IDAHO DEPARTMENT OF FISH AND GAME
Stephen P. Mealey, Director
Project W-160-R-24
Subproject 43
Completion Report
SEASONAL HABITAT USE, POPULATION CHARACTERISTICS, AND MANAGEMENT OF CHUKAR PARTRIDGE (ALECTORIS CHUKAR) IN WEST- CENTRAL IDAHO
Study I: Chukar Partridge Habitat Use and Management in West-central Idaho
Job 1: Implement the study plan
July 1, 1996 to June 30, 1997
By:
Andrew Lindbloom, Graduate Student, University of Idaho Kerry Reese, Professor, University of Idaho Pete Zager, Principle Wildlife Research Biologist, Idaho Dept. of Fish and Game
November 1997 Boise, Idaho Findings in this report are preliminary in nature and not for publication without permission of the Director of the Idaho Department of Fish and Game.
The Idaho Department of Fish and Game adheres to all applicable state and federal laws and regulations related to discrimination on the basis of race, color, national origin, age, sex, or handicap. If you feel you have been discriminated against in any program, activity, or facility of the Idaho Department of Fish and Game, or if you desire further information, please write to: Idaho Department of Fish and Game, 600 S. Walnut, Box 25, Boise, ID 83707; OR the Office of Human Resources, U.S. Fish and Wildlife Service, Department of the Interior, Washington, DC 20240.
TABLE OF CONTENTS
ABSTRACT ...... 1
INTRODUCTION...... 1
STUDY AREA ...... 3
METHODS ...... 4 Trapping ...... 4 Radio Telemetry ...... 6
CHAPTER I. Methods for Handling And Radio-marking Chukar Partridge ...... 8 INTRODUCTION...... 8 METHODS ...... 9 Transmitter Attachment ...... 9 Restraining Device ...... 9 RESULTS ...... 10 Transmitter Attachment ...... 10 Restraining Device ...... 10 DISCUSSION ...... 11
CHAPTER II. Seasonal Habitat Use and Selection of Chukar Partridge in West-central Idaho ..13 INTRODUCTION...... 13 METHODS ...... 14 Data Collection ...... 14 Statistical Analyses ...... 15 RESULTS ...... 18 Data Collection ...... 18 Statistical Analyses ...... 19 DISCUSSION ...... 21 MANAGEMENT IMPLICATIONS ...... 22
LITERATURE CITED ...... 24
Chukar Comp97 i LIST OF TABLES
Table 1.1. A comparison of observations of penned captive-reared chukars with necklace- and backpack-mounted transmitters, 1995 ...... 31
Table 2.1. Areas and percentages of spring and summer cover types measured for the Lower Salmon River study area, west-central Idaho, 1995 and 1996 ...... 31
Table 2.2. Densities of yellow starthistle for the Lower Salmon River study area, west- central Idaho, 1995 and 1996 ...... 32
Table 2.3. Physiographic characteristics of spring and summer habitat use locations for chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995-96 ...... 32
Table 2.4. Average, minimum, and maximum percentages of rock, shrub, grass, forb, and agriculture estimated for habitat use locations of each cover type for chukar partridge, west-central Idaho, 1995 and 1996 ...... 33
Table 2.5. Maximum-likelihood analysis-of-variance table for log-linear model of dependent variables gender, season, year, and cover for chukar partridge habitat use in west-central Idaho, 1995 and 1996 ...... 34
Table 2.6. Analyses of habitat use per cover type between spring and summer for chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995 and 1996 ...... 35
Table 2.7. Bonferroni simultaneous confidence intervals for spring habitat selection analysis of chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995 and 1996 ...... 36
Table 2.8. Bonferroni simultaneous confidence intervals for summer habitat selection analysis of chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995 and 1996 ...... 37
Table 2.9. Maximum-likelihood analysis-of-variance table for log-linear model of dependent variables sex, year, and density of yellow starthistle used by chukar partridge in west- central Idaho, 1995 and 1996 ...... 38
Chukar Comp97 ii LIST OF FIGURES
Figure 0.1. Approximate distribution and relative abundance of chukar partridge in Idaho (Idaho Upland Game Management Plan 1986-90) ...... 39
Figure 0.2. Study area for chukar partridge research located in the lower Salmon River canyon, west-central Idaho, 1995-96 ...... 40
Figure 1.1. Dimensions of restraining device for handling and radio-tagging chukar partridge ...... 41
Figure 1.2. Chukar partridge in restraining device ...... 42
Figure 1.3. Percentages of radio-tagged birds lost due to depredation, slipped collars, radio- caused mortalities, and other miscellaneous causes compared between chukars with necklace- and backpack-mounted transmitters, west-central Idaho, 1995 and 1996 ...... 43
Figure 1.4. Weight change from first capture of chukars collared with necklace-mounted radio transmitters in 1995, west-central Idaho ...... 44
Figure 1.5. Weight change from first capture of chukars collared with backpack-mounted radio transmitters in 1996, west-central Idaho ...... 45
Chukar Comp97 iii
COMPLETION REPORT STATEWIDE WILDLIFE RESEARCH
STATE: Idaho SUBPROJECT: Seasonal Habitat Use, Population PROJECT NO.: W-160-R-24 Characteristics, and Management of Chukar SUBPROJECT NO.: 43 Partridge (Alectoris chukar) in West-central Idaho STUDY NO.: I STUDY NAME: Chukar Partridge Habitat Use JOB NO.: 1 and Management in West-central Idaho
PERIOD COVERED: July 1, 1996 to June 30, 1997
ABSTRACT
Field work was completed during FY 97. Data analysis is nearly complete and 2 chapters of the final report are attached. The remainder will be completed and submitted by January 1, 1998.
INTRODUCTION
Native to southern Europe and Asia, chukar partridge (Alectoris chukar) were first introduced into the United States in 1893 (Cottam et al. 1940). Chukars filled what was considered an empty niche in many arid to semi-arid regions characterized by steep terrain, rocky outcrops, talus slopes, and introduced annuals. Chukars were first introduced into Idaho in 1933 (Christensen 1970), and 25 counties now have viable chukar populations. Introductions are believed to have occurred in all suitable habitats throughout the state (Bizeau 1963) and the chukar partridge has become 1 of Idaho's most important upland game bird species.
The present distribution of chukar partridge in North America is from southern interior British Columbia southward through eastern parts of Washington, Oregon, and California to the northern part of Baja California, and east in the Great Basin uplands through Nevada, Idaho, Utah, western Colorado, and Montana (Johnsgard 1973). Small populations of uncertain status exist in Nebraska, Texas, Arizona, New Mexico, western South Dakota, and southern Alberta (Christensen 1970, Johnsgard 1973). The present range of chukars in Idaho is from Latah County in northern Idaho, to Owyhee County in southwestern Idaho, to Caribou County in southeastern Idaho (Fig. 0.1). The denser populations of chukar partridge occur in the southwestern and west-central part of the state (Idaho Upland Game Plan 1986-90).
The first hunting season in Idaho was initiated in 1953. Hunting seasons have been traditionally liberal and daily bag limits of 10 and possession limits of 20 were normal during the 1950s and 1960s. Currently, bag and possession limits are 8 and 16, respectively, with a season length from mid-September to late December. Season lengths and limits are believed to have a negligible effect on populations due to relatively low harvest rates caused by the inaccessibility of areas that chukars inhabit (Harper et al. 1958, Christensen 1970).
Chukar Comp97 1 Concern over Idaho's chukar populations has increased in the past decade. Harvest estimates from 1963 to 1981 range from a low of 123,000 birds to a high of 228,700 (Idaho Fish and Game News 1993). From 1982 until present, harvest estimates have not exceeded 100,000 birds annually. Harvest estimates of upland game birds will vary in any state from year to year, either due to errors in the estimation process or natural population fluctuations. Populations fluctuate due to varying natural regulating factors such as weather, predation, or food availability. Due to the popularity of the chukar partridge as a hunted species, however, wildlife biologists are often challenged to reduce these fluctuations and to manage populations for higher numbers. In order to do this, managers need to know more about the basic ecology of the chukar partridge.
Although a notable amount of information is available in the literature regarding the chukar partridge, certain fundamental ecological data are lacking. Chukars in the Negev Desert of Israel have been studied to reveal responses of dietary water content (Degen et al. 1984), diet and water turnover rates (Alkon et al. 1985), and time-energy budgets (Carmi-Winkler et al. 1987). Though important from a biological perspective, these studies provide little assistance to the wildlife manager. Galbreath and Moreland (1953), Bohl (1957), Harper et al. (1958), Christensen (1970), and Molini (1976) provide details on introductions, life history, and management of the chukar partridge. These early reports have been valuable in providing some basic information about chukars, but much of the data are only anecdotal. In addition, these reports suggest that more research on chukar partridge is needed.
It is imperative that biologists understand and predict how management practices affect wildlife populations. Despite overwhelming interest in the chukar as a popular game bird, limited research has been conducted to understand the biology and ecology of the species. To obtain further knowledge of chukar partridge and provide more effective management practices, habitat use, movements, home range, nest site characteristics, nesting chronology, productivity, and survival/depredation rates of chukar partridge were investigated along the Lower Salmon River, west-central Idaho. Chukars were trapped and radio-tagged during winter and monitored during spring and summer. In Chapter 1, the methods used to radio-mark chukar partridge were reported. The applicability of using a restraining device developed to handle wild birds and the comparison of backpack- and necklace-mounted transmitters was also investigated.
In Chapter 2, habitat use for both spring and summer months was examined. The observed use of habitats relative to the availability of these habitats was analyzed to assess selection. Physical characteristics of use locations such as slope, aspect, elevation, percent of cover type, and percent of vegetative species were reported. In addition, the use of habitats was compared with the abundance and distribution of yellow starthistle (Centaurea solstitialis), an exotic which is increasingly abundant in chukar habitats.
Nest site characteristics and reproduction will be examined in Chapter 3 which will be included in the final report. Micro-habitat nest measurements included visual obstruction, overhead canopy cover, and ground cover; macro-habitat measurements were the same as those mentioned for Chapter 2. Comparisons were made between first and second nesting attempts, and
Chukar Comp97 2 successful versus unsuccessful nests. Nesting chronology, productivity, and brood habitat use were also recorded.
Chapter 4, to be completed, will contain movement and survival data for chukar partridge in west-central Idaho. Elevational changes between spring and summer locations, and spring- summer home range sizes were reported and discussed. Overall spring-summer survival rates and seasonal survival rates were calculated and compared; depredation by avian and mammalian predators was also examined.
STUDY AREA
The study area was located in the western portion of Idaho County, in west-central Idaho, approximately 14 km south from the town of Cottonwood (Fig.0.2). Trap sites were centered primarily on Rock Creek and its confluence with the Lower Salmon River. Boundaries of the study area were delineated by chukar movements and natural physiographic barriers. The upper boundary extended approximately 100 m into the agricultural fields of the Camas prairie, a biological barrier, while the lower boundary was defined as the Lower Salmon River. Side boundaries of the study area were established 1 ridge past the farthest location of a radio-marked chukar. Area of study was approximately 2036 ha (5030 acres).
The delineation of study area boundaries will affect the proportional distribution of habitat. Porter and Church (1987), however, showed that where habitat dispersion patterns are regular or random, the impact of study area delineation on inferential analysis is unimportant; such was the case with this study.
The general climate of the lower Salmon River region was semi-arid, characterized by hot, dry summers and fairly mild winters with only limited and ephemeral snow in the valley bottoms (Tisdale 1986). Tisdale (1986) reported mean temperatures from the Cottonwood weather station for January and July of -3.3 C and 18.6 C, respectively. Annual precipitation was reported at 504 mm.
Elevations of the study area ranged from approximately 402 m (1320 feet) at the Salmon River to 1108 m (3634 feet) at the top of Pine Tree Gulch. The dominant surface rock formation was Columbia River basalt. Rapid cutting of valleys following regional uplift and vigorous stream flow resulted in over-steepening of the canyon sides, so that slopes of 45 to 75% predominated (Tisdale 1986). Weathering of the Columbia River basalt produced a series of vertical cliffs and talus slopes. Most of the soils developed from a mixture of residual and colluvial materials, and contained a high proportion of gravel and stone (Tisdale 1986). Intermittent springs, creeks, and livestock watering ponds were interspersed infrequently throughout the area.
The natural vegetation of the study area was developed from the Pacific Northwest region flora and was strongly dominated by bunch grasses (Horton 1972). Tisdale (1986) reported plant communities characterized by bluebunch wheatgrass (Agropyron spicatum), Idaho fescue (Festuca idahoensis), and hood sedge (Carex hoodii), respectively, occupied most of the
Chukar Comp97 3 grassland area in west-central Idaho. Sand dropseed (Sporobolus cryptandrus) and red threeawn (Aristida longiseta) were reported occurring at low elevations. Small inclusions of shrub-grass types were dominated by stiff sagebrush (Artemisia rigida), snowberry (Symphoricarpos albus), smooth sumac (Rhus glabra), curlleaf mountain mahogany (Cercocarpus ledifolius) and netleaf hackberry (Celtis reticulata) (Tisdale 1986). Invasion of exotic annuals like cheatgrass (Bromus tectorum) and perennial forbs such as yellow starthistle occurred on many depleted sites.
Canyon and plateau portions of the study area were mostly in private ownership. Riparian habitats of the Salmon River were administered by the Bureau of Land Management and a section along Rock Creek was owned by the State of Idaho.
Livestock grazing was the primary land use of the study area. Range use from 1890 through 1940 was heavy, with stocking rates many times greater than what is presently considered sustainable (Tisdale 1986). As a result, most easily accessible portions of the lower Snake River valley were overgrazed and their vegetation greatly altered (Tisdale 1986). Cultivation for crop production was minor in the study area. Only small portions along the river valley bottoms had topography suitable for farming. The plains area above the river canyons, adjacent to the study area, was extensively farmed for wheat.
METHODS
Trapping
Beginning in mid-January of 1995, the study area was surveyed to determine chukar population status. Surveying was accomplished by hiking through the area with a Labrador retriever. Areas that repeatedly produced birds were mapped for potential trap sites.
Traps were placed on areas that were believed to be frequently used by chukars. Chukars were trapped from January to May of 1995 and 1996 using modified clover leaf walk-in traps baited with a mixture of cracked corn, whole wheat, sunflower seeds (black oil), and dried peas. Chicken wire leads, approximately 1 m height and 10 m length, were extended from the entrance of the traps to help funnel the birds toward the trap. Leads were also used to connect multiple traps in order to cover large areas with few traps. Some circular and rectangular traps with funnel entrances were used, but no birds were captured in these traps. Twenty-one winter/spring traps sites were established in 1995 and 54 established in 1996. Furthermore, 3 summer trap sites were also established in 1995 but were not successful.
A spot-lighting technique was developed and tested for the capture of chukar partridge on night roosts. The technique involved the use of a portable, hand-held spotlight and a fishing net with a long, extended handle. Various cliffs and talus areas of known concentrations of birds were hiked at night in an attempt to find roosting birds. The method involved 1 person shining the spotlight in the bird's eyes, while another person approached from the side and placed the net over the bird. Six attempts were made, but locating the birds at night proved to be unsuccessful. The cover was either sufficiently dense or uneven that the birds were not spotted until they
Chukar Comp97 4 flushed. Five attempts were also made to spotlight radio-collared birds with the same results. No birds were successfully captured using this method.
In 1995, traps were initially checked twice daily (in the morning and the evening). However, because chukars were frequently flushed near the immediate vicinity of the traps, traps were later checked only once daily. Because this still continued to disturb birds in the vicinity of traps, subsequent checking of traps occurred only at night. Very few birds were flushed at night and trapping success increased.
In 1996, all traps were checked at night from the start of the field season. However, after checking the traps the following morning on several occasions, it was apparent that rodents were eating all the bait during the night. From observations in the field and numerous comments from local landowners, the rodent population appeared to have increased in 1996. Because this left no bait to lure the birds during the day, traps were checked early in the morning usually starting at 3 or 4 a.m.
Twenty-three chukars, 9 females and 14 males, were trapped during the 1995 field season. Twenty-two birds were radio-collared; 1 chukar was found depredated in the trap. Thirty-three chukars, 14 females and 18 males, were trapped in 1996. One bird was depredated in the trap and sexing was not possible. In addition, 2 females died from handling stress and 1 male was pecked to death by a larger male. This left 29 birds to be radio-tagged in 1996; a total of 51 birds radio-tagged from both field seasons. In addition, several Meadowlarks (Sturnella neglecta), Rufous-sided Towhees (Pipilo erythrophthalmus), Gray Partridges (Perdix perdix), American Robins (Turdus migratorius), 1 Northern Flicker (Colaptes auratus), and 112 Black-billed Magpies (Pica pica) were incidentally trapped and released.
Necklace-mounted radio transmitters, manufactured by AVM Instrument Company, Ltd. (Livermore, CA), were used on all birds in 1995. Because of numerous complications with necklace mounts (see Chapter 1), backpack-mounted transmitters (AVM Instrument Co.) were used in 1996. Besides weighing approximately 3-g more, these transmitters had identical specifications to the necklace-mounted transmitters. To enable 1 researcher to attach the backpack transmitter, a modified quail restraining device (Demaso and Peoples 1993) was used successfully with all the birds in 1996 (see Chapter 1).
Due to the lack of plumage dimorphism exhibited by male and female chukars, several measurements were taken to distinguish sex. Gender was determined primarily from shank length measurement, which is considered 93% accurate (Woodard et al. 1986). Shank lengths were determined by measuring (using calipers) the distance from the foot pad to the top of the hock joint in legs flexed at 90 degrees from the tibia. Birds with shank lengths >60 mm were classified as male, and those <60 mm as female. Because of the possibility of error in this measurement, when shank lengths were within 0.3 mm of the 60 mm mark, I also used the following measurements to assist in sex determination (listed in order of importance; Appendix A):
Chukar Comp97 5 1) weight: Weights of male chukars should be between 536 and 729 g, and females between 462 and 550 g (Christensen 1954). Weights were measured by placing the bird in a cloth sack and suspending this from a 1000-g Pesola scale.
2) spur-tarsus diameter (Cunningham 1959): The diameter of the spur-tarsus in males should be between 7.5 to 11 mm, and females between 6 and 9 mm. Spur- tarsus diameters were measured from the point of the spur (or knob) to the opposite side of the tarsus by use of calipers. In birds without spurs, the diameter of the tarsus was measured.
3) middle toe length (Cunningham 1959): Middle toe lengths for males should be between 37 and 46 mm, and females between 34 and 43 mm. Toe lengths were measured by placing a millimeter ruler against the web between the toes and measuring the total length of the middle toe.
Chukars that had no distinguishing shank lengths were usually sexed by examination of weight. If weight was in the overlap range between males and females, then spur-tarsus diameter was used to assess sex. Middle toe lengths were not considered representative.
Radio Telemetry
Birds were located approximately weekly using a Telonics Scanner/Receiver (Telonics, Inc., Mesa, AZ) and a hand-held 3-element Yagi antennae using the loudest signal method (Springer 1979). In order to prevent possible bias in locations due to time of day, I attempted to locate birds evenly throughout 3 diurnal time periods: (1) sunrise to 1000 hours, (2) 1001 to 1400 hours, and (3) 1401 hours to sunset. Locations were recorded on 7.5 minute topographic maps and later transferred to data sheets using Universal Transverse Mercator Coordinates (UTM). I recorded cover type, percent cover type, common vegetative species, density of yellow starthistle, general rock formation, slope, aspect, elevation, time, covey size, and miscellaneous comments for every chukar location.
Large error polygons typical of performing telemetry in rugged terrain (Hupp and Ratti 1983) may yield erroneous location data. Because of the preponderance of steep topography and rock cliff faces throughout the study area, and the resulting radio-wave bounce, numerous attempts to locate birds by triangulation failed. Hence, most attempts to gather location data resulted in flushing the bird. If birds held tight and did not flush, such that I could hike an approximate 20- m radius circle centered at the bird location, no attempt was made to flush the bird.
During the nesting season, efforts were made to prevent disturbance by keeping greater distances between me and the chukars. Almost all birds held tight when nesting, so nest sites were obtained by circling the bird. I measured general nest characteristics, such as clutch size and stage of incubation (Westerkov 1950), when the bird was off the nest. Harper et al. (1958) found nests by observing hens feeding from daylight to 8:00 a.m. and from 5:30-6:30 p.m. Attempts were made to search for nests during these time periods, assuming the bird had left the nest site.
Chukar Comp97 6 Macro- and micro-habitat characteristics were measured only after the nest was hatched, abandoned, or depredated.
Precautions to prevent disturbance were also taken during the brood rearing season. Most brood locations gathered within the first few weeks after hatch were obtained by circling the bird because chukars with broods not capable of flight were very reluctant to flush. Often these sites were flagged and returned to later for measurements in order to minimize human presence near brood. When broods reached flight capability (approximately 2 weeks), hens and broods were flushed with a Labrador retriever in order to obtain accurate brood counts.
Chukar Comp97 7 CHAPTER I.
METHODS FOR HANDLING AND
RADIO-MARKING CHUKAR PARTRIDGE
INTRODUCTION
Upland game birds pose many challenges to wildlife biologists using radio telemetry to research population characteristics. Numerous radio attachment techniques have been employed on different species with varying degrees of success. Cochran (1980) suggested that the number and variation of attachment methods for birds is an indication of the uncertainties inherent in attaching a transmitter to an animal so greatly dependent for its survival upon the uninhibited use of its environment.
The diversity in physiognomy, size, and behavior among birds limits generalizations on the attachment of telemetry devices (Samuel and Fuller 1994). Few studies have been conducted that examine transmitter attachment on chukar partridge. Results from an investigation conducted by Slaugh et al. (1989) favored the use of backpack-mounted transmitters instead of ponchos. The ponchos, however, weighed 30 g or 5.9% of an average 506 g female chukar. Mating complications that could be associated with backpack antenna angle were reported negligible on captive-reared birds (Slaugh et al. 1990).
The attachment of transmitters on collars is a common and successful method for several species of upland game birds such as ring-necked pheasants (Phasianus colchicus; Marcstrom et al. 1989), sharp-tailed grouse (Tympanuchus phasianellus; Amstrup 1980, Gardner 1997), sage grouse (Centrocercus urophasianus; Amstrup 1980, Fischer et al. 1997), and bobwhite quail (Colinus virginianus; Burger et al. 1995). Because of the extensive use and success of collars on upland game birds, it was assumed collars would likewise be compatible for chukar partridge. In 1995, chukars were radio-marked with necklace-mounted transmitters as part of a study on the ecology of chukar partridge in west-central Idaho. Because of numerous complications associated with necklace mounts, backpack-mounted transmitters (Slaugh et al. 1989) were used the second year of the study.
The implications of using necklace- and backpack-mounted transmitters on chukar partridge are herein reported. Field observations of wild birds and records from an experiment on captive-reared birds were compared. In addition, the use of a restraining device for handling and radio-tagging chukars was investigated.
Chukar Comp97 8 METHODS
Transmitter Attachment
In 1995, 22 birds were trapped and radio-tagged with necklace-mounted radio transmitters. Manufactured by AVM Instrument Company, Ltd. (Livermore, CA), each transmitter was equipped with a 12-month battery and mortality sensor and weighed approximately 10.8 g [(2.1% of average 506 g female (Christensen 1954)]. The necklace collar was plastic-coated copper wire inserted in a sleeve of surgical tubing. The copper wire was extended through an open tube in the transmitter and secured with a plastic crimper.
In the summer of 1995, an experiment was conducted to more closely observe complications associated with necklace-mounted transmitters and to investigate the use of backpack-mounted transmitters. Nineteen captive-reared chukars were placed in an outdoor pen; 6 with necklace- mounted transmitters, 6 with backpack-mounted transmitters, and 7 uncollared as a control. Birds were observed approximately 2-3 times daily for the first week, and approximately 2 times per week for the following 1.5 months.
Because both captive and wild birds failed to acclimate well to necklace-mounted transmitters, all birds trapped in 1996 were radio-tagged with AVM wingstrap backpacks (Slaugh et al. 1989). These transmitters weighed slightly more than necklace-mounted transmitters (average 14 g or 2.8% of the average female), but otherwise had identical specifications. Parachute cord (3 mm diameter) was used to attach the transmitter to the chukar. The ends of the cord were crimped together with a 3/8-inch (9.5-mm) hogring (Slaugh et al. 1989) and covered with heat-shrink tubing (Edelmann pers. comm., Schumacher et al. 1978).
Restraining Device
To enable 1 researcher to attach a backpack transmitter, a restraining device for handling chukars was developed and tested. Design and construction were based on slight modifications of a quail restraining device constructed by Demaso and Peoples (1993). The device was similar to a small folding table with a 2-1/4-inch (5.8 cm) diameter hole cut in the middle (Fig.1.1). The top platform was made of 1/4-inch (6.4 mm) wafer board and the legs of 1-inch (2.6 cm) hollow square aluminum tubing. The legs were hinged to the top board so that the device could be folded flat and slid into a backpack. The center crossbar was attached with a hitch pin on each side, and was removable. Once a bird was in hand, its legs were secured together with a single buckle leather strap and slid through the hole (Fig. 1.2). The strap was next fastened to the bottom crossbar. An additional nylon strap was secured over the upper legs and attached to 1 of the table legs as a safety strap.
Chukar Comp97 9 RESULTS
Transmitter Attachment
Of the 22 birds collared with necklace-mounted transmitters between February and May of 1995, only 3 persisted to the end of the field season in mid-August. Two of these 3 birds were recaptured the following field season, both without collars. Nine of the 22 radio-collared birds (41%) were able to slip the collar over their heads, or break the attachment (Fig. 1.3). Four birds (18%) died because of predation. In addition, 3 birds (14%) died due to the attachment of the radio; these birds were found dead, but the entire carcasses were intact. One transmitter slipped below the crop, apparently preventing food from passing any further down the digestive tract. The bird died emaciated with a full crop.
Of the 29 birds radio-marked with backpack-mounted transmitters in 1996, 7 persisted to the end of the field season. Fifteen birds (52%) died because of predation (Fig. 1.3). No birds slipped the transmitter or died because of the attachment.
Changes in body weight between capture events and the number of recaptures were also examined as potential indicators of transmitter effects. Chukars with necklace-mounted transmitters lost an average of 21.1 g [n = 9, SE = 41.1, range = (-95, 35)] between first and second capture events, whereas chukars with backpack-mounted transmitters lost an average of 1.8 g [n = 9, SE = 29, range = (-45, 40)]. The difference in weight loss between chukars with necklace vs. backpack transmitters, however, was not significant (P = 0.27, t = 1.15, df = 16). The number of recaptures was greater for birds with necklace-mounted transmitters than those with backpacks (Fig. 1.4 and 1.5). Six birds with necklaces were captured >2 times and 1 was captured 9 times; 1 bird mounted with a backpack was captured >2 times (5 captures).
Observations of penned birds (Table 1.1) indicated that backpacks were more compatible with chukar partridge. Immediate response from birds collared with necklaces occurred with 5 of the 6 birds (83%); birds flipped in the air, walked backwards pecking at necklace, and/or laid on their backs scratching at transmitter. No immediate response was observed from birds with backpack-mounted transmitters. Other responses observed included (1) pecking at transmitter by both bird wearing the transmitter and other birds, (2) transmitter slipping off balance or to the side (askew), (3) preening around transmitter, (4) bird walking backwards for no apparent reason, (5) eye irritation from antenna, and (6) bird removing or slipping transmitter off.
Restraining Device
The restraining device was used with 19 captive-reared birds and 29 wild chukars and proved to be successful in restraining the chukars while mounting backpack transmitters. A few birds showed resistance to the device by excessive struggling and wing flapping, but were easily subdued by placing a black cloth hood over their heads.
Chukar Comp97 10 DISCUSSION
Most species require restraint, sedation, or anesthesia while being radio-tagged (Samuel and Fuller 1994). The successful modification of a restraining device (Demaso and Peoples 1993) for chukar partridge should improve the care given to chukar partridge during research, and may also reduce costs by allowing 1 researcher to handle and radio-tag animals.
Previous studies suggest transmitter weights should not exceed 5% of body mass and transmitters weighing >10% should not be used on birds released in the field (Caccamise and Hedin 1985, A.O.U. 1988). Samuel and Fuller (1994) reported that many species seem to tolerate packages that are 4% of their body weight, but distribution of weight and a variety of other factors may be more important. Given that necklace- and backpack-mounted transmitters were only 2.1% and 2.8% of the average female body weight, respectively, it is unlikely weight was a detrimental factor of transmitter attachment.
Methods of attachment are dependent upon the body form of the animal under investigation. Knowledge of anatomical and behavioral function, along with trial and error, are necessary for development of optimal attachment methods (Cochran 1980). Numerous species known to acclimate well to poncho or necklace-mounted transmitters (e.g., ring-necked pheasant, sharp- tailed grouse, sage grouse) have long necks and large heads. Chukars differ from many upland game birds because their necks are short and tapered, greatly limiting the applicability of attachment around the neck. In addition, Slaugh et al. (1989) noted the intolerance of chukar partridge to any type of ventral attachment.
Results from this study are in agreement with experiments conducted by Slaugh et al. (1989); wingstrap backpack-mounted transmitters are the preferred method of attachment for chukar partridge. Concerns regarding this attachment method, however, developed from observed 1) feather and skin loss that resulted from attaching wing-straps, 2) time spent preening around the transmitter, and 3) predation rates. Greenwood and Sargent (1973) noted that chafing and feather loss may result in reduced insulation for waterfowl. Slaugh et al. (1989) observed minor feather loss that was attributed to loose attachment of transmitter. Both feather and skin loss were observed during this study, regardless of harness tightness. Whether this could negatively impact chukars needs further investigation.
Perry (1981) reported that canvasback ducks marked with backpacks spent an inordinate amount of time on shore picking at the transmitter, which eventually resulted in significant weight loss. Calculations of weight change between first and second captures were used to investigate whether chukars were exhibiting the same behavior (Fig. 1.5). The average change in weight for birds with backpack-mounted transmitters was -1.8 g, suggesting that backpacks do not cause significant weight loss.
Differences between predation rates of birds with necklace versus backpack mounts is probably of little value; few birds with necklaces were available to predation because of other losses and hence depredation rates are not comparable to birds with backpacks. The observed predation rate
Chukar Comp97 11 of 52% of chukars with backpack-mounted transmitters, however, may be of some concern. Some birds were noted to flush last in a covey or appeared to resist flight. Behavior of such may increase predation. Unfortunately this was not quantified, however, and many radio-tagged birds appeared to behave similar to birds without collars. Numerous studies on upland game birds and waterfowl have suggested negative impacts of backpack-mounted transmitters (Hessler et al. 1970, Boag 1972, Greenwood and Sargeant 1973, Lance and Watson 1977, Erikstad 1979, Johnson and Berner 1980, Warner and Etter 1983, Small and Rusch 1985, Marcstrom et al. 1989). Further research should be conducted to more thoroughly address the impacts of backpacks on chukar partridge survival.
Chukar Comp97 12 CHAPTER II.
SEASONAL HABITAT USE AND SELECTION OF CHUKAR
PARTRIDGE IN WEST-CENTRAL IDAHO
INTRODUCTION
Habitat use is a critical facet in the management of a wildlife species (White and Garrott 1990). Estimates of habitat use and selection for chukar partridge are essentially nonexistent. Most habitat use records do not expand beyond a general description of cheatgrass-sagebrush, steep topography, rock outcroppings, talus slopes, and arid climate. The most detailed description was probably given by Galbreath and Moreland (1953), who specified optimum chukar habitat as the following: about 50% sage-cheatgrass-bunchgrass; 45% talus slopes, rock outcroppings, cliffs and bluffs; and 5% brushy creek bottoms and swales.
Seasonal habitat use among spring and summer is understood even less. Food habits and general movements during different seasons have been documented, but the proportional use of habitats relative to availability has not been studied. Galbreath and Moreland (1953), Harper et al. (1958), and Christensen (1970) suggested that habitat use during the summer is greatly influenced by the availability of water. Chukars can be found concentrated at water sources during early morning and late evening hours. During winter months, snow depths are believed to limit habitat use of chukars (Alcorn and Richardson 1951, Christensen 1970, Kuz’mina 1992). Wind-swept slopes and cultivated fields are where most chukars will be found in winters of deep snowfall. However, macro- and micro-habitat characteristics specific to spring and summer are not known. Whether use is significantly different between sexes and seasons has not been examined. In addition, knowledge of habitat use specific to Idaho is needed.
The impacts of invading yellow starthistle in chukar habitats are also unknown. Yellow starthistle is an introduced annual knapweed from Europe that has infested nearly 3 million acres (1,215,000 ha) in Idaho, California, and Washington (Callihan et al. 1989). Eradication by chemical application is possible but expensive and used mainly in cultivated areas. Biological control, such as introduction of native insects from the Mediterranean, is still in the experimental stages (R. Callihan, pers. comm.). Limited occurrence of chukars in or near yellow starthistle in Idaho has been reported during annual helicopter surveys (C. Johnson, BLM Area Biologist, pers. comm.), but actual use is unknown. No record exists of yellow starthistle in chukar diets. If this is indicative of no use, much of the chukar habitat in Idaho is in jeopardy.
This study addressed habitat use and selection through a 2-year radio-telemetry investigation. The following null hypotheses were tested: 1) habitat use by chukars does not differ between spring and summer; 2) habitat use by chukars does not differ between males and females; 3)
Chukar Comp97 13 chukars use cover types in proportion to their availability; and 4) chukars use areas dominated by yellow starthistle in proportion to their availability.
Habitat use was defined as the quantity of habitat used by the animal in a fixed period of time (Johnson 1980); habitat use was estimated by the proportion of radio locations within each cover type. Habitat selection was defined as the process in which an animal actually chooses a particular habitat (Johnson 1980, Peek 1986). Habitat use was considered selective if habitats were used disproportionately to their availability. According to Johnson (1980), a natural hierarchical ordering of selection processes can be identified. Second order selection, the determination of home range (macro-habitat) characteristics of an individual or social group within the geographical range, was the scale of selection investigated.
METHODS
Data Collection
Radio-marked chukars were located approximately weekly from April to August. Each locale of use consisted of the area within a 10-m radius circle centered at locations of radio-marked birds and from incidental, unmarked birds. Locations from unmarked birds occurred during initial surveys of the canyon and from pursuit of radio-tagged chukars. Slope, aspect, elevation, cover type, percent of each cover type, common vegetative species, and density of yellow starthistle were recorded for all habitat use locations. Because of the circular nature of aspect data, measurements were categorized into 4 quadrants: 1) Northeast (0-90°), 2) Southeast (91-180°), 3) Southwest (181-270°), or 4) Northwest (271-360°).
Four cover types were identified for the habitat use and selection analysis: 1) rock (talus, outcrop, cliff), 2) shrub, 3) grass/forb, and 4) agriculture. Because most locations contained more than 1 vegetative and/or physical characteristic, determination of cover type was based on the percentage of the cover types present inside the 10-m radius circle. Agricultural cover types were characterized by areas containing 50% or more agricultural crop, while shrub cover types contained 20% or more shrubs. Rock cover types were characterized by areas containing 20% or more rock cover, but less than 20% shrub. Grass/forb cover types were characterized by areas containing less than 20% rock or shrub, with grass and forbs making up the highest percent cover for the location.
The percent of cover types found within each location was most often determined from visual estimation. Due to potential errors in visual estimations, the percentages of cover types at numerous bird locations were measured using the line intercept method (Canfield 1941). These measurements were used to increase the accuracy of estimations and were completed throughout the seasons in order to sample plants in various phenological stages (about 15 line intercepts completed each year). Line intercepts indicated visual estimations were usually within 5-10% of the actual measurement.
Chukar Comp97 14 To assess cover type availability, all cover types were delineated onto 7.5-minute-series ortho- photoquads from interpretation of aerial photographs. Cover type availability was measured by overlaying the ortho-photoquads with 100-dot per square inch dot grids. The number of dots per cover type was calculated to estimate the area (ha) and proportion of the total study area occupied by each cover type (Table 2.1).
The availabilities of yellow starthistle densities were determined during the peak bloom (late July). Two categories of density were measured: 1) 0% < starthistle < 5%, and 2) starthistle > 5%. These categories were delineated on aerial photographs by walking and driving the study area and using a spotting scope. Data were later transferred to ortho-photoquads and area per density category was measured using dot grids (Table 2.2).
Two cover types, trees and Conservation Reserve Program (CRP) fields, accounted for 0.5% and 2.1% of the total study area respectively. Habitat use locations, however, did not occur in either cover type. Thomas and Taylor (1990) recommended that when 1 or more cover types are commonly available but rarely used the analyses should be presented including and excluding these cover types. However, these cover types were neither used nor commonly available. The inclusion of cover types that were not used but available would result in a more significant chi- square goodness of fit test. Influences on the Bonferroni tests would be an increased number of categories that would affect the upper standard normal table value. The resulting effects would be wider confidence intervals, which could influence significance. Recalculations of the chi- square value and confidence intervals, however, resulted only in trivial numeric changes, none of which affected significance. Therefore, trees and CRP were not included in the analyses.
Statistical Analyses
The pooling of homogeneous data may result in a more powerful analysis (Zar 1996). However, as demonstrated by Schooley (1994) and Zar (1996), misleading inferences may result from analyses performed on pooled heterogeneous data. Therefore, before habitat use and selection were analyzed, habitat use data were tested to evaluate the appropriateness of pooling between (1) radio-marked bird locations (marked) and incidental locations (unmarked), (2) males and females, (3) 1995 and 1996, and (4) spring and summer. All statistical analyses were conducted at α = 0.05 level of significance.
Chi-square homogeneity analyses were first conducted to examine the data for differences between marked and unmarked bird locations. Because gender was unknown for unmarked locations, differences between male and female locations were not examined in this analysis. Chi-square analysis tested the null hypothesis that habitat use data between years, seasons, and marked/unmarked locations was homogeneous.
To assess differences in habitat use between sexes, log-linear models were constructed without the unmarked data. A categorical data modeling procedure (PROC CATMOD; SAS Institute, Inc. 1990) was used to construct log-linear models that allowed for the examination of 2-, 3-, and 4-way interactions between cover type, year, season, and sex. Due to insufficient sample size,
Chukar Comp97 15 this method was not used to test between marked and unmarked data; however, chi-square homogeneity tests did allow for 2-way interactions analysis. Because homogeneity analyses and log-linear models both suggested that habitat use differed between seasons, Z-tests for comparing 2 binomial proportions were constructed to determine which cover types were used differently.
Chi-square goodness of fit tests were used to assess whether proportional use of cover types during spring or summer differed significantly from proportional availability in the study area (Neu et al. 1974). Because the test involved pooled data, the Yates correction for continuity factor was used (Zar 1996). Sample size must be sufficiently large to allow for a chi-square approximation for the goodness of fit test statistic. Dixon and Massey (1969) recommended > 1 expected observation in each cover type and that < 20% of all categories contain < 5 expected observations. These expected frequencies were met and exceeded.
Bonferroni simultaneous confidence intervals were constructed to detect which cover types were used less than expected, more than expected, or not significantly different than expected (Neu et al. 1974, Byers et al. 1984). The following sample size requirements of the Bonferroni tests were met: N(pi) and N(1 - pi) must both be ≥5, where N is total number of locations and pi is the proportion of observations in the ith habitat (Dixon and Massey 1969). Cherry (1996) recommended 2 alternative methods to the Bonferroni confidence intervals that compute intervals for multinomial proportions. Differences between the Bonferroni method and the alternatives were most extreme when sample sizes were small and the number of categories large. Given the small number of habitat categories (k = 4) and adequate sample sizes in this study, however, the traditional method of Bonferroni is comparable to the alternative methods.
In order to address the possible avoidance of starthistle, a categorical data modeling procedure (PROC CATMOD; SAS Institute, Inc. 1990) was used to construct log-linear models that assessed 2- and 3-way interactions between use of yellow starthistle density categories, year, and gender. Because the thistle does not attain full growth (and spines) until summer, use and availability were measured and analyzed only for this season. A chi-square homogeneity analysis was conducted to test for differences in yellow starthistle use between locations of marked and unmarked birds. A goodness of fit test was used to test proportional use and availability of various densities of yellow starthistle. Bonferroni confidence intervals were constructed to assess where, if any, differences existed between observed and expected use. Sample size requirements for goodness of fit tests and Bonferroni tests were met.
Several factors were considered for choosing the above methods to analyze habitat selection. The methods of weighting observations, assumptions, and characteristics of my data set were most influential in this decision (Alldredge and Ratti 1992). The Neu et al. (1974) method gives equal weight to each observation on each animal. Animals are not weighted equally but in accordance to the number of locations obtained from that individual. Habitat use data from animals with higher numbers of locations is more accurate and hence more representative of the population than those with only a few locations. If differences in numbers of locations per bird were extremely different, as presented by Alldredge and Ratti (1992), aberrant behavior by some
Chukar Comp97 16 individuals could bias the results. Given the minor differences observed in my data set, however, I do not believe this to be an issue.
The analysis of habitat selection and the numerous statistical methods used requires many assumptions. The chi-square goodness of fit test and related multiple comparisons (Neu et al. 1974, Byers et al. 1984) requires the assumption that availabilities are known constants, not estimated quantities (Thomas and Taylor 1990). Because resource availability was partitioned into cover types based on aerial photographs and ortho-photoquads, this assumption was met (Thomas and Taylor 1990, White and Garrott 1990). In addition, the following assumptions were met: 1) the animal has an opportunity to select any of the habitat which is deemed available; 2) observations are collected in a random, unbiased manner; and 3) habitat availability is the same for all animals (Neu et al. 1974, Alldredge and Ratti 1986). Furthermore, the possible bias related to visibility of animals relative to cover (Neu et al. 1974, Thomas and Taylor 1990) was not a problem. Radio-telemetry eliminated this bias for radio-marked animals, and the use of a bird dog substantially reduced any potential bias in flushing unmarked birds.
Yet another assumption to be met is that observations for 1 animal are independent of observations for other animals. All covey locations were counted as only 1 habitat location; therefore, only the assumption of independence of covey groups was made. If more than 1 radio- tagged chukar was in a covey, data were recorded for only 1 randomly selected bird (Alldredge and Ratti 1992). Animal locations were also considered to be independent of each other because locations were made only once or twice a week -- sufficient time for a chukar to sample all cover types given the spatial heterogeneity between cover types and the mobility of the animal. In addition, in order to prevent bias of habitat use towards time of day, I attempted to systematically obtain observations in equal amounts in 3 diurnal time periods: 1) sunrise to 1000 hours, 2) 1001 to 1400 hours, and 3) 1401 hours to sunset (Alldredge and Ratti 1992).
Even with the measures taken to prevent dependence between locations, many would argue that the use of locations instead of animals as sampling units constitutes pseudoreplication (Hurlbert 1984, White and Garrott 1990, Ratti and Garton 1994) and, consequently, lack of complete independence. Nonetheless, given the characteristics of my data set, analysis of data per individual bird was not feasible. Because of rough terrain, far-distanced locations, and bird losses due to a variety of reasons (radio transmitter failures, slipped radios, depredation), low numbers of locations per individual resulted. White and Garrott (1990) suggested that in some cases when few observations are taken on many animals, pooling of data may be justified.
The number of radio locations per animal determines the accuracy with which its habitat use is estimated (Aebischer et al. 1993). Given that the area contains 4 main cover types, I believed that 8 locations per individual (allowing 2 locations per cover type if all were used equally) would be a minimum representation of habitat use per bird. After birds with less than 8 locations were eliminated, the remaining sample size was 20 animals. Aebischer et al. (1993) suggested that there be at least 10 individuals per category for comparisons between categories. Twenty animals was an insufficient sample size to test for differences between years, seasons, and sexes. Consequentially, I chose to use the number of radio locations as sample size. This prevented the
Chukar Comp97 17 loss of data from radioed birds with less than 8 locations (31 birds) and allowed for the use of unmarked data -- and hence a larger sample size and more powerful test.
The consequences of conducting an experiment with pseudoreplication results in limiting the inference population to individuals radio-collared (Ratti and Garton 1994). Because the radio- tagged birds were random samples from the study area, however, I maintain that results from these individuals can be used to make inferences concerning the population of chukars in this section of the Lower Salmon River canyon.
RESULTS
Data Collection
During February-May 1995, 23 chukars were trapped. One bird was depredated in the trap and the remaining 22 were radio-marked. In 1996, 33 chukars were trapped between February and May, 29 of which were radio-marked. Two trapped birds died because of handling stress, 1 bird was depredated in the trap, and 1 bird died from intraspecific aggression in the trap (pecked to death).
Habitat use data from 1995-96 was obtained from 51 radio-tagged birds (30 males, 21 females) and numerous incidental unmarked birds. Locations which resulted from mortalities or non- independence were censored and not used. In 1995, 79 locations from radio-tagged birds and 45 from unmarked birds were used in habitat analyses. In 1996, 205 locations from radioed birds and 17 locations from unmarked birds were used; resulting in 346 locations in 1995-96.
Habitat use data were separated into 2 seasons -- spring and summer. Spring habitat use locations were gathered between February 16 and June 10 and were reflective of pair formation, breeding, and 1st nesting attempts. Summer habitat use locations were gathered between June 11 and August 13, and were reflective of renesting and brood rearing.
Slope and elevation for spring habitat use locations averaged 51% and 3158 m respectively (Table 2.3). For summer locations of habitat use, slope and elevation averaged 60% and 2887 m, respectively. Results of t-tests show that slope measurements at bird locations were significantly different between spring and summer locations, but elevations were not.
Measurements of aspect for locations in the spring and summer ranged from 18-358° and 14- 359°, respectively (Table 2.3). Four percent of spring locations were categorized as Northeast aspect, while 24%, 54%, and 18% were categorized as Southeast, Southwest, and Northwest, respectively. For summer locations, 5%, 12%, 54%, and 29% were categorized as Northeast, Southeast, Southwest, and Northwest, respectively. Chi-square homogeneity analysis on the frequency of locations in each quadrant resulted in the rejection (χ2 = 11.5, df = 3, P = 0.009) of the null hypothesis that aspect measurements from spring locations were homogeneous to measurements from summer locations. Binomial proportions Z-tests detected significant
Chukar Comp97 18 differences in the use of Southeast (Z = 2.9, P = 0.002) and Northwest (Z = -2.4, P = 0.008) quadrants, but no differences in Northeast (Z = 0.132, P = 0.448) and Southwest (Z = -0.108, P = 0.456) quadrants. During summer months, slopes of Southeast aspect were used less and slopes of Northwest aspect were used more.
Within the use locations categorized as rock cover type, the percent of talus, cliff, or outcrop averaged 39% (Table 2.4). Forbs were also abundant at 20%. Within the shrub cover types that were used, the percent shrubs averaged 34%. Grass and forb averaged 19% and 26%, respectively, in the grass/forb cover type (Table 2.4).
The more common shrub species found in use locations were curleaf sagewort (Artemisia ludoviciana), Ribes spp., syringa (Philadelphus lewisii), common snowberry, and Rosa spp. The common grass species were brome spp., bluebunch wheatgrass, Poa spp., medusahead (Taeniatherum asperum), and fescue spp., and the common forb species were yellow starthistle, arrowleaf balsamroot (Balsamorhiza sagittata), Lomatium spp., western yarrow (Achillea lanulosa), and tonella (Tonella floribunda).
Statistical Analyses
Because minimum expected frequencies of agriculture cover type did not meet the requirements by Dixon and Massey (1969), and because frequencies were too low to be analyzed in log-linear models, tests of homogeneity were performed without the agriculture cover type. Hence, the use of agriculture between years, seasons, sexes, or marked and unmarked bird locations could not be tested. For purposes of selection analysis, however, agriculture cover types were tested in goodness of fit statistics.
The null hypothesis that samples taken from marked and unmarked data during the springs of 1995 and 1996 were homogeneous was not rejected (χ2 = 8.648, df = 6, P = 0.1944). Therefore, data from marked and unmarked chukars for both springs could be pooled. Also, the null hypothesis that samples taken from marked and unmarked data during the summers of 1995 and 1996 were homogeneous was not rejected (χ2 = 8.226, df = 6, P = 0.222). Thus, data from marked and unmarked birds could be pooled for both summers. Using the pooled data (marked and unmarked for 1995-96), the null hypothesis that samples of habitat use taken during the spring and summer were homogeneous was rejected (χ2 = 18.619, df = 2, P = 0.00009).
Locations from unmarked birds were removed from the data set to construct log-linear models (CATMOD; SAS Institute, Inc. 1990) that examined interactions between cover type use, year, season, and sex. The ideal model is one that best describes the data relationships and is simple. However, there is no all-purpose, best method of model selection (Fienberg, 1977). The saturated model was chosen first and all higher order interactions which were nonsignificant were eliminated until the significance of all interactions of interest were realized. No attempt was made to detect the best fit, simplest model, because interest in the model was only to assess interactions between dependent variables. All 3- and 4-way interactions were nonsignificant, resulting in a final model of main effects and 2-way interactions (Table 2.5). The interaction
Chukar Comp97 19 between season and cover was significant (P = 0.0056), agreeing with the results of the previous chi-square test. There were no significant interactions between the dependent variables year and cover (P = 0.1757), or gender and cover (P = 0.097). The likelihood ratio statistic compares this model with the saturated model and is an appropriate goodness of fit test for the model; the nonsignificant p-value of 0.4806 attests the model fits.
Because homogeneity tests and log-linear models suggested data may be pooled between marked and unmarked birds, years, and sexes but not between seasons, habitat use, and selection were analyzed per season using all the data pooled. Binomial proportions Z-tests revealed seasonal differences in use of rock (P = 0.0087) and shrub (P = < 0.001) cover types, while seasonal use of grass/forb and agriculture were not different (Table 2.6). Birds increased use of shrubs in summer, while rock cover types were used less.
The degree of habitat utilization was also examined by ranking proportional use of each cover type. During spring, grass/forb was the most used cover type, followed by rock, shrub, and agriculture (Table 2.6). During summer, shrub was the most used cover type, followed by grass/forb, rock, and agriculture.
Habitat selection was analyzed by first testing for differences between use and availability. The null hypothesis that the observed frequency of cover type use in spring was proportional to the expected frequency was rejected (χ2 = 172.23, df = 3, P<0.001).
During spring, chukars used rock and shrub more than expected, grass/forb less than expected, and agriculture equal to expected (P < 0.05 for all pairwise comparisons; Table 2.7). The null hypothesis that the observed frequency of cover type use in the summer was proportional to the expected frequency was rejected (χ2 = 329.33, df = 3, P < 0.001). During summer, chukars used rock and shrub more than expected, grass/forb less than expected, and agriculture less than expected (P < 0.05 for all pairwise comparisons; Table 2.8).
Analysis of yellow starthistle data was completed using several tests. Log-linear models (CATMOD; SAS Institute, Inc. 1990) revealed that the 3-way interaction between use of yellow starthistle densities, years, and sexes was not significant (P = 0.1202). Next, all nonsignificant 2-way interactions were removed from the 2-way model to construct the final model. Because the maximum likelihood analysis-of-variance table (Table 2.9) revealed a significant interaction between year and density (P = 0.0024), yellow starthistle use data could not be pooled between years. Data from 1996 was used because of larger sample size (n = 125 for 1996, n = 34 for 1995). Two-way interactions between starthistle density and sex were not significant; thus, male and female data could be pooled. The homogeneity analysis null hypothesis that samples of yellow starthistle use taken for males, females, and unmarked birds were homogeneous was not rejected (χ2 = 1.025, df = 2, P = 0.599). Hence, marked and unmarked data could be pooled.
The null hypothesis that the observed frequency of chukar use of yellow starthistle densities in the summer was proportional to the expected frequency was rejected (χ2 = 57.3, df = 1, P < 0.001). Bonferroni confidence intervals revealed that both categories of densities were used
Chukar Comp97 20 differently than expected (P < 0.05 for both pairwise comparisons). Areas of less dense yellow starthistle (0 - 5%) were used more than expected (selected; 74% observed vs. 41% expected use), while areas of more dense starthistle ( >5%) were used less than expected (avoided; 25% observed vs. 59% expected use) by chukar partridge.
DISCUSSION
Chukar partridge exhibited selection for certain cover types. This selection varied slightly between spring and summer; however, differences in habitat use between seasons were more obvious. Differential diet selection among seasons may be depicted by differential habitat use. Alkon et al. (1985) suggested that the chukar’s food was largely a function of forage availability, which is closely linked to meteorological and phenological events. Other studies have also demonstrated seasonal variation in chukar food habits (Moreland 1950, Christensen 1952, Galbreath and Moreland 1953, Sandfort 1954). Galbreath and Moreland (1953) recorded a dominance of cheatgrass seeds and increased percentages of serviceberry, hawthorn, chokecherry (Prunus sp.) fruits and animal matter in summer/fall diets. In Nevada, increased consumption of seeds, fruits, and insects was also noted (Christensen 1952, Weaver and Haskell 1967). In this study, use of shrub habitats by chukar partridge increased from spring to summer (20% increase); shrubs were ranked 3rd in use during spring and 1st in use during summer. The previous reported increases in fruit consumption during summer months may be a contributing factor to the observed increased proportional use of shrubs (most of which provide fruits throughout the summer) by chukar partridge in the Lower Salmon River canyon.
There are other possible explanations to why chukars exhibit differential seasonal use of habitats. Galbreath and Moreland (1953) noted that chukars can tolerate temperatures as high as 118° F (48 C) without great distress. Even though summer temperatures rarely exceed 118° F in this region, the cooler temperatures found in shrub habitats may be desirable. Not only did proportional use of shrubs increase significantly in summer, proportional use of rock habitats decreased significantly. It is likely that talus slopes and rock outcrops were hotter during summer months, while shrub habitats were cooler. In addition, the use of southern exposed slopes was less in summer than in spring, further suggesting that chukars may be exhibiting behavior to avoid higher temperatures.
Chukar Comp97 21 The availability of standing water is also an important habitat factor. Although Greenhalgh (1955) noted that 8 adult chukar survived an experiment of being penned without water for 81 days, it is unlikely that chukars in the wild exhibit the desire to repress drinking for this long. Water is believed to limit the distribution of chukar partridge during summer months (Galbreath and Moreland 1953, Harper et al. 1958, Christensen 1970). As water sources and vegetation begin to dry up, chukar partridge concentrate at the few sources left during early morning and evening hours (Alcorn and Richardson 1951, Galbreath and Moreland 1953, Bohl 1957, Harper et al. 1958, Christensen 1970). Degen et al. (1984) suggested that chukars do not require drinking water from early winter to late spring, when succulent green forage is available, but need free water during summer and autumn. Galbreath and Moreland (1953), Harper et al. (1958), and Christensen (1970) suggested that habitat use during the summer is greatly influenced by the availability of water. Perhaps in west-central Idaho, the increase in proportional use of shrubs from spring to summer, many of which were riparian habitats, could be partly a result of the availability of water present in these areas.
The use and impact of agricultural crops by chukar partridge during certain times of the year has been previously questioned. Wheat, rye, and barley were reported eaten by chukars in Washington and potato crop damage was witnessed during dry summers when weather forced chukars out of their natural range into cultivated lands for water (Galbreath and Moreland 1953). Damage to corn and wheat occurred in Nevada during winter months (Alcorn and Richardson 1951). Tomlinson (1960) reported that alfalfa, wheat, and barley are eaten by chukar. However, Christensen (1970) believed that these incidents have been unusual, and so infrequent, that for all practical purposes there is no conflict between chukars and agriculture. In addition, Harper et al. (1958) reported that crop depredation by chukar was light in all cases and of little economic importance. The results of this study are in agreement with Christensen (1970) and Harper et al. (1958). Selection of agricultural fields during spring or summer did not occur. Spring proportional use (0.03) of agriculture was not significantly different than proportional availability, and summer proportional use (0.01) was significantly less than availability.
Chukar partridge use of areas heavily infested by yellow starthistle was less than expected. Chukars selected areas with low densities of starthistle, suggesting the potential negative impacts of this noxious weed on chukar populations. This comes as no surprise, given the long sharp spikes surrounding each flower head that often produces a nearly impassable field of thorns. Populations in the study area are presently at all-time lows since introductions occurred, and have not shown much upward change in the last 5 years (Hemker 1997). Coincidentally, yellow starthistle has drastically increased in the last decade (Craig Johnson, BLM Area Biologist, pers. comm.).
MANAGEMENT IMPLICATIONS
Habitat analyses are an important component of wildlife management because habitat provides food, cover, and other factors essential for the population to survive (Samuel and Fuller 1994). Use/availability-related analyses are helpful in identifying patterns of habitat selection, but interpretation is not simple (Litvaitis et al. 1994). White and Garrott (1990) suggested that
Chukar Comp97 22 selection for habitat type does not provide evidence that the habitat type is necessary to the animal’s survival and reproduction.
Given these restrictions, results from this study could nevertheless provide guidelines specific to the management of chukar partridge. Selection by chukar partridge of shrub cover types in both spring and summer suggests the importance of this habitat for populations in similar canyon grassland habitats. It is probable that some variable associated with shrub cover types is related to fitness or survival. Some of these shrub habitats are found on the outer edges of large talus slopes, but many make up the narrow riparian communities. Managers should encourage or practice management activities which allow for the maintenance of these shrub communities. It is hypothesized that management activities that negatively impact the persistence of these shrub communities would be detrimental to chukar populations.
Land management practices such as grazing should be closely monitored to evaluate the effects on shrubs. Roath and Krueger (1982b), Kauffman et al. (1983a), and Sedgwick and Knopt (1991) reported increases in riparian shrub utilization during late season grazing. Because many of these shrub communities used by chukars are riparian habitats, ecologically-sensitive riparian grazing systems that consider the shrub component should be practiced. Early spring grazing is a preferred method in many areas because livestock are better distributed due to upland vegetation being succulent at this time and because livestock may avoid the wetter riparian soils (Clary and Webster 1989, 1990; Platts 1990; Clary and Booth 1993). Land managers should also monitor utilization rates and practice management techniques such as development of alternative water sources (Floyd et al. 1988, Smith et al. 1993), herding (Kauffman and Krueger 1984, Davis 1986, Chaney et al. 1990), and culling (Skovlin 1957, Roath and Krueger 1982a, Davis 1986, Ohmart and Anderson 1986, Thomas 1991) to maintain the well-being of these riparian communities.
Land management practices on upland communities must also be monitored. Although grass/forb cover types were labeled avoided, this was not reflective of low use. Because 78% of the study area was covered in grass/forbs, it is unlikely any amount of moderate use of this cover type when compared with availability would be labeled as not avoided. However, use of this cover type ranked first and second in summer and spring, respectively. It is probable this habitat component contains important requirements, such as food and cover, for chukar partridge.
Eradication of yellow starthistle may be necessary to increase chukar populations to preinfestation levels. The use of herbicides does not appear feasible, given the enormity of chukar habitat and associated costs of chemicals and application. Biological control agents may be the only method possible of reducing starthistle densities. Efforts and experiments to approve the import of Mediterranean insects capable of controlling yellow starthistle should be continued. In addition, halting the spread of starthistle into other chukar ranges currently uninvaded should be addressed seriously.
Chukar Comp97 23 LITERATURE CITED
Alcorn, J. R., and F. Richardson. 1951. The chukar partridge in Nevada. J. Wildl. Manage. 15:265-275.
Aebischer, N. J., P. A. Robertson, and R. E. Kenward. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74:1313-1325.
Alkon, P. U., A. A. Degen, B. Pinshow, and P. J. Shaw. 1985. Phenology, diet, and water turnover rates of Negev Desert chukars. J. Arid Environ. 9:51-61.
Alldredge, J. R., and J. T. Ratti. 1986. Comparison of some statistical techniques for analysis of resource selection. J. Wildl. Manage. 50:157-165.
______, and ______. 1992. Further comparison of some statistical techniques for analysis of resource selection. J. Wildl. Manage. 56:1-9.
American Ornithologists' Union (A.O.U.). 1988. Report of Committee of wild birds in research. Auk 105 (1, Suppl.): 1A-41A.
Amstrup, S. C. 1980. A radio-collar for game birds. J. Wildl. Manage. 44:214-217.
Bizeau, E. 1963. Chukar partridge in Idaho. Idaho Wildl. Review 15:3-4.
Bohl, W. H. 1957. Chukars in New Mexico. 1931-1957. Bull. No. 6. New Mexico Dep. Game and Fish, Santa Fe. 69 pp.
Boag, D. A. 1972. Effect of radio packages on behavior of captive red grouse. J. Wildl. Manage. 36:511-518.
Burger, L. W., Jr., T. V. Dailey, E. W. Kurzejeski, and M. R. Ryan. 1995. Survival and cause- specific mortality of northern bobwhite in northern Missouri. J. Wildl. Manage. 59:401-410.
Byers, C. R., R. K. Steinhorst, and P. R. Krausman. 1984. Clarification of a technique for analysis of utilization-availability data. J. Wildl. Manage. 48:1050-1053.
Caccamise, D. F., and R. S. Hedin. 1985. An aerodynamic basis for selecting transmitter loads in birds. Wilson Bull. 97:306-318.
Callihan, R. H., F. E. Northam, J. B. Johnson, E. L. Michalson, and T. S. Prather. 1989. Yellow starthistle control. Univ. of Idaho, College of Agr., CIS 634.
Chukar Comp97 24 Canfield, R. H. 1941. Application of the line intercept method in sampling range vegetation. J. For. 39:388-394.
Carmi-Winkler, N., A. A. Degen, and B. Pinshow. 1987. Seasonal time-energy budgets of free- living chukars in the Negev Desert. Condor 89:594-601.
Chaney, E., W. Elmore, and W. S. Platts. 1990. Livestock grazing on western riparian areas. Environmental Protection Agency, Washington, D. C. 45 pp.
Cherry, S. 1996. A comparison of confidence interval methods for habitat use-availability studies. J. Wildl. Manage. 60:653-658.
Christensen, G. C. 1952. Overwintering of the chukar partridge (Alectoris graeca) in Nevada, U. S. A. J. Bombay Nat. Hist. Soc. 51:277-279.
______. 1954. The chukar partridge in Nevada. Nevada Fish and Game Commission. Bio. Bull. No. 1, 77 pp.
______. 1970. The Chukar Partridge. Its introduction, life history, and management. Nevada Department of Fish and Game. Biological Bulletin No.4, 82 pp.
Clary, W. P., and G. D. Booth. 1993. Early season utilization of mountain meadow riparian pastures. J. Range Manage. 46:493-497.
______, and B. F. Webster. 1989. Managing grazing of riparian areas in the Intermountain Region. USDA Forest Serv., Gen. Tech. Rep. INT-263.
______, and ______. 1990. Riparian grazing guidelines for the intermountain region. Rangelands 12:209-212.
Cochran, W. W. 1980. Wildlife Telemetry. Pages 507-520 in S. D. Schemnitz, ed. Wildlife management techniques manual. Fourth ed. The Wildlife Society, Bethesda, MD.
Conover, W. J., and R. L. Iman. 1981. Rank transformations as a bridge between parametric and nonparametric statistics. Amer. Stat. 35:124-129.
Cottam, C., A. L. Nelson and L. W. Saylor. 1940. The chukar and Hungarian partridges in America. U. S. D. I. Biol. Surv. Wildl. Leaflet 159. 6 pp.
Cunningham, E. B. 1959. Influence of variable diets and time of hatch on development and age determination of the chukar partridge. Game bird survey W-50-R-8. Wyoming Game and Fish Commission. 98 pp.
Chukar Comp97 25 Davis, J. W. 1986. Options for managing livestock in riparian habitats. N. A. Wildl. and Nat. Res. Conf. 51:290-297.
Degen, A. A., B. Pinshow, and P. J. Shaw. 1984. Must desert chukars (Alectoris chukar sinaica) drink water? Water influx and body mass changes in response to dietary water content. Auk 101:47-52.
Demaso, S. J., and A. L. Peoples. 1993. A restraining device for handling Northern Bobwhites. Wildl. Soc. Bull. 21:45-46.
Dixon, W. J., and F. J. Massey, Jr. 1969. Introduction to statistical analysis. 3rd ed. McGraw- Hill, New York. 638 pp.
Erikstad, K. E. 1979. Effects of radio packages on reproductive success of willow grouse. J. Wildl. Manage. 43:170-175.
Fienberg, S. E. 1977. The analysis of cross-classified categorical data. MIT Press, Cambridge, MA. 151 pp.
Fischer, R. A., W. L. Wakkinen, K. P. Reese, and J. W. Connelly. 1997. Effects of prescribed fire on movements of female sage grouse from breeding to summer ranges. Wilson Bull. 109:82-91.
Floyd, D., P. Ogden, B. Roundy, G. Ruyle, and D. Stewart. 1988. Improving riparian habitats. Rangelands 10:132-134.
Galbreath, D. S., and R. Moreland. 1953. The chukar partridge in Washington. Bio. Bull. No. 11. Washington State Game Department. 54 pp.
Greenhalgh, C. M. 1955. Utah chukar partridge report. Western states chukar committee quarterly report 2, June-August.
Greenwood, R. J., and A. B. Sargeant. 1973. Influence of radio packs on captive mallards and blue-winged teal. J. Wildl. Manage. 37:3-9.
Harper, H. T., B. H. Harry, and W. D. Bailey. 1958. The chukar partridge in California. California Fish and Game 44:5-50.
Hemker, T. 1997. Upland Game. Job Prog. Rep., Proj. W-170-R-20, Idaho Dept. Fish and Game, Boise.
Hessler, E. J. R., J. R. Tester, D. B. Siniff, and M. M. Nelson. 1970. A biotelemetry study of survival of pen-reared pheasants released in selected habitats. J. Wildl. Manage. 34:267- 274.
Chukar Comp97 26
Horton, L. E. 1972. Ecological Analysis: a preliminary investigation of vegetation structure and ecosystem function of the lower Salmon River. USDA For. Serv., Intermountain Reg. Rep. 85 pp.
Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54:187-211.
Hupp, J. W., and J. T. Ratti. 1983. A test of radio telemetry triangulation accuracy in heterogeneous environments. Proc. Int. Wildl. Biotelemetry Conf. 4:31-46.
Idaho Fish and Game News. Fall 1993. Idaho Department of Fish and Game. Boise, ID.
Idaho Upland Game Plan. 1986-1990. Idaho Department of Fish and Game Species Management Plan. Boise, ID.
Johnsgard, P. A. 1973. Grouse and Quails of North America. Univ. Nebraska Press. Lincoln, NE. 553 pp.
Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65-71.
______. 1994. Population analysis. Pages 419-444 in T. A. Brookout, ed. Research and management techniques for wildlife and habitats, 5th ed. The Wildlife Society, Bethesda, MD.
Johnson, R. N., and A. H. Berner. 1980. Effects of radio transmitters on released cock pheasants. J. Wildl. Manage. 44:686-689.
Kachigan, S. K. 1982. Multivariate statistical analysis: a conceptual introduction. Radius Press, New York.
Kauffman, J. B., and W. C. Krueger. 1984. Livestock impacts on riparian ecosystems and streamside management implications...a review. J. Range Manage. 37:430-438.
______, ______, and M Vavra. 1983a. Effects of late season cattle grazing on riparian plant communities. J. Range Manage. 36:685-691.
Kuz'mina, M. A. 1992. Tetraonidae and Phasianidae of the USSR. Oxonian Press, New Delhi. 302 pp.
Lance, A. N., and A. Watson. 1977. Further tests of radio-marking on red grouse. J. Wildl. Manage. 41:579-582.
Chukar Comp97 27 Litvaitis, J. A., K. Titus, and E. M. Anderson. 1994. Measuring vertebrate use of terrestrial habitats and foods. Pages 254-274 in T. A. Brookout, ed. Research and management techniques for wildlife and habitats, 5th ed. The Wildlife Society, Bethesda, MD.
Marcstrom, V., R. E. Kenward, and M. Karlbom. 1989. Survival of ring-necked pheasants with backpacks, necklaces, and leg bands. J. Wildl. Manage. 53:808-810.
Molini, W. A. 1976. Chukar partridge species management plan. Nevada Department of Fish and Game, 53 pp.
Moreland, R. 1950. Success of chukar partridge in the state of Washington. N. Am. Wildl. Conf. 15:399-409.
Neu, C. W., C. R. Byers, and J. M. Peek. 1974. A technique for analysis of utilization - availability data. J. Wildl. Manage. 38:541-545.
Ohmart, R. D., and B. W. Anderson. 1986. Riparian habitat. pages 169-199 in Cooperriden et al. Inventory and Monitoring of wildlife habitat.
Peek, J. M. 1986. A review of wildlife management. Prentice-Hall, Englewood Cliffs, N. J.
Perry, M. C. 1981. Abnormal behavior of canvasbacks equipped with radio transmitters. J. Wildl. Manage. 45:786-789.
Platts, W. S. 1990. Managing fisheries and wildlife on rangelands grazed by livestock. White Horse Assoc., Smithfield, UT. 445 pp.
Porter, W. F., and K. E. Church. 1987. Effects of environmental pattern on habitat preference analysis. J. Wildl. Manage. 51:681-685.
Ratti, J. T., and E. O. Garton. 1994. Research and experimental design. Pages 1-23 in T. A. Brookout, ed. Research and management techniques for wildlife and habitats, 5th ed. The Wildlife Society, Bethesda, MD.
Roath, L. R., and W. C. Krueger. 1982a. Cattle grazing and behavior on a forested range. J. Range Manage. 35:332-338.
______, and ______. 1982b. Cattle grazing influence on a mountain riparian zone. J. Range Manage. 35:100-103.
Samuel, M. D., and M. R. Fuller. 1994. Wildlife radiotelemetry. Pages 370-418 in T. A. Brookout, ed. Research and management techniques for wildlife and habitats, 5th ed. The Wildlife Society, Bethesda, MD.
Chukar Comp97 28 Sanfort, W. W. 1954. Evaluation of chukar partridge range in Colorado. West. Assoc. State Game and Fish Comm. Proc. 34:244-250.
SAS Institute Inc. 1990. SAS/STAT User's guide, Version 6, 4th edition. SAS Institute, Cary, NC. 1686 pp.
Schooley, R. L. 1994. Annual variation in habitat selection: patterns concealed by pooled data. J. Wildl. Manage. 58:367-374.
Schumacker, R. W., C. J. Perkins, A. D. Sullivan, and D. E. Wesley. 1978. Heat-shrinkable tubing as a means of securing harness knots. J. Wildl. Manage. 42:685-686.
Sedgwick, J. A., and F. L. Knopf. 1991. Prescribed grazing as a secondary impact in a western riparian floodplain. J. Range Manage. 44:369-373.
Skovlin, J. M. 1957. Range-riding -- The key to range management. J. Range Manage. 10:269-271.
Slaugh, B. T., J. T. Flinders, J. A. Roberson, M. Ray Olson, and N. P. Johnston. 1989. Radio transmitter attachment for chukars. Great Basin Naturalist 49:632-636.
Slaugh, B. T., J. T. Flinders, J. A. Roberson, and N. P. Johnston. 1990. Effect of backpack radio transmitter attachment on chukar mating. Great Basin Naturalist 50:379-380.
Small, R. J., and D. H. Rusch. 1985. Backpacks vs. ponchos: survival and movements of radio- marked ruffed grouse. Wildl. Soc. Bull. 13:163-165.
Smith, M. A., J. D. Rodgers, J. L. Dodd, and Q. D. Skinner. 1993. Habitat selection by cattle along an ephemeral channel. Rangelands 15:120-122.
Springer, J. T. 1979. Some sources of bias and sampling error in radio triangulation. J. Wildl. Manage. 43:926-935.
Tisdale, E. W. 1986. Canyon grasslands and associated shrublands of west-central Idaho and adjacent areas. Forestry, Wildlife, and Range Exp. Stn. Bull. No. 40, Univ. of Idaho, Moscow.
Thomas, H. S. 1991. The importance of rancher input in solving riparian problems. Rangelands 13:83-84. Thomas, D. L., and E. J. Taylor. 1990. Study designs and tests for comparing resource use and availability. J. Wildl. Manage. 54:322-330.
Tomlinson, R. E. 1960. Is New Mexico climatically suitable for chukars? West. Assoc. of State Game and Fish Comm. Proc. 39:16.
Chukar Comp97 29
Warner, R. E., and S. L. Etter. 1983. Reproduction and survival of radio-marked hen ring- necked pheasants. J. Wildl. Manage. 47:369-375.
Weaver, H. R., and W. L. Haskell. 1967. Some fall foods of Nevada chukar partridge. J. Wildl. Manage. 31:582-584.
Westerkov, K. 1950. Methods for determining the age of game bird eggs. J. Wildl. Manage. 14:56-67.
White, G. C., and R. A. Garrott. 1990. Analysis of wildlife radio-tracking data. Academic Press, Inc., San Diego, CA. 383 pp.
Woodard, A. E., J. C. Hermes, and L. Fuqua. 1986. Shank length for determining sex in chukars. Poultry Sci. 65:627-630.
Zar, J. H. 1996. Biostatistical Analysis. Third ed. Englewood Cliffs: Prentice-Hall Inc., New Jersey. 662 pp.
Chukar Comp97 30 Table 1.1. A comparison of observations of penned captive-reared chukars with necklace- and backpack-mounted transmitters, 1995.
Observation Necklace-mounted birds Backpack-mounted birds immediate responsea 5 0 pecking at transmitter 5 0 transmitter askew 5 0 preening around attachment 1 4 walking backwards 5 0 eye irritation from antenna 1 0 slipped transmitter 1 0 a Frantic behavior observed directly after mounting of transmitter.
Table 2.1. Areas and percentages of spring and summer cover types measured for the Lower Salmon River study area, west-central Idaho, 1995 and 1996.
Cover type Area (ha) % of study area
Rock (cliff, talus, outcrop) 224 11.3
Shrub 126 6.3
Grass/forb 1537 77.5
Agricultural field 97 4.9
Total 1984 100
Chukar Comp97 31 Table 2.2. Densities of yellow starthistle for the Lower Salmon River study area, west-central Idaho, 1995 and 1996.
Density of yellow starthistle Area (ha) % of study area
0% < density < 5% 815 41.1
Density > 5% 1169 58.9
Total 1984 100
Chukar Comp97 32
Table 2.3. Physiographic characteristics of spring and summer habitat use locations for chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995-96.
Spring Summer F-test t-test
Ave. + SE range na Ave. + SE range na P-value P-value slope 51 + 21 0 - 94 169 60 +17 8 - 90 156 0.003 <0.001b aspect N/A 18 - 358 167 N/A 14 - 359 156 N/Ac N/Ac elevation 3158 + 390 1950 - 3680 171 2887 + 444 1780 - 3690 158 0.048 0.054b
a Sample size. b Two-tailed t-test assuming unequal variances. c Not appropriate tests for circular data.
Chukar Comp97 33
Table 2.4. Average, minimum, and maximum percentages of rock, shrub, grass, forb, and agriculture estimated for habitat use locations of each cover type for chukar partridge, west-central Idaho, 1995 and 1996.
COVER TYPE
ROCK SHRUB GRASS/FORB AGRICULTURE (n = 100) (n = 91) (n = 123) (n = 7)
Ave. + S.E. Min.a Max.b Ave. + S.E. Min.a Max.b Ave. + S.E. Min.a Max.b Ave. + S.E. Min.a Max.b
% rock 39 + 19 18 100 20 + 19 0 80 9 + 9 0 40 0 + 0 0 0
% shrub 4 + 5 0 15 34 + 13 15 75 3 + 5 0 16 0 + 0 0 0
% grass 13 + 9 0 50 9 + 7 0 40 19 + 14 1 65 4 + 10 0 25
% forb 20 + 14 0 65 13 + 9 0 45 26 + 14 0 60 2 + 4 0 10
% agric. 0 + 0 0 0 0 + 0 0 0 0 + 0 0 0 93 + 19 50 100
a Minimum value observed. b Maximum value observed.
Chukar Comp97 34
Table 2.5. Maximum-likelihood analysis-of-variance table for log-linear model of dependent variables gender, season, year, and cover for chukar partridge habitat use in west- central Idaho, 1995 and 1996.
Source DF Chi-squarea P-value
Gender 1 14.06 0.0002
Season 1 0.00 0.9721
Year 1 49.91 0.0000
Cover 2 8.56 0.0139
Gender x Cover 2 4.67 0.0970
Season x Cover 2 10.38 0.0056
Season x Year 1 9.87 0.0017
Year x Cover 2 3.48 0.1757
Likelihood Ratio 11 10.56 0.4806
a Chi-square test for each effect is a Wald test (SAS Inc. 1990) based on the information matrix from the likelihood.
Chukar Comp97 35
Table 2.6. Analyses of habitat use per cover type between spring and summer for chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995 and 1996.
Cover Spring Spring rank Summer proportional Summer rank test P value Resultsb proportional use of use use (n=158) of use statistica (n=179)
Rock 0.374 2 0.253 3 2.384 0.009 * used less
Shrub 0.173 3 0.386 1 4.381 < 0.001 * used more
Grass/forb 0.419 1 0.354 2 1.214 0.113 none
Agricultur 0.034 4 0.006 4 N/Ac N/Ac N/Ac e
Asterisks indicate significance (P < 0.05). a Binomial proportions Z-test. b Differences of habitat use for summer compared with spring. c The use of agriculture between seasons could not be tested because of small sample sizes.
Chukar Comp97 36
Table 2.7. Bonferroni simultaneous confidence intervals for spring habitat selection analysis of chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995 and 1996.
Cover Expected proportional Observed proportional Bonferroni 95% Confidence Results use use Interval
Rock 0.113 0.374 (0.284, 0.465)* selecta
Shrub 0.063 0.173 (0.125, 0.244)* selecta
Grass/forb 0.775 0.419 (0.327, 0.511)* avoidb
Agriculture 0.049 0.034 (0.000, 0.067) nonec
• Asterisks indicate significance (P < 0.05). a Select indicates observed proportional use significantly greater than expected proportional use. b Avoid indicates observed proportional use significantly less than expected proportional use. c None indicates observed proportional use not significantly different than expected proportional use.
Chukar Comp97 37
Table 2.8. Bonferroni simultaneous confidence intervals for summer habitat selection analysis of chukar partridge in the Lower Salmon River Canyon of west-central Idaho, 1995 and 1996.
Cover Expected proportional Observed proportional Bonferroni 95% Confidence Results use use Interval
Rock 0.113 0.253 (0.167, 0.340)* selecta
Shrub 0.063 0.386 (0.289, 0.483)* selecta
Grass/forb 0.775 0.354 (0.259, 0.450)* avoidb
Agriculture 0.049 0.006 (0.000, 0.021)* avoidb
• Asterisks indicate significance (P < 0.05). a Select indicates observed proportional use significantly greater than expected proportional use. b Avoid indicates observed proportional use significantly less than expected proportional use.
Chukar Comp97 38
Table 2.9. Maximum-likelihood analysis-of-variance table for log-linear model of dependent variables sex, year, and density of yellow starthistle used by chukar partridge in west-central Idaho, 1995 and 1996.
Source DF Chi-squarea P-value
Gender 1 15.69 0.0001 Year 1 34.52 0.0000 Density 1 3.15 0.0760 Year x Density 1 9.24 0.0024
Likelihood Ratio 3 5.73 0.1254 a Chi-square test for each effect is a Wald test (SAS Inc., 1990) based on the information matrix from the likelihood.
Chukar Comp97 39
Figure 0.1. Approximate distribution and relative abundance of chukar partridge in Idaho (Idaho Upland Game Management Plan 1986-90).
Chukar Comp97 40
Figure 0.2. Study area for chukar partridge research located in the lower Salmon River canyon, west-central Idaho, 1995-96.
Chukar Comp97 41
Figure 1.1. Dimensions of restraining device for handling and radio-tagging chukar partridge.
Chukar Comp97 42
Figure 1.2. Chukar partridge in restraining device.
Chukar Comp97 43
Figure 1.3. Percentages of radio-tagged birds lost due to depredation, slipped collars, radio- caused mortalities, and other miscellaneous causes compared between chukars with necklace- and backpack-mounted transmitters, west-central Idaho, 1995 and 1996.
Chukar Comp97 44
Figure 1.4. Weight change from first capture of chukars collared with necklace-mounted radio transmitters in 1995, west-central Idaho.
Chukar Comp97 45
Figure 1.5. Weight change from first capture of chukars collared with backpack-mounted radio transmitters in 1996, west-central Idaho.
Chukar Comp97 46
Submitted by:
Peter Zager Principal Wildlife Research Biologist
Approved by:
IDAHO DEPARTMENT OF FISH AND GAME
John Beecham Wildlife Game and Research Manager Federal Aid Coordinator
Steven M. Huffaker, Chief Bureau of Wildlife
Chukar Comp97 47