THE SOUTHWESTERN 46(3) :354-357 NATURALIST 46(3):354-357
SEPTEMBER 2001
BED-SITE SELECTION BY DESERT MULE DEER IN SOUTHERN ARIZONA JOHN C. TULL, PAUL R. KRAUSMAN,*AND ROBERTJ. STEIDL School of RenewableNatural Resources,325 Biological SciencesEast Building, The Universityof Arizona, Tucson, AZ 85721 Present address ofJCT: 21495 DortortDrive, Reno, NV 89511 * Correspondent:
[email protected] ABSTRACT-Wecompared bed sites selected by desert mule deer (Odocoileus hemionus eremicus) to nearby random sites to assess bed site features. Thermal cover of bed sites (i.e., vegetation -75 cm high that provided shade for a deer) was highest in summer (X = 43.7% + 4.1 SE) compared to spring (29.7% + 4.4), winter (33.5% ? 3.1), and autumn (39.8% + 4.0); however, selection for bed sites with thermal cover that differed most from that available randomly was highest in spring (X difference between bed and random sites = 14.8% + 5.2) compared to summer (8.3% ? 4.1), winter (6.6% + 3.5), and autumn (8.6% ? 3.5). Thermal cover is likely important as a contribution to thermoregulation, escape cover, protection of fawns, and fawn survival. RESUMEN-Comparamos los echaderos seleccionados por el venado bura (Odocoileus hemionus eremicus) a sitios cercanos al azar para determinar las caracteristicas de los echaderos. La cubierta termal de los echaderos (o sea, vegetaci6n >75 cm de altura que provey6 sombra para el venado) fue mas alta en el verano (X = 43.7% + 4.1 SE), comparado con la primavera (27.9% ? 4.4), invierno (33.5% + 3.1), y otono (39.8% + 4.0). Sin embargo, la selecci6n de echaderos con cubierta termal que difiri6 mas que lo disponible al azar fue mas alta en la primavera (Xdiferencia entre echaderos y sitios al azar = 14.8% + 5.2) comparada con el verano (8.3% ? 4.1), invierno (6.6% ? 3.5), y otono (8.6% + 3.5). La cubierta termal es quizas importante como una contribuci6n a la termorregulaci6n, cubierta de escape, protecci6n de ciervos, y supervivencia de ciervos.
The scale at which biologists measure and assess the ways in which wildlife use landscapes influences the patterns of use observed (Kotliar and Wiens, 1990). Habitat use by ungulates, for example, is studied most often at the home range (macrohabitat) scale with radiocollared animals. Movements, forage, and cover selection by desert mule deer (Odocoileus hemionus eremicus) all have been studied
at the
home-range scale in the southwestern United States (Dickinson and Garner, 1979; Koerth et al., 1985; Ordway and Krausman, 1986; Bellantoni and Krausman, 1990; Albert and Krausman, 1993), but not at the smaller, microhabitat scale. Studies on use of thermal cover by deer in northern climates suggest that dense cover of conifer forest canopies benefits deer most during winter (Cox, 1938; Verme and Ozoga, 1971; Ozoga and Gysel, 1972; Moen, 1976; DelGiudice and Riggs, 1996). Conversely, tem-
perature extremes in southwest deserts occur during summer; therefore, thermal cover (i.e., vegetation -75 cm high that provided shade for a deer) can aid in thermoregulation (Anthony, 1972; Leopold and Krausman, 1987; Hayes and Krausman, 1993). Our objectives were to determine if deer selected (Hall et al., 1997) bed sites with different amounts of thermal cover compared to paired random sites (50-100 m from bed sites) and to understand vegetation characteristics of bed sites. AND METHODS-We studied deer in MATERIALS
AvraValley,Pima Co., Arizona, in an area bounded by the Tucson Mountains to the east and the Roskruge Mountainsto the west, 20 km west of Tucson. Elevations ranged between 610 and 1,429 m. Land ownershipwas patchy with municipal, state, federal, Tohono O'odham Nation, and private landholders (e.g., housing developments and ranches) represented.
Tull et al.-Bed-site
September 2001
355
selection by mule deer
TABLE 1-Thermal cover (%) selection for bed versus nearly random sites for desert mule deer in Avra Valley, Arizona, 1996-1997.
X
SE
X
SE
X difference
29.7 43.7 39.8 33.5
4.37 4.11 4.00 3.06
14.9 35.3 31.2 26.9
3.99 4.35 4.20 3.06
14.75 8.32 8.58 6.64
Bed Season Spring Summer Autumn Winter
n 26 49 38 45
Bed
Random
F (df= 1, n)
P
8.12 4.16 5.96 3.65
0.0086 0.0470 0.0195 0.0625
rection of which we determined randomly. We then Vegetation associations in the area were common to the Sonoran Desert and included: creosote (Lar- selected randomly a paired location 50 to 100 m rea tridentata)-bursage (Ambrosia) in undisturbed from the bed-site and repeated vegetation measureflats; desert grassland along portions of major wash- ments. es and associated flood plains; ironwood (Olneya teWe used logistic regression for matched case-consota)-canyon ragweed (Ambrosiaambrosioides)in wash- trol studies to compare vegetation characteristics of es; mesquite (Prosopis velutina)-burroweed (Isocoma beds to random sites for males and females comtenuisecta) in disturbed flats (i.e., abandoned agri- bined (Hosmer and Lemeshow, 1989). We used stepcultural lands); and palo verde (Cercidium)-mixed wise selection (P = 0.25 to enter; P = 0.15 to recacti on bajadas. move) to determine those characteristics that best differentiated bed from paired-random sites. Precipitation is bimodal, occurring during the July to September monsoon and December to March winter showers (Reitan and Green, 1968). Mean preRESULTS-We captured 17 desert mule deer, cipitation was 28.3 cm for Tucson, Arizona, 1996 3 males and 14 females, a ratio similar to all (National Climatic Data Center, 1996). Average nordeer observed in the study area between July mal seasonal temperatures for 1993 were 24.3, 30.2, 1996 and March 1997 (39 males and 154 fe16.6, and 13.6? C for spring (April toJune), summer males). Thermal cover of bed sites used by (July to September), autumn (October to Decemdeer was highest in summer compared to othber), and winter (January to March), respectively er seasons (Table 1). Deer showed evidence of (National Climatic Data Center, 1993). These seaselection for bed sites in areas of relatively high sons were defined based on weather patterns and thermal cover in all seasons (X2 - 3.60, P < biology of desert mule deer (Krausman and Etch0.06). However, deer selected bed sites with berger, 1995). We used net-guns to capture and collar desert thermal cover that differed most from that mule deer (Krausman et al., 1985) in November available in spring when available thermal cov1995 and February 1996 to obtain a representative er was lowest (14.9%) compared to other seasample of the mule deer in Avra Valley. Each collar sons when available thermal cover was much was color-coded to assist in visual identification of Table 1). Characteristics (26.9-35.3%; higher individuals. that distinguished bed sites from random sites Locations of individual deer were obtained >24 h also varied seasonally (Table 2). apart during daylight hours. We hiked to collared deer and attempted to obtain a visual location to the high temperatures in DIscussION-Given determine if the deer was bedded or not. If the deer the desert Southwest, we expected amount of was bedded, we marked the location for vegetation measurements. thermal cover to be an important determinant When we saw bedded deer, we recorded their loof bed-site selection by desert mule deer durcation. We used the line-intercept method (Canfield, selecing spring and summer. Consequently, 1941) to measure percent dominant vegetation of tion for thermal cover was strongest in spring, each species and percent thermal cover in and 20 when available thermal cover was lowest (Table m around the bed site. We measured percent cover 1). In summer, bed sites used had the highest of all plant species; grouped percent-cover values for cover of all seasons but because available thertrees, shrubs, forbs (i.e., annual herbaceous plants), mal cover was higher, selection was less strong succulents (i.e., fleshy plants with water-storing leaves or stems), vines, and grasses; total percent cov- than for spring (Table 1). Amount of available thermal cover was highest in summer because er for all plant species; and percent thermal cover. We measured vegetation along a 20-m line, the di- foliage growth is increased because of mon-
356
The SouthwesternNaturalist
vol. 46, no. 3
TABLE 2-Vegetation parameters selected by stepwise procedures, their standard errors, Wald Chi-square, P-value, and odds ratios for the probability of bed versus random sites for desert mule deer in Avra Valley, Arizona, 1996-1997. Estimates represent the difference in % cover of parameters between bed and random sites as measured along a 20-m line transect.
Season
Parameter
Estimate
Spring Summer
Autumn
Tree Total Grass Vine Total
Winter
Mesquite Thermal
0.07 0.04 0.08 -0.13 0.08 0.08 0.02
SE
Wald x2
P
Odds ratio
0.027 0.016 0.054 0.071 0.050 0.035 0.014
6.72 7.38 2.31 3.59 2.48 5.82 3.18
0.0096 0.0066 0.1282 0.0582 0.1152 0.0158 0.0746
1.073 1.046 1.085 0.874 1.082 1.088 1.026
soon rains. Selection also occurred during winter and autumn, but to a lesser degree, which suggests thermoregulation is less of a factor during these mild seasons. Further, selection of thermal cover in winter may be confounded by deer seeking escape cover during winter and autumn. We observed female mule deer fawning as early as 8 August 1996 (Fox and Krausman, 1994). Female deer may have selected vegetation with high cover to protect fawns and to gain access to forage with greater nutritional benefits. Because the first 45 days of a fawn's life are most crucial to survival (Trainer, 1975), female nutritional benefits from rains may improve lactation quality or quantity and improve fawn survival (Smith and LeCount, 1979; Urness, 1981). Additionally, increased thermal cover, which implies increased escape cover, may improve fawn survival during summer. Dense stands of mesquite made it difficult to approach deer while remaining undetected, suggesting that success for non-ambush predators may be reduced in areas of dense cover. Selection for bed sites was stronger in spring, summer, and autumn compared to winter (Table 1). We expected bed site selection in winter to be weaker than other seasons because deer are not thermally stressed. In winter, however, deer selected bed sites with characteristics that indicated high thermal cover (Table 2). Escape and thermal cover were indistinguishable by our measures, therefore selection for thermal cover may, in part, reflect a selection for protective cover. Our view of thermal cover at bed sites reflects our scale of measurement and the paired design we used. It is clear, however, that deer
are able to locate and use areas with high thermal cover despite seasonal periods of low available cover. Additional research is needed to determine the ecological significance thermal cover has on deer condition, survival, productivity, and recruitment (Parker and Gillingham, 1990; Cook et al., 1998). J. T. Avey and R. J. Popowski assisted with data collection. The United StatesBureauof Reclamation provided helicopters for locating deer.J. P. Charpentier provided computer support. L. Fox and an anonymousreferee provided constructivecomments on an earlier draft. S. Moralesprepared the Spanish resumen. Funding wasprovidedby the United States Bureau of Reclamation. LITERATURECITED 1993. Desert mule S. K., ANDP. R. KRAUSMAN. ALBERT,
deer and forage resources in southwestArizona. SouthwesternNaturalist38:198-205. ANTHONY, R. G. 1972. Ecological
relationship be-
tween mule deer and white-taileddeer in southeast Arizona. Unpublished Ph.D. dissertation, Universityof Arizona, Tucson. 1990. MoveBELLANTONI,E. S., AND P. R. KRAUSMAN.
ments of desert mule deer and collared peccaries in Saguaro National Monument. The University of Arizona Cooperative National Park Studies Unit, Technical Report 42. CANFIELD,R. H. 1941. Application of the line intercept method in sampling range vegetation.Journal of Forestry39:388-394. COOK, J. G., L. L. IRWIN, L. D. BRYANT, R. A. RIGGS, ANDJ. W. THOMAS. 1998. Relations of forest cover
and condition of elk: a test of the thermal cover hypothesisin summer and winter.WildlifeMonographs 141. Cox, W. T. 1938. Preventing deer concentrations. Journal of Wildlife Management2:1-2.
September2001
Tull et al.-Bed-site selection by mule deer
G. D., AND M. R. RIGGS.1996. LongDELGIUDICE, term research on the white-tailed deer-conifer thermal cover relationship: aligning expectations with reality. Transactions of the North American Wildlife and Natural Resources Conference 61: 416-428. T. G., AND G. W. GARNER.1979. Home DICKINSON, range use and movements of desert mule deer in southwestern Texas. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 33:267-278. 1994. Fawning habFox, K. B., ANDP. R. KRAUSMAN. itat of desert mule deer. Southwestern Naturalist 39:269-275. AND M. L. MORRISON. HALL,L. S., P. R. KRAUSMAN, 1997. The habitat concept and a plea for standard terminology. Wildlife Society Bulletin 25: 173-182. 1993. Nocturnal HAYES,C. L., ANDP. R. KRAUSMAN. activity of female desert mule deer. Journal of Wildlife Management 57:897-904. 1989. Applied HOSMER,D. W.,JR., ANDS. LEMESHOW. logistic regression. John Wiley & Sons, New York.
357
MOEN, A. N. 1976. Energy conservation by whitetailed deer in the winter. Ecology 56:192-198. NATIONAL DATACENTER.1993. ClimatologCLIMATIC ical data annual summary: Arizona. National Oceanic and Atmospheric Administration. Asheville, North Carolina. NATIONAL CLIMATIC DATACENTER.1996. Climatological data annual summary: Arizona. National Oceanic and Atmospheric Administration. Asheville, North Carolina. 1986. Habitat ORDWAY, L. L., AND P. R. KRAUSMAN. use by desert mule deer. Journal of Wildlife Management 50:677-683. OZOGA,J. J., AND L. W. GYSEL.1972. Response of white-tailed deer to winter weather. Journal of Wildlife Management 50:471-474. K. L., ANDM. P. GILLINGHAM. 1990. Estimates PARKER, of critical thermal environments for mule deer. Journal of Range Management 43:73-81. REITAN,C. R., ANDC. R. GREEN.1968. Weather and climate of desert environments. In: McGinnies, W. J., B. J. Goldman, and P. Paylore, editors. Deserts of the world. University of Arizona Press, Tucson. Pp. 21-92. KOERTH, B. H., JR., B. K. SOWELL,F. C. BRYANT,AND SMITH,R., ANDA. LECOUNT.1979. Some factors afE. P. WIGGERS. 1985. Habitat relations of mule fecting survival of desert mule deer fawns. Jourdeer in the Texas Panhandle. Southwestern Natnal of Wildlife Management 43:657-665. C. 1975. Direct causes of mortality in mule uralist 30:579-587. TRAINER, deer fawns during summer and winter periods of N. B., ANDJ.A. WIENS.1990. Multiple scales KOTLIAR, Steens Mountain, Oregon-a of patchiness and patch structure: a hierarchical progress report. framework for the study of heterogeneity. Oikos Proceedings of the Western Association of Game and Fish Commissioners 55:163-170. 59:253-260. URNESS, P.J. 1981. Desert and chaparral habitats. In: P. R., AND R. C. ETCHBERGER. 1995. ReKRAUSMAN, Wallmo, O. C., editor. Mule and black-tailed deer of desert to a water in sponse ungulates project of North America. University of Nebraska Press, Arizona. Journal of Wildlife Management 59: Lincoln. Pp. 347-385. 292-300. OZOGA.1971. Influence of winP. R., J. J. HERVERT, AND L. L. ORDWAY. VERME,L.J., ANDJ.J. KRAUSMAN, ter weather on white-tailed deer in Upper Mich1985. Capturing deer and mountain sheep with igan. In: Haugen, A. 0., editor. Snow and Ice a net-gun. Wildlife Society Bulletin 13:71-73. Symposium. Iowa State University Press, Ames. 1987. Diurnal LEOPOLD,B. D., ANDP. R. KRAUSMAN. Pp. 16-28. activity patterns of desert mule deer in relation to temperature. Texas Journal of Science 39:49Submitted20 March 2000. Accepted18 August 2000. 53. AssociateEditor was David R. Edds.
