep losses to uncontrolled predators on a range lambing operation in southeastern New Mexico. Study Area The study area was the privately owned Leroy McKnight Ranch, in Lincoln County in southeastern New Mexico. The ranch encompasses 3,84 acres in the Upper Hondo Soil Conservation District. The ranch is in the Deama-Limestone Rock Land soil association [Maker et al., 97], at an ' Former research associate and associate professor of wildlife sciences. elevation of approximately 6, feet. The predominant slope ranges from 2 to 6+ percent. The soils are predominantly shallow with a large amount of rock on the surface and numerous rock outcrops. Approximately 97 percent of the Limestone Rock Land Association is non-irrigable. The dominant vegetation association consists of Juniperus spp. canopy interspersed with Pinus edulis. The dominant shrub was Quercus sp. on northfacing slopes with a complex variety of grasses and forbs interspersed with occasional stands of cholla (Opuntia imbricata) on level and south-facing slopes. The McKnight ranch is composed of two main pastures-the north pasture, which contains approximately two sections, and the south pasture, which contains approximately four sections. The north pasture was used primarily in fall, winter. and early spring by ewes and year around by 8 to 22 rams. The south pasture was used in spring as a lambing pasture and during the summer and early fall, up to marketing, as the major production pasture for ewes and lambs. Most of the sheep were in the south pasture during the study. This study was designed for no predator control measures to be used on the study area. On the north border of the area was a cattle ranch with no predator control program. On the east was a conglomerate of a portion of the cattle ranch, and two sheep ranches, with relatively ineffective predator control. On the southern border was an area of the Mescalero Apache Indian Reservation which had neither predator control nor human inhabitation. Along three miles of the western boundary was a sheep ranch with intensive predator control. This ranch, approximately three

miles wide, was bordered on its western side by the Lincoln National Forest, where predator control was limited to steel traps. In the study design, it was planned that the cattle ranch, Mescalero Apache Indian Reservation, and Lincoln National Forest would function as predator reservoirs for the study area and fill any voids in predator populations created by past predator control on the Leroy McKnight Ranch. Sheepmen with ranches adjacent to the McKnight Ranch believed that the absence of control activities on that ranch would contribute to their predation problem. Therefore, predator control practices were increased on the adjacent sheep ranches in 97. Permission was sought and granted for Jim Gibson, the sheepman on the western boundary, to initiate predator control on the cattle ranch adjacent to the study area and also on the Mescalaro Apache Indian Reservation. These intensive predator control efforts were essentially as effective as if they had been conducted on the study area, since there were no known resident predators on the Leroy McKnight Ranch. All predators known to prey on domestic sheep were transients. The disturbance created by daily field searches for sheep mortalities possibly prohibited predator species from establishing residency within the study area, but more likely the study area was too small to support resident populations of coyotes and bobcats (Felis rufa). Being transient caused the animals to be subject to the intensified predator control operation when entering and leaving the study area. The Leroy McKnight Ranch had a range lambing operation, which is the primary type of operation in New Mexico. Under this management practice, ewes lamb on the range with no intervention or help from herders. Lambing on the McKnight Ranch occurred in the four-section south pasture, where it was impractical to observe or assist ewes. Normally, sheep were rounded up and sheared in April before lambing began in May. Lambs were docked, vaccinated, marked, and castrated in early June and marketed in October. PROCEDURES At shearing, the adult sheep were counted to determine the size of the flock. This was the most complete round-up and thus the most accurate count of sheep on the McKnight Ranch. Lambs were first counted at marking, and the final lamb count was conducted at marketing. Radio telemetry, observation of avian carrion feeders, and daily intensive field searches were used to locate dead sheep. Lambs were equipped at marking with mortality transmitters, which used a motion-sensitive mercury switch to keep them inactive. Movement of the lamb, activated the mercury switch before a pre-determined interval, which was regulated by a capacitor within the transmitter, and discharged the capacitor so that no signal was transmitted from the unit. If the mercury switch remained motionless for the necessary time, as it would on a dead lamb, the capacitor became fully charged and activated the transmitter. Dead lambs were located with the aid of a portable receiver and directional, handheld antenna [Kolz et al. 973]. The daily field searches incorporated vehicles, horses, and walking as modes of transportation. Carcasses were located primarily by observing the activities of avian carrion feeders, actual sighting of a carcass, and the use of radio telemetry. Turkey vultures (Cathartus aura) were the most prevalent species of the avian carrion feeders and were efficient in locating carcasses. During early morning hours, immediately following sunrise, searches were made from vehicles and horseback for carcasses before evidence was destroyed by turkey vultures. Following this rapid coverage of the study area, a more intensive search was made during the later hours of the morning to check the more remote, inaccessible pockets and canyons. After a carcass was located, it and the surrounding area were photographed, and signs which might aid in determining the cause of death were recorded at the scene. A necropsy of the carcass followed, and additional signs which would provide information helpful in determining the cause of death were recorded. The necropsy involved an external examination of the carcass, examination of the hide and carcass after complete skinning, and an internal examination. A predator kill was distinguished from carrion feeding by evidence of heart beat at the time of injury. Punctures such as those made by the canine teeth of a predator are visible inside the hide. If the heart is beating when the punctures are made, massive hemorrhaging occurs in the surrounding tissue. These hemorrhages were highly visible on the underside of the hide and immediate underlying muscle tissues of the carcass. If the wounds were made after the heart had stopped beating, the canine punctures were visible, but the massive hemorrhaging in the surrounding tissues was not present. Carcasses which had deteriorated and dried so that the hide was no longer pliable were removed from the field, labeled with the necropsy number, and soaked in water until the hide was once again pliable. The hide could then be inverted and

examined for punctures and hemorrhaging. The hemorrhaged areas would appear as dark brown stains on the underside of the hide. These were the primary criteria used to determine predator kills. Information gathered from the necropsy and the surrounding area was used to classify the causes of death as predation including species of predator whenever possible, other than predation, or unknown. Massive hemorrhaging on a carcass or the underside of the hide was the criterion for a predator kill. This established the minimum number known killed by predators. Cause of death other than predation included carcasses which could be necropsied but did not exhibit the massive hemorrhaging of a predator kill. The "unknown" category was comprised of carcasses that had deteriorated so that no information on the cause of death could be obtained. Mr. McKnight was compensated for all predator losses at a rate determined by average weight and price per pound received for other lambs at shipping time. Values of ewes lost to predators were determined by adding the average price of wool lost for one year to the market price for breeding ewes in October of that year. In return, Mr. McKnight agreed to practice no predator control on his ranch during the study. RESULTS Flock Size and Composition The McKnight flock consisted of 633 ewes and 22 rams in 974. They were observed from early lambing in March until marketing on October 2, 974. In 97, the flock was comprised of 29 adult ewes and 8 rams. Observation began in December of 974 and terminated at marketing on October 4, 97. The flock was sheared ami counted between April 2 and April 4, 974 and April 2 and April 4, 97. Early lambing in 974 occurred in February, the main lambing period came in May, and a few lambs were being born through the latter part of June. In 97, the early lambs began dropping on March, the main lambing came in the last half of April through the month of May, and lambing was completed by June 4. The early lambs were from ewes bred by adjacent ranch rams, which attained access to the McKnight Ranch on the eastern and southern boundaries of the study area. For a lamb to be considered early. it had to be born before the end of the -day gestation period which began when rams were put out for breeding. The desired lambing period for 974 was May and for 97 the last half of April and May. In 974, 27 lambs were gathered at marking and 266 (98%) of these were fitted with mortality transmitters. In 97, 378 lambs were marked and 23 (6%) were fitted with mortality transmitters. The count at marking in 974 was 27 lambs and ewes for a 4.2 percent lamb crop. Many ewes that did not have lambs were not gathered at marking time. In 97, the count at marking was 378 lambs and 29 ewes for a 7 percent lamb crop. These numbers are the first estimates of the lamb crop on a range lambing operation. Lamb losses from birth to marking are not considered in figuring the percentage lamb crop. In a range lambing operation, an accurate birth rate is almost impossible to obtain. The topography and vegetative cover on the McKnight Ranch probably made it as difficult as any area in the West to locate and count sheep accurately. A more precise count of the lamb crop was obtained by adding the marking count to known unmarked lambs and lambs that were known to be dead before marking, each year. That count is still low for the true lamb crop. In 974, totals of 633 ewes and 339 lambs gave a known lamb crop of 3.6 percent. In 97, totals of 29 ewes and 43 lambs gave an 8.3 percent known lamb crop. Sheep Mortalities The first mortality recorded in the 974 study period occurred on March 3 and the last on October 9. During the next study period, the first mortality was recorded December 7, 974 and the last on October 3, 97. During the first period, 44 sheep carcasses were located and examined for evidence of cause of death. Of these, 28 were adult sheep (27 ewes and ram) and the remaining 6 were lambs (table ). Replacement ewe lambs are kept in the breeding herd rather than being marketed. This category in tables and 2 provides a set of base-line values from which losses in preceding age classes occurred. During 974, no losses of adult sheep were attributed to predation. Of the know 6 lamb mortalities, 3 (46 I were attributed to predation and 63 (4%) to all causes of death other than predation (includes 9 unknown causes). Of the 3 predation mortalities, 4 (77.4%) were due to coyotes, (9.4%) to bobcat, 2 (3.8%) to black bear, and (9.4%) to undetermined predator species. Location of the predation mortalities for 974 are shown in figure. In 97, 4 sheep carcasses were examined, including 36 adult ewes and 78 lambs (table 2). Five of the 36 adult ewe losses were attributed to predation, 27 to causes other than predation, and 4 to unknown causes. Of the 78 lamb mortalities. 3

Table. Age distribution of all sheep mortalities on the McKnight Ranch during 974 Number of Total Numbers Lost Year Born Replacement Ewe Lambs Numbers Lost Numbers Lost To Predation Other Than Predation 974 4 973 6 972 8 97 97 3 969 968 2 967 966 2 Table 2. Age distribution of all sheep mortaliti n the McKnight Ranch during 97 Number of Total Numbers Lost Year Replacement Numbers Numbers Lost Other Than Born Ewe Lambs Age Lost To Predation Predation 2 were attributed to predation, 22 to causes other than predation, and 4 to unknown causes. All predation mortalities were attributed to coyote. The location of the predation mortalities are shown in figure 2. The losses not due to predation were from diverse causes. These are shown for 63 lambs and 28 adults in 974 and 26 lambs and 3 adults in 97 (table 3). Numbers of all sheep mortalities by categories of loss and the month in which mortalities occurred are shown in table 4. The loss of adult sheep was greatest both years during May, which coincided with the primary lambing periods. This period of high adult mortality may partially be explained as a result of added stress and problems associated with lambing. May was also the month in which most lamb losses were attributed to predation. Three potential explanations for this phenomenon are ) predators are opportunists and the incidence of lamb encounters was greatest during this month; 2) very young lambs are more susceptible to predation; and 3) predators are attracted to lambing flocks. Both years, the decrease in monthly predation lamb losses after May indicates an inverse relationship between lamb size and susceptibility to predation. After the major loss period in 97, September was the only month in which there was a subsequent increase in predation, and this was due to one male coyote. The relationship between lamb size and its susceptibility to predation may be stronger with bobcats or inexperienced coyotes than with older coyotes experienced in killing domestic sheep. After lambs weighed 2 to 3 pounds in 974. the bobcat, apparently a small one, had considerable difficulty in killing a lamb. During both study periods living lambs were observed to have wounds inflicted by coyotes. This suggests that either the lambs were large and strong enough to escape from the predator, or that the lambs were not fully acceptable as prey for the coyote, or that some ewes presented defenses of their lambs that discouraged some coyote attacks. One adult male coyote in the study area killed lambs as large as 82 pounds during 97. All lambs in a given growing season may fall into such a predator's range of susceptible prey. A possible

**>

Table 3. Sheep losses on the McKnight Ranch not attributed to predators, 974 and 97 Numbers Lost Category of Loss Physiological Disorders (Heart, Lung, Urinary, Etc Stress Malnutrition and Mismothering Lambing Complications Stillborn Docking Injury Rattlesnake Bite Accident Infection No Sign of Predation Not enough evidence available to further categorize these losses, but difinitely not predation. 2 Carcass remains either decomposed or destroyed past point of determining any cause of death. Table 4. Numbers of all sheep mortalities on the McKnight Ranch by category and month, 974 and 97 Month and Year Adi jits Other than Predation Lambs Other thar i Predation Adults Lost to Predation Lambs Lost to Predation Adults Unknown Lambs Unknown Total January 974 97 February 974 97 March 974 97 April 974 97 May 2 4 6 4 3 4 3 2 6 974 97 July 974 97 974 97 September 974 97 October 974 97 November 974 97 December 974 97 -Th lis mortality is included in the 97 study year although with this chronological tabulation it is shown as 974.

factor regulating what constituties prey for the more experienced coyotes would be the acceptability of the prey to the individual predator instead of the prey's susceptibility to all predators of a given species. Losses of lambs to causes other than predation in 974 were highest in June, one month later than losses to predation. Normally, these losses would have been highest in May, when young lambs were most abundant; the peak in June was due primarily to adverse weather conditions. During April and early May, 974, there was sufficient green vegetation for ewes to produce ample quantities of milk for their lambs. By early June, lack of precipitation and hot days greatly reduced the quantity of green forage and ewes began to go dry. This lack of milk and green forage probably caused nutritional problems for lambs and poor ewe-to-lamb bonds, which resulted in increased lamb losses from causes other than predation. Rains began in July resulting in a green-up of vegetation and a pronounced decrease in lamb mortalities to causes other than predation. The belief that the June, 974, lamb losses to causes other than predation was due to adverse weather conditions was supported by the 97 data. The 97 losses in this category were highest in May, when young lambs were numerous, but fewer lambs were lost in 97. Weather was again the major factor regulating the losses. Forage production was excellent in 97. Winter moisture carried the spring green-up into the period of summer rains. With high-quality forage, the flock was healthier and deaths due to causes other than predation decreased. In 974, lambs lost to unknown causes reached a high in May, remained relatively high in June, and decreased in July. In 97, the lamb mortalities categorized as unknown were insignificant. As expected, the greatest losses in the unknown category should, and did, coincide with the months of greatest lamb mortality. Data on numbers of sheep and losses to predators for 972 through 97 are presented in table. Predation on adult sheep was minor compared with lamb predation. The numbers of lambs lost to predation were relatively stable. There may be a correlation between a relatively stable predator population and the number of lambs these predators kill during a given year. If the predator population was under intensive control prior to and during the study periods, as was previously suggested, then it probably has remained a relatively stable population for the past four years. It was estimated that the coyote numbers on the study area for 974 and 97 were six and five transients, respectively. This estimate was based on signs observed by the landowner, federal trapper, and the field personnel on the study. No coyotes were known to den on the McKnight Ranch or to remain on the ranch for longer than 24 hours during the study. When it was possible to track a coyote, no discernible differences were found in the age of the tracks entering and leaving the study area. Frequency of Predation Losses of sheep to predators by months for 974 and 97 are shown in table 6. In 974, one lamb was lost to a predator every.48 days during the primary lambing period (May). The frequency rate remained high in June (/.88 days) while lambing was still occurring. From July i through October 2, 974, when lambs were marketed, there was an inverse relationship between time and the intensity of predation. Lamb predation in 97 was intermittent. The increased intensity in March and September of 974 can partially be explained by sheep distribution during these two periods. It was known that a problem coyote was using the northern portion of the north pasture. The majority of the sheep flock, except for 8 rams, was run in the south pasture. Sheep from the south pasture were transferred to the north pasture during round-up for shearing beginning on April 2, 97. This increase in sheep numbers in the north pasture and the fact that early lambs began dropping on March, 97, Table. Numbers of sheep and losses to predators on the McKnight Ranch, 972-97 Numbers of Sheep Losses t o Predators Year Ewes Lambs Lamb Crop Ewes Percent Lambs Percent 972* 973* 974 97 77 747 633 29 4 3 339 43 9. 4.2 3.6 8.3....9 26 4 3 2 4.8..6 2. a previous interview survey [Howard, V. W. Jr. 974].

Table 6. Predation intensity, number of mortalities and number of days per mortality, on the McKnight Ranch, 974 and 97 974 Adult 97 Adult *Lambs marketed October 2, 974 and October 4, 97 may explain the increased intensity of predation. The increase for the month of September, 97, also may be partially explained by the high concentration of sheep in the north pasture before marketing. The peak of the 97 lamb predation occurred in May (/.3 days) which again coincided with the peak lambing period. Predation on adult sheep for 97 was low and intermittent and no discernible pattern for either area or time period was exhibited. Location of Wounds Wound location was used in determining the species of predator responsible for the mortality (table 7). Bobcats wounded sheep on the dorsal surface of the head and neck four times and on the back the fifth time. The fifth carcass may have had wounds in the neck, but the neck hide and underlying muscle tissues had been eaten by either the bobcat or turkey vultures. The two sheep killed by a black bear had ventral neck wounds. The size of the canine punctures and distance between these punctures distinguished these sheep from those killed by coyotes. The dorsal and ventral neck-head regions were the major locations for coyote-inflicted wounds. This is in agreement with Davenport et al. [973] and Nesse [973].Rosko [948] and Gier [ 97] reported that a coyote typically gashes the underside of a lamb's throat. The gashed throat type of wound was not observed in this study. In the throat region, separate punctures into the esophagus were typical (figure 3). Gier [97] reported that coyotes grab small lambs by the head and damage the skull. Damage to the head of small lambs in this study was comprised of punctures into or crushing of the skull cap and shearing off the posterior ends of the lower mandibles (figure 4). Parts of Animals Consumed The parts of sheep consumed by predators are given in table 8 for 2 bobcat, 2 bear, and 37 coyote kills. These were mortalities in which a determination of the portions consumed by each species could be made. Carcasses were not counted if avian carrion feeders or species of mammals other.than the predatory species had altered them so that portions consumed could not be determined readily. Neck muscles were consumed on two of the five sheep killed by bobcat. Damage to the other three carcasses did not allow accurate classification of portions consumed. The two sheep killed by bear Table 7. Locations of wounds on lambs by species of predator on the McKnight Ranch, 974 and 97 Year, Predator, and Sheep Head & Neck Region Dorsal (a) Ventral (b) No. %/sp No. %/sp Other Areas on Body (c) a&b No. %/sp No. %/sp Multiple Wounds a&c b&c No. %/sp No. %/sp a&b&c No. %/sp 974 Bobcat Bear Coyote Unk. Sp. Total 97 4. 2. 26.3

Fig. 3. The skin of a lamb killed by a coyote showing puncture wounds in the ventral and lateral neck and top of the head. Fig. 4. Skull and neck of a young lamb killed by a coyote showing canine punctures throughout the skull cap and sheared lower mandibles. Table 8. Parts of sheep carcasses consumed by predators o n the McKnight Ranch, 974 and 97 Bobcat Bear 974 97 974 Coyote Totals Lamb 974 97 97 974-7 Neck Body ai I Thoracic Contents Thoracic Contents Hindquarters No Consumption Digestive Tract EWE were both fed on more extensively than those killed by any of the other predator species. The portions of the carcasses not consumed by the bear were the skull, hide, lower legs, stomach, and intestines. All other portions of the carcasses, including bones, were consumed entirely (figure ). The portions consumed for 37 sheep killed by coyotes were documented for 974 and 97. In 974, 9 (42.9%) of 2 lamb carcasses had the thoracic contents consumed, and two (9.%) other carcasses had portions of the hindquarters eaten by coyotes. There was no evidence of coyotes having fed on the remaining (47.6%) carcasses which had not been altered by other animals. In 97 there were 6 sheep killed by coyotes on which a positive determination could be made of the portions consumed. Fourteen of these were lambs and two were adult ewes. Portions consumed on these 4 lambs including 4 (28.6%) thoracic contents, 2 (4.3%) portions of the hindquarters, 2 (4.3%) digestive tract, and 6 (42.8%) had no evidence of consumption. Portions of the hind-quarters of the two adult sheep had been eaten. Some evidence of coyote consumption was present on (2.4%) of 2 mortalities in 974 and (62.%) of 6 mortalities in 97. These data (table 8) are presented for only those mortalities where a positive determination could be made. These data should not be extrapolated to all predator-inflicted mortalities unless one accepts the assumption that carcasses on which consumption has occurred are altered to the point that no determination can be made at the same rate as carcasses on which no consumption occurred. The authors have no data to either support or refute this assumption. Effort of Search The primary means for locating a carcass were by direct sighting, observating of avian carrion feeders, and the mortality transmitters. In 974 and 97,

Table 9. Mode of transportation utilized in locating sheep mortalities on the McKnight Ranch, 97 Month Transportation i Hours Number of Searched Carcasses Hours per Carcass March Vehicle Horseback Walking 33. 8. 2. 2 33. 3.7. April Vehicle Horseback Walking 9. 82. 8. 8 9. 6..6 May Vehicle Horseback Walking 236. 34.2 68. 6 8 39.3 7. 9.8 Hi PHI Fig.. Carcass of a lamb killed by a black bear with only head, hide, lower legs, and viscera remaining. All fleshy parts and most hones were consumed by the bear. respectively, 77 and 66 carcasses were located by sight; avian carrion feeders aided in finding 37 and 43 carcasses; and mortality transmitters accounted for 3 and carcasses. The mortality transmitters were not available to aid in location until May 2 7. 974 and June 7, 97. The man-hours of field search for locating mortalities from March to October 4, 97, is given by mode of travel in table 9. Walking appeared the most efficient mode, but this was not a random search. Due to the necessity of covering large areas as rapidly as possible, in most instances walking was incorporated after searchers had some indication of the location of a mortality, such as vulture activity. Therefore, walking appeared to be the most efficient in hours spent locating a sheep mortality. Traveling horseback was more efficient in hours per mortality than was search from vehicles. Travel by both horseback and vehicles afforded the searchers rapid mobility. However, the searcher traveling horseback was not restricted to roads or areas observable from roads. These average times for locating mortalities may be misleading as they do not account for time each carcass was available. The availability of carcass was designated as the interval between the estimated time (date) of death and the time (date) the mortality was located. The following data estimations are given for 974 and 97, respectively. Within one day after the estimated date of death, 72 (.%) and 3 (46.%) carcasses were located. Between two and seven days, 36 (2.%) and 3 (27.2%) carcasses were located. From eight to 4 days, 2 (8.3%) and 8 (7.%) were located. Carcasses available more than 4 days were 24 (6.7%) and 22 (9.3%). Those available for June July August Vehilce Horseback Walking Vehicle Horseback Walking Vehicle Horseback Walking Vehicle Horseback Walking Vehicle Horseback Walking 9. 23.7 69.7 22.7 84.7 3. 4. 34. 4. 6. 3...2 9.2 279.2 4 7 2 4 3 38. 66 4. 42.6 26.4 6.2 72. 33.6 more than 4 days were conbined since it was difficult to determine the date of death. Cumulative values of these data for 974 and 97, respectively, show that 72 (.%) and 3 (46.%) sheep carcasses were found within one day of their availability. Within seven days 8 (7. I and 84 (73.7%) were located, and within 4 days 2 (83.3%) and 92 (8.7%) carcasses were located These values suggest intensity and success of searches were similar for the two study periods although sheep numbers and mortalities were obviously different. The similarity of the cumulative percentages for the two years probably results from having most of the sheep in the south pasture when lambs were available. Size of the pasture and vegetative cover were stable for the two periods, so that the degrees of difficulty in locating carcasses were similar. The fewer carcasses in 97 may account for the slightly lower percentages of searching success because fewer carcasses generally become more difficult to locate in an area of a given size whenever other factors remain fairly constant. 4.8

DISCUSSION No Control Area The study did not achieve its designed goal of no predator control. Instead, the area was moderately to intensively controlled because of the intensified control programs on surrounding areas. For a truer representation of a range lambing operation without predator control, it would be necessary to either obtain jurisdiction over a large buffer area around the perimeter of the study area or locate a sheep rancher whose cattle-raising neighbors do not control predators. To obtain the necessary lands and retain jurisdiction over those lands would necessitate the use of federal or state lands. A large land area which could be manipulated to satisfy the needs of the study would allow a true represention of a range lambing operation without predator control. The quality of data would be enhanced if the sheep were also managed by the research team. The primary problem in this study was that the flock management practices were not designed for pertinent data collection. For example, round-ups at docking or marketing were incomplete. At shearing in 974, ewes were ear-tagged so that accurate counts could be obtained at docking and marketing. With these counts, sheep mortalities not located by the search team could have been estimated. Lambs also were marked with numbered ear tags at docking to obtain an estimate of the unlocated lamb mortalities. With the ranching practices used on the study area, round-ups were incomplete and the flock was handled in the corrals so rapidly that accurate counts of marked animals could not be obtained. If the ownership or jurisdiction of the flock were controlled by the research team the flock management practices could be altered to attain information pertinent to the study. Radio Telemetry Although the use of radio telemetry was responsible for locating relatively few carcasses in this study, it is potentially the most efficient means of locating sheep mortalities. Changes in transmitter design, which would allow a better type harness to be used, may be necessary to obtain higher success. In this study transmitters were attached to lambs by means of an elastic harness sewn to a canvas sack which contained the mortality transmitter. The elastic was necessary to allow for the rapid growth of the lambs between docking and marketing. A major problem occurred whenever a lamb kicked at its harness or its ear. The hind leg sometimes passed between a harness strap and the lamb's body, stretching the elastic. The high-stretch elastics initially used in harnesses would break whenever entanglement occurred and the transmitter would be dropped in the field. The stronger lowstretch elastic, selected to reduce the numbers of dropped transmitters, would not break whenever entanglement occurred, but lambs were often injured by prolonged entanglement. The wounds became infected and blown by flies, and caused loss of the lamb if medical treatment was not applied. Future studies to document lamb mortality should require a portion of the lambs be equipped with transmitters as soon after birth as possible. Ewes could be shed-lambed and released along with their lambs as soon as the lambs were strong enough to follow the ewes in the field. This procedure would enhance locating lambs during the period of greatest mortality. Predation This study verified that the problem of predation on domestic sheep, specifically that of coyotes, in southeastern New Mexico is real. The losses documented during 974 and 97 are similar to those reported by Mr. McKnight on a previous questionnaire survey for 972 and 973. Coyotes were responsible for 77 and percent of all predator losses for 974 and 97, respectively. In addition, all predator species were responsible for 37 and percent of all field mortalities for the corresponding two years. Furthermore, predators, which prey primarily on lambs in New Mexico, were responsible for the death of 6 and 2 percent of the lamb crop during the two years, respectively. Certainly this level of predation is not present on every sheep ranch in New Mexico [Howard 974]. Therefore, a blanket predator control program is not the solution to the problem but rather control applied on the individual problem-area basis. The level of predator control necessary to maintain the sheep industry will vary among and within geographical areas. Perhaps the best approach for either federal or state agencies is to concentrate on key problem areas at specific times during a year and whenever possible on problem individuals within the predator population. Losses to Other Causes Losses of sheep which are due to causes other than predation also need to be reduced. This reduction of losses possibly could be accomplished by incorporating better animal husbandry methods along with improved ranch management practices. The specific methods to be used would have to be tailored to the individual operation. The evaluation of the problems and their possible solutions could probably be handled best through the use of agricultural and wildlife extension specialists working directly with the individual ranch operators.

LITERATURE CITED Cain, S. A., J. A. Kadlec, D. L. Allen, R. A. Cooley, M. G. Hornocker, A. S. Leopold, and F. H. Wagner. 972. Predator Control-97. Report to the Council on Environmental Quality and the Department of the Interior by the Advisory Committee on Predator Control. Univ. of Mich. Press, Ann Arbor. Davenport, J. W., J. W. Bowns, J. P. Workman, and D. B. Nielsen. 973. Assessment of Sheep Losses, p. 3-7. In F. H. Wagner et al. Final Report to the Four Corners Regional Commission on Predator Control Study, Contract FCRC No. 62-366-44. Utah State Univ., Logan. Gier, H. T. 97. Coyotes in Kansas. Kansas State Univ. Agri. Exp. Sta. Bull. 393. Howard, V. W., Jr. 974. Sheep losses in southeastern New Mexico. Coyote Research Workshop. Denver, Colorado. Kilz, A. L., G. W. Corner, and R. E. Johnson. 973. A multiple-use wildlife transmitter. USDI. Fish and Wildlife Service. Denver, Colorado. Spec. Sci. Rep.-Wildl. No. 63. Maker, H. J., M. T. Turner, W. W. Gallman, and J. U. Anderson. 97. Soil Associations and Land Classification for Irrigation, Lincoln County. N. Mex. State Univ. Agri. Exp. Sta. Res. Rep. 22. Nesse, G. E. 973. Predation in the Sheep Industry in Glenn County, California. M.S. Thesis, Univ. of Calif., Davis. Rosko, L. 948. Losses of Sheep from Predatory Animals on Summer Ranges in Iron County, Utah. Utah Coop. Wildl. Res. Rep.

New Mexico State University's Agricultural Experiment Station publishes many bulletins and research reports of interest to residents of New Mexico. You may obtain a copy of the latest list of such publications by contacting the County Extension Office in your county, or by writing to: Bulletin Office Department of Agricultural Information New Mexico State University Drawer 3AI Las Cruces, New Mexico 883