The Condor8956-14 0 The CooperOrnithological Society 1987
HABITAT USE BY MOUNTAIN QUAIL IN NORTHERN CALIFORNIA’
LEONARD A. BRENNAN,* WILLIAM M. BLOCK,~ AND R. J. GUTIBRREZ Department of Wildlife, Humboldt State University,Arcata, CA 95521
Abstract. We studied habitat use by Mountain Quail (Oreortyx pictus) at four sites in northern California. Vegetative cover types (macrohabitats) were used in proportion to availability. Significant microhabitat variables which distinguished used from available microhabitat structureincluded proximity to water and tall, dense shrubs.Mountain Quail population densitiesranged from 9 to 30 birds per 100 ha; populationswith greaterdensities used a larger range of the available microhabitat structure. Key words: Habitat use;habitat availability; habitat selection;population density;logistic regression;northern California; Mountain Quail.
INTRODUCTION The objectives of this paper are to (1) provide The Mountain Quail (Oreortyx pictus) inhabits a quantitative analysis of the structural and flo- upland forest and woodland habitats in the west- ristic aspectsof habitat use by Mountain Quail ern United States and Baja California Norte, during spring and summer, and (2) examine the Mexico (AOU 1983). Although this quail is men- relationship between habitat use and variation tioned in at least 270 published accounts (Gu- in local population density. tiirrrez 1975) most of these reports consist of STUDY AREAS anecdotal information and only a few (e.g., Grin- nel et al. 1918, McLean 1930, Rahm 1938, Miller The four study areas(Fig. 1) were chosenbecause and Stebbins 1964) contain significant infor- they represented the major physiognomic re- mation about the natural history of this species. gions inhabited by Mountain Quail in northern Gutierrez (1980) has provided the only quan- California. The topographic and vegetative titative assessmentof habitat use by Mountain structure differed significantly among areas (Fig. Quail. He reported that a dense tree canopy and 2) and a breeding population of Mountain Quail steep slopes were important components of was present at each area (Brennan 1984). In con- Mountain Quail habitat in the Coast Range trast to the resident population studied by Gu- Mountains ofcentral California. Becausehis study tierrez (1980) populations of Mountain Quail at was conducted at only one area, a general pattern our study areas were migratory, forced by deep of habitat use was not established. snow to lower elevations each winter. Areas The purposeof this study was to examine hab- ranged from 500 to 1,200 ha in size, and from itat use by Mountain Quail at four sites located 1,100 to 2,100 m in elevation. over a broad geographic area. Characterization COAST RANGE (SECS. 14,23; T4N R4E) of the habitat features, with which birds are as- sociated,is the foundation of life history, behav- This area was within the mixed evergreen forest ioral, and evolutionary studies (Thorpe 1945, with chinquapin (Castanopsis chrysophylla) Southwood 1977, Rotenberry 1981). Our em- (Ktichler 1977) of Humboldt County. Three cov- phasis was on how different populations of this er types were present: mixed evergreen forest, quail were related to an array of vegetative and mixed brush, and oak (Quercus spp.) woodland. topographic features. The mixed evergreen forest was dominated by Douglas-fir (Pseudotsuga menziesii), and tanoak (Lithocarpus denszj7orus). The areas of mixed brush were composed of deerbrush (Ceanothus I Received 14 January 1986. Final acceptance 3 1 integerrimus), blue elderberry (Sambucus ceru- July 1986. 2Present address:Department of Forestry and Re- lea), and willow (Salix spp.). The oak woodland sourceManagement, 145 Mulford Hall, University of was dominated by Oregon white oak (Quercus California, Berkeley, CA 94720. garryana).
[661 HABITAT USE BY MOUNTAIN QUAIL 67
KLAMATH MOUNTAINS (SEC% 34,35; T46N I 122O R8W) This area was within the Klamath yellow pine forest (Kiichler 1977) of Siskiyou County. Two cover typeswere present:mixed forest and mixed brush. The mixed forest was composedof Jeffrey pine (Pinusjeffreyi) and Douglas-fir. The areas of mixed brush were dominated by whiteleaf manzanita (Arctostaphylosviscida) and deer brush. CALIFORNIA MODOC PLATEAU (SECS.29,30; T41N R6E) This area was within the yellow pine/shrub forest (Ktichler 1977) of Modoc County. Three cover types were present; pine-juniper forest, shrub- steppe, and basalt lava reefs. The pine-juniper forest had an overstory of Jeffrey pine and west- ern juniper (Juniperusoccidentalis), with occa- sional Douglas-fir and incense cedar (Calocedrus decurrens).Shrub-steppe areas were dominated by big sagebrush(Artemesia tridentata), green- FIGURE 1. Geographiclocations of the Mountain leaf manzanita (Arctostaphylospatula), and an- Quail studyareas in northernCalifornia. telope bitterbrush (Purshia tridentata). Pure stands of curlleaf mountain mahogony (Cerco- carpusledzfolius) covered many areas. Although tion of individuals,” whereas habitat selection the lava reefs were only sparsely vegetated, this implies that organisms “consciously choose cover type was included in the analysis because among alternative habitats” (Hutto 1985:457). Mountain Quail (1) used prominent lava rocks This distinction doesnot, however, eliminate the as crowing sites during the breeding season, (2) possibility that processesof habitat selection commonly crossedlarge expansesof lava to ob- might be responsible for the patterns of habitat tain drinking water, and (3) used crevices and use we observed (see Discussion). other openings in the lava as escapecover. Observed patterns of habitat use by birds often vary as a function of the scale at which the pop- NORTHERN SIERRANEVADA (SECS.25,26; T29N RllE) ulations were sampled (Wiens 198 1, 1985). With this in mind, we compared the habitats used by This area was within the Sierra montane forest and available to Mountain Quail at two spatial (Ktichler 1977) of Lassen and Plumas counties. scales:(1) macrohabitat (among cover types) and Two cover types were present: mixed forest and (2) microhabitat (within cover types). We defined mixed brush. The mixed forest had an overstory available habitat as the array of macro and mi- of sugar pine Jeffrey pine, (Pinuslambertiana), crohabitatsthat could be usedby Mountain Quail Douglas-fir, white fir and in- (Abiesconcolor), at our study areas. We estimated the density of cense cedar. Areas of mixed brush were domi- Mountain Quail at each study area so that we nated by Sierra chinquapin (Castanopsissem- could evaluate our habitat analyses in relation snowbrush and pervirens), (Ceanothusvelutinus), to local population abundance (cf. Wiens and greenleaf manzanita. Rotenberry 198 1, James et al. 1984). METHODS Evaluationof macrohabitatuse. We examined macrohabitat use by Mountain Quail (i.e., the EXPERIMENTAL DESIGN use of cover types in proportion to availability) This study was designedas a distributional rather at each area by using a chi-square contingency than a behavioral investigation. Therefore, analysis (Zar 1974:59). Expected frequency of throughout this paper, we employ the term hab- use values were based on the proportions of itat userather than habitatselection because the available cover types which we estimated from term habitat useindicates “the actual distribu- 1:24,000 scaleorthophotographic quadrangles of 68 L. A. BRENNAN, W. M. BLOCK AND R. J. GUTIBRREZ
selecting a random compass bearing, and then randomly choosinga distance between 0 and 100 m along this bearing. Subsequent plots were spacedat 50- or 100-m intervals (depending on the patch size of the cover type) along the random bearing. A total of 100 randomly located plots (25 at each area) were contrasted with an equal number of organism-centered plots for our with- in-site analyses. In our pooled analyses we con- trasted 114 organism-centered plots with 100 randomly-located ones. We contrasted microhabitat use and avail- ability in both univariate and multivariate space. FIGURE 2. Regional variation in the microhabitat Univariate tests for differences between and structure used by Mountain Quail in northern Cali- among individual variables were based on a fornia. Discriminant function analysis was based on Brown-Forsythe one-way analysis of variance 13 habitat variables (Table 2) measuredon 25 circular, (ANOVA) without the assumption of equal with- 0.02-ha nlots at each area (100 plots total). Ninetv- in-group variation (Brown and Forsythe 1974). nine percent of all sampleswere-associated with the correctstudy site. The two discriminantfunctions shown We felt justified in pooling data from acrossfour account for 91% of the variation across the four re- distinct areasbecause we wanted to obtain a gen- gions. Some points represent more than one sample. eral estimate of habitat use by Mountain Quail. Ellipses illustrate the range of discriminant scoresfor Floristic aspectsof use and availability were ex- each site. amined using Spearman’s rank-correlation anal- ysis (Zar 1974:243). Our multivariate contrast of each study area. Observed frequency of use val- use and availability was based on a logistic ues were based on the number of sightings of regressionanalysis (Cox 1970, Engelman 198 1) quail in each cover type. of the structural and topographic variables that Evaluationof microhabitatuse. We examined had statistical differences between the used and structural and topographic aspectsof microhab- available groups(see Results). In general, logistic itat use with a series of variables (based on Gu- regressionis used to derive a classification func- tierrez 1980, see Appendix 1, this paper) that tion from a series of “predictor” (in this case were measured on 0.02-ha (15 m diameter) cir- habitat) variables and then assesshow well this cular plots. We examined floristic aspectsof mi- function can predict which group the samples crohabitat use with percent relative cover values came from, much like discriminant analysis. A of speciesof woody plants that were known to fundamental difference between logistic regres- provide food resourcesfor Mountain Quail (based sion and discriminant analysis is that in logistic on food habits data from 559 Mountain Quail regressionthe analysis must be constrained to a collected from throughout California listed in two-group contrast.During the exploratory phase Appendix 2 of Gutitrrez 1977). of our analysis, we observed that logistic regres- The center of each habitat plot corresponded sion was superior to discriminant analysis for either to locations used by Mountain Quail or evaluating our use and availability data in multi- randomly located points. We usedorganism-cen- variate space(Brennan et al. 1986). tered habitat samples to estimate the microhab- Densityestimation. We estimated Mountain itat structure used by Mountain Quail. The lo- Quail density with variable-width line transect cation of the first quail (whether in a covey, in methods using the Fourier series estimator a pair, or a single bird) detected was used as the (Burnham et al. 1980). The observers walked center of a habitat plot. We sampled available transects(total lengths ranged from 12 to 14 km habitat structure with a systematic random de- at each area) and recorded distance and sighting sign where the number of plots at each area was angle deviation from the transect line to each stratified by the proportions of cover types pres- detectedquail (methods detailed in Brennan and ent. Within each cover type, we located a random Block 1986). We used the program TRANSECT starting point. From each starting point, the lo- (Laake et al. 1979) to calculate our density es- cation of the first habitat plot was obtained by timates. HABITAT USE BY MOUNTAIN QUAIL 69
TABLE 1. Chi-squarecontingency analysis of ma- crohabitatuse by MountainQuail at foursites in north- em California. : .z.:. . . . . : . .: : t:::.:...... 0 ...... Numberofquail ...... sightings Study area Propor- Ob- EX- Cover type= tion served pectedb P CoastRange : Mixed forest 0.32 16 13 t:: i . : i Mixed brush 0.60 30 35 t::.. :...... i t:.::: .:.. . . . :. . Oak woodland 0.08 4 2 >O.lO ...... Klamath Mtns. 0 0.25 M 0.50 0.75 1.0 Mixed forest 0.36 11 14 Mixed brush 0.64 29 26 >O.lO Modoc Plateau Pine-juniper 0.32 14 16 Shrub-steppe 0.36 31 18 Lava reefs 0.32 5 16
Sampling effort. We spent 250 person-days in the field locating and/or censusing Mountain shrub canopy cover) using logistic regression Quail (16 June to 10 October 1982, 14 April to showed a substantial improvement over classi- 7 October 1983). We collected density data dur- fication of the organism-centered and randomly- ing May and June 1983. located plots based solely on prior probabilities of group membership (Table 4). Histograms of RESULTS the predicted probabilities of group membership MACROHABITAT USE illustrate the classification results of all habitat The Modoc Plateau was the only one of the four samplespooled acrossthe four areas(Fig. 3). The areaswhere macrohabitat use by Mountain Quail distribution of the organism-centered samples was not in proportion to the relative areas of was skewed toward the high end of the proba- available cover types (Table 1). Mountain Quail bility scale.Conversely, probability scoresfor the used all of the vegetative cover types in propor- habitat samplesfrom the randomly-located group tion to their availability. The Modoc area was were distributed more-or-less evenly across the unique insofar as the lava reefs were used infre- entire probability scale (Fig. 3). The samples in quently by Mountain Quail. the randomly-located group contained a wider variety of habitat components (including the pro- MICROHABITAT USE portion of microhabitat structure usedby Moun- The average microhabitat structure used by tain Quail) than the organism-centered plots. Mountain Quail varied greatly among the four The relative cover values of speciesof woody study areas; 8 of the 13 variables showed statis- plants that were known to provide foods eaten tically significant differences (P < 0.01) across by Mountain Quail were greaterin the organism- the four areas (Table 2). When we contrasted centered group than in the randomly-located microhabitat use with availability, we detected group (Table 5). Nine of the 12 speciesused in significant differences(P < 0.0 1) in 4 of 13 vari- this analysis were shrubs. The observed r, (0.6) ables (Table 3). A mathematical combination of did not exceed the critical r, (0.9; df = 8; P > these four variables (distance to water, distance O.OS),therefore the two groups of relative cover to cover, maximum shrub height, and percent values appeared to be statistically independent. 70 L. A. BRENNAN, W. M. BLOCK AND R. J. GUTIERREZ
TABLE 2. Average microhabitat structurein four regions of the breeding range of Mountain Quail. Data are from 25 organism-centeredcircular plots (0.02 ha) measuredat each region (100 plots total).
Coast Range Klamath Mtns. Modoc Plateau N. Sierra Nevada Variable R SE f SE + SE R SE F-r&9
Basal area (mz/ha) 0.1 0.06 0.2 0.16 0.2 0.14 0.2 0.06 Elevation (m) 1,231.0 16.0 1,152.0 14.0 1,445.0 4.0 1,896.0 22.0 4:;.0
Distance to edgecover (m) (m) 2.01.1 0.440.34 0.43.2 0.181.2 0.13.5 0.061.2 2.21.2 0.440.42 n3,.’ Distance to water (m) 94.0 21.4 166.0 32.0 199.0 30.0 94.0 17.4 4.1 Litter depth (cm) 1.4 0.26 2.1 0.18 2.0 0.3 2.4 0.2 2.8 Maximum shrub height (m) 3.3 0.3 2.3 0.16 2.7 0.22 1.8 0.14 8.3 Minimum shrub height (m) 0.1 0.02 0.4 0.04 0.2 0.04 0.1 0.02 11.6 Percent dead material 19.7 4.1 15.0 4.4 17.2 3.7 23.5 5.1 ns Percent herb cover 25.5 4.3 7.1 3.8 26.8 4.7 14.3 5.3 4.0 Percent shrub canopy 37.1 4.4 48.8 5.2 41.8 4.7 50.6 5.6 ns Percent tree canopy 21.3 6.1 25.1 4.2 10.3 3.6 20.9 5.9 Slope (“) 24.6 2.3 19.9 2.3 5.1 1.5 20.3 1.9 f17s.5
a See Appendix for an explanation of the habitat variables. DAll F-ratios significant at P < 0.01, unless noted as not significant (ns); one-way analysis of variance.
DENSITY IN RELATION TO MICROHABITAT affected our ability to distinguish used from USE AND AVAILABILITY available microhabitat structure. Density estimates of Mountain Quail ranged from 9.0 (Modoc Plateau) to 30.0 (Klamath Moun- DISCUSSION tains) birds per 100 ha. There was an inverse Our analyses were based on statistical correla- relationship between density values and the av- tions and thus can only indicate a mathematical erage percentage of correctly classified habitat association between Mountain Quail and certain plots among the four study areas (Fig. 4). A sam- habitat features (cf. Wiens 1985). Therefore, we ple size of n = 4 precluded a regression analysis can only infer that mechanisms which govern of this relationship; however, it was apparent habitat selection might be responsible for the sta- that variation in local population abundance in- tistical differences between use and availability fluenced patterns of habitat use which in turn that we observed. Several potential biases may
TABLE 3. Average valuesof the microhabitat structureused by and available to Mountain Quail. Values given were obtained by pooling habitat samples from four regions in northern California.
Organism centered Randomly located (n = 114) (n = 100) Variable* R SE Range + SE Range F-ratiob
Basal area (m*/ha) 0.12 0.18 O-3.7 0.24 0.04 O-2.6 ns Elevation (m) 1,470.o 28.0 1,050-2,16 1 1,494.o 32.0 1,033-2,066 ns Distance to cover (m) 0.83 0.20 O-13.0 3.9 0.66 O-45.0 15.2 . Distance to edge (m) 2.49 0.362 O-27.0 3.90 0.68 O-45.0 Distance to water (m) 131.0 12.0 O-500 255.0 23.0 O-500 2’;:s Litter depth (cm) 1.8 0.11 O-6.7 1.8 0.10 O-4.6 ns Maximum shrub height (m) 2.4 0.12 O-6.4 1.9 0.12 O-5.0 8.2 Minimum shrub height (m) 0.28 0.07 O-l.0 0.14 0.02 O-l.0 ns Percent dead material 16.0 1.5 O-87.0 20.9 2.31 O-96.0 ns Percent herb cover 17.7 2.0 O-99.0 18.5 2.55 O-98.0 Percent shrub canopy 45.8 2.3 O-10 32.1 2.7 O-100 13s2 Percent tree canopy 18.4 2.2 O-100 25.5 3.17 O-100 ns Slope (“) 17.6 1.1 O-59.0 19.4 1.19 O-59 ns
p See Appendix for an explanation of the habitat variables. b F-ratios significant at P < 0.0 I, unless noted as not significant (ns); one-way analysis of variance. HABITAT USE BY MOUNTAIN QUAIL 7 1
TABLE 4. Classification results of Mountain Quail 100 MODOC PLATEAU and randomly-locatedhabitat plots by stepwiselogistic regression,based on four habitat variables.’
Percentageof habitat samples classifiedin the correctgroupb Studyarea samplesize Organism Randomly (quail : random) centered located Average N. SIERRA NEVADA Coast Range I 0 (2525) 88.5 68.0 78.4 Klamath Mtns. (25:25) 83.0 60.0 71.4 COAST RANGE Modoc Plateau I 0 (25:25) 91.0 100.0 96.0 70 N. Sierra Nevada i KLAMATH MTNS. (25:25) 78.6 84.0 81.3 All Areas T (114:lOO) 85.1 60.0 73.0 ol =Distance to water,distance to cover,maximum shrubheight, percent 10 20 30 shrubcanopy. b Prior probabilityof correctclassification based on relative samplesize BIRDS PER 100 HECTARES of eachgroup. For individual studyareas prior probabilitieswere 50:50, for the analysisacross all areasthey were 53:47. FIGURE 4. Relationship between Mountain Quail density estimates from the four study sites (abscissa) and the average percentageof organism-centeredand have affected these statistical differences. First, randomly-locatedhabitat plots from eachsite that were our results may be biased in favor of the most correctly classifiedby logistic regression(ordinate). conspicuous individuals because we used “the first quail seen” as the criterion for our samples Mountain Quail were in mixed evergreen forest of habitat use. Second, the effect of observer in- and chaparral cover types. Ninety-five percent fluence is difficult to assess.Most of the time it of our observations were also in mixed forest and was impossible to know if the quail detected us brush cover types. before we detectedthem, which may have caused Macrohabitat use by Mountain Quail differed them to move (and thus bias the location of the from macrohabitat availability only on the Mo- sample). Thus, we often did not know if the first dot Plateau. The availability of lava reefs clearly quail seen had been feeding, moving to water, exceeded the frequency with which Mountain roosting, involved in a social interaction with Quail used them. Lava reefs were the only type another quail, or moving in responseto the ap- of habitat at our study areas not previously iden- proaching observer. Using the locations of quail tified as a habitat used by Mountain Quail (cf. other than the first seen would have introduced even more bias. Becauseof these problems, we made no distinctions between different types (e.g., TABLE 5. Relative cover values of speciesof woody feeding, roosting, etc.) of habitat use, and instead plants that may provide food resourcesused by Moun- tain Quail. Data were obtained from 25 0.02-ha plots chose to treat the observations within a basic at four northern California areas (100 plots total). context of distribution. Time-dependent inter- actions were another potential sourceof bias. We Relative cover (%p tried to minimize this confounding effectby sam- Organism centered REz’y pling habitat use throughout the entire day. A Species study of time-activity patterns with prolonged Arbutus menziesii 11.5 1.0 observations of undisturbed quail would be a Arctostaphylos~pp.~ 16.0 15.8 2.0 good way to avoid thesesources of potential bias. Castanopsissempervirens 3.0 Ceanothusspp. ’ 34.0 26.7 Suchan approachwas, however, beyond the scope Sambucuscerulea 1.9 0.4 of this study. Symphoricarposalbus 3.9 1.5 Quercus~pp.~ 1.4 0.1 MACROHABITAT USE - Relative cover valuesdo not sum to 100% in eachcategory because Macrohabitat useby Mountain Quail in this study 25 additionalsnecies of woodvplants make up the balanceofthe relative COYU values. was similar to what Gutitrrez (1980) reported. bIncludes A. vrscida, A. patula. r IncludesC. ~uneafw C. integerrmus. C. velurinus. Ninety-seven percent of his observations of d IncludesQ. kelloggir. Q. garryana. 72 L. A. BRENNAN, W. M. BLOCK AND R. J. GUTIeRREZ
Grinnell et al. 19 18). At best, lava reefs provided stressthan low sparseshrubs. Thus, it seemsrea- some escape cover and prominent places from sonable that this quail should use areas where which territorial quail call. the shrub height and canopy coverage is greater that what is generally available. Although we em- MICROHABITAT USE phasize the role of shrubs in our analysis, other The identification of habitat featureswhich differ variables which showedlittle difference between in a use vs. availability contrast is indirect evi- the used and available groups (such as percent dence that birds recognize a specific configura- herb cover or litter depth) might also be impor- tion of habitat structure and settle in these areas tant biologically. to feed, loaf, roost, or breed (Hildtn 1965). Our Our results illustrate the importance of sam- results indicated that Mountain Quail were con- pling avian habitat use over a broad geographic sistently associatedwith a microhabitat config- area, and at different spatial scales(Wiens 198 1, uration that consists of tall and dense shrubs 1985). We detected significant differences be- which are in close proximity to drinking water tween use and availability at the microhabitat and escapecover (Table 3, Fig. 3). These statis- level, and few differences at the macrohabitat tical results can be interpreted in terms of the level. By sampling microhabitat use and avail- biology of this quail, especially when considered ability over a wide geographicarea, we brought in the context of a physiological need for water into question the value of slope as an important and food resources.Although we emphasize the component of Mountain Quail habitat. Gutitr- importance of food and water in our interpre- rez (1980) reported that steepslopes were an im- tation, we cannot completely discount the pos- portant factor in distinguishing between the hab- sibility that other processessuch as predation itats used by California and Mountain quails. (Hildtn 1965), climatic stress(Walsberg 1985) Our data indicated that topography alone prob- or even parasitism (Freeland 1983) may have ably has little value as a component of Mountain influenced the patterns of habitat use we ob- Quail habitat. Rather, it is the juxtaposition of served. Interspecific competition with California tall, dense shrubsin proximity to available water Quail (Callipepla calijixnica) was largely ruled that characterizesthe general pattern of habitat out by Gutitrrez (1980) as a process that has use by Mountain Quail in northern California. influenced habitat use by Mountain Quail. The respective ecologiesand biogeographic histories DENSITY IN RELATION TO MICROHABITAT USE of these two quails are vastly different (cf. Gu- titrrez 1980, Gutitrrez et al. 1983). Hildkn (1965), Noon et al. (1980), and Wiens Mountain Quail require drinking water during and Rotenberry (198 1) recognizedthat variation hot weather, and juveniles must drink soon after in local population abundance can have marked hatching if they are to survive (Grinnell and effectson patterns of habitat use by birds. Our Swarth 1913, McLean 1930, Rahm 1938). Be- results also support this contention. As the den- causeof this physiological requirement, the pres- sity of Mountain Quail increased acrossthe four ence of available drinking water is one habitat areas, we observed that a greater range of avail- component with clear ecological significance for able microhabitat structure was used (as shown this quail. The maximum distance of adult and by the correspondingdecrease in successfulclas- immature Mountain Quail from water observed sifications of the organism-centered and ran- by Miller and Stebbins (1964) at Joshua Tree domly-located plots). The inverse relationship National Monument was 1.6 and 0.8 km re- shown in Figure 4 indicates that Mountain Quail spectively-values which are slightly greaterthan density may be related to microhabitat use. Some those we observed (Tables 2 and 3). ultimate measure of habitat quality such as sur- Mountain Quail spent a great deal of time be- vivorship or reproductive success(cf. Van Home neath the perennial vegetationthat providesmany 1983) must, therefore, be obtained before we can food resources(GutiCrrez 1980). Our analysis of conclude that population density is an indicator the relative amounts of shrub speciesindicated of habitat quality for this quail. an apparent preference for those speciesthat are known to provide food resources.Presumably, ACKNOWLEDGMENTS the largeand well-developed shrubsprovide more C. W. Barrows,C. E. Braun, J. A. Crawford,R. L. food than small shrubs. Tall dense shrubs might Hutto, W. D. Klimstra, M. L. Morrison, M. G. Ra- also provide more shadeand relief from thermal phael,J. L. Roseberry,and J. Vemer reviewedearlier HABITAT USE BY MOUNTAIN QUAIL 73 versions of this paper and provided many suggestions Mountain and California Quails in the Carmel which improved it. G. M. Allen, B. R. Noon, and J. Vallev. California. Living Bird 18:7l-93. 0. Sawyer advised on the sampling design and statis- GUT&R&, R. J., R. M. ZINK~ANDS. Y. YANG. 1983. tical analyses.T. L. Pruden helped with proofreading Genie variation, systematicand biogeographicre- and provided important moral support. D. Bright, T. lationships of some galliform birds. Auk 100:33- Burton, A. Franklin, D. Garner, R. Golightly, C. Sisco, 47. and J. P. Ward helped in numerousways. F&ding was HILD~N,0. 1965. Habitat selectionin birds: a review. vrovided in vart bv The International Ouail Founda- Ann. Zool. Fenn. 2~53-75. iion, and thecalifdrnia Department of forestry-For- HUTTO, R. L. 1985. Habitat selection by nonbreed- est ResourcesAssessment Program. Computer services ing, migratory land birds, p. 455-475. Zn M. L. were provided by the Humboldt StateUniversity Com- Cody [ed.], Habitat selection in birds. Academic puter Center. Press, Orlando, FL. JAMES,F. C., R. F. JOHNSON,N. 0. WAMER, G. J. LITERATURE CITED NEIMI, AND W. J. BOECKLEN.1984. The Grin- nellian niche of the Wood Thrush. Am. Nat. 124: AMERICANORNITHOLOGISTS UNION.’ 1983. Check- 17-30. list of North American Birds. 6th ed. American K~~CHLER,A. ih7. 1977. Potential natural vegetation Ornithologists’ Union, Washington, DC. of California. Map. In M. G. Barbourand J. Major BRENNAN,L. A. 1984. Summer habitat ecology of [eds.], Terrestrial vegetation of California. John Mountain Ouail in northern California. Wiley and Sons, New York. M.S.thesis, H;mboldt State Univ., Arcata, CA. LAAKE,J. L., K. P. BURNHAM,AND D. R. ANDERSON. BRENNAN,L. A., ANDW. M. BLOCK. 1986. Line tran- 1979. User’s manual for program TRANSECT. sectestimates of Mountain Quail density. J. Wildl. Utah State Univ. Press, Logan. Manage. 501373-377. MCLEAN,D. D. 1930. The quail of California. Calif. BRENNAN,L. A., W. M. BLOCK,AND R. J. GUTIBRREZ. Dev. Fish Game Bull. 2: l-47. 1986. The use of multivariate statistics for de- MILLER;A. H., AND R. C. STEBBINS.1964. The lives veloping habitat suitability index models, p. 177- of desert animals in JoshuaTree National Mon- 182.In J. Vemer,M. L. Morrison, and C. J. Ralph ument. Univ. California Press, Berkeley. [eds.], Wildlife 2000: modeling habitat relation- NOON, B. R., D. K. DAWSON,D. B. INKLEY, C. S. ships of terrestrial vertebrates. Univ. Wisconsin ROBBINS,AND S. H. ANDERSON. 1980. Consis- Press, Madison. tency in habitat preference of forest bird species. BROWN, M. B., ANDA. B. FORSYTHE.1974. The small Trans. N. Am. Wildl. Nat. Resour. Conf. 45:226- sample behavior of some statisticswhich test the 244. equality of severalmeans. Technometrics 16:129- RAHM, N. M. 1938. Quail rangeextension in the San 132. BernardinoNational Forest. Calif. Fish Game 24: BURNHAM,K. P., D. R. ANDERSON,AND J. L. LAAKE. 133-158. 1980. Estimation of density from line transect ROTENBERRY,J. T. 1981. Why measurebird habitat?, sampling of biological populations. Wildl. Mono- p. 29-32. In D. E. Capen [ed.], The use of multi- gr. 72:1-202. variate statisticsin studiesofwildlife habitat. U.S. Cox, D. R. 1970. The analysisofbinary data. Methu- For. Serv. Gen. Tech. Rep. RM-87. en, London. SOUTHWOOD,T.R.E. 1977. Habitat, the templet for ENGELMAN,L. 1981. Stepwise logistic regression,p. ecologicalstrategies? J. Anim. Ecol. 46:337-365. 330-344. In W. J. Dixon, L. Engelman,J. W. Frane, THORPE,W. H. 1945. The evolutionary significance M. A. Hill, R. I. Jennrich,and J. D. Toporek [eds.], of habitat selection.J. Anim. Ecol. 14:67-70. Biomedical computer programs, P-series. Univ. VAN HORNE,B. 1983. Density as a misleading in- California Press, Los Angeles. dicator of habitat aualitv._ . J. Wildl. Manage. 47: FREELAND,W. J. 1983. Parasitesand the coexistence 893-901. of animal host species.Am. Nat. 121:223-236. WALSBERG,G. E. 1985. Physiologicalconsequences GRINNELL.J., H. C. BRYANT.AND T. I. STORER.19 18. of microhabitat selection, p. 389-413. In M. L. The game birds of Caiifomia. Univ. California Cody ted.], Habitat selection in birds. Academic Press, Berkeley. Press, Orlando, FL. GRINNELL,J., AND H. S. SWARTH. 19 13. An account WIENS,J. A. 1981. Scaleproblems in avian censusing. of the birds and mammals of the San Jacinto area Stud. Avian Biology 6:5 13-52 1. of southern California. Univ. Calif. Publ. Zool. WIENS. J. A. 1985. Habitat selection in variable en- 10:197-406. v&nments: shrub-steppebirds, p. 227-25 1. In M. GUTII&REZ,R. J. 1975. Literature review and bibli- L. Cody [ed.], Habitat selectionin birds. Academic ography of the Mountain Quail (Oreortyx pictus). Press, Orlando, FL. Svecialrevort. USDA Forest ServiceRenional Of- WIENS,J. A., ANDJ. T. ROTENBERRY.198 1. Censusing fice, San l%ancisco,CA. and the evaluation of avian habitat occupant< GUTIBRREZ,R. J. 1977. Comparative ecology of the Stud. Avian Biolow 6~522-532. Mountain and California Quails in the Carmel ZAR, J. H. 1974. BiosGtisticalanalysis. Prentice Hall, Valley, California. Ph.D.diss., Univ. California, Engelwood, NJ. Berkeley. GUTIBRREZ,R. J. 1980. Comparative ecology of the 74 L. A. BRENNAN, W. M. BLOCK AND R. J. GUTIBRREZ
APPENDIX 1. Explanations of the methods used to measure habitat variables on 0.02-ha, 15-m diameter plots at four areas in northern California. Plots cor- respond to Mountain Quail sightings and a random sample of available habitat stratifiedby vegetativecov- er type.
Variable Explanation
Basal area Sum of basal area at diameter of all trees within a plot at 1.37 m above ground. Elevation Altitude (m) above sea-level, based on an altimeter or to- pographic map. Distance to cover Meters to nearest escapecover measuredwith a tape or range finder. Distance to edge Meters to nearest patch of vegetation different from that at plot center. Distance to water Meters to nearest water from plot center. Litter depth Average litter depth taken from 10 measurementsalong a 15-m line intercept. _ Maximum shrub Height of the tallest shrub on height the plot. Minimum shrub Height of the shortest shrub on height the plot. Percent dead ma- Percentageof a 15-m tape in- terial tercepted by fallen logs, branches,and dead shrubs. Percent herb cov- Percentageof a 15-m tape in- er tercepted by grassesand forbs. Percent shrub Percentageof a 15-m tape in- canopy= tercepted by shrub foliage. Percent tree cano- Percentageof a 15-m tape in- a tercepted by tree foliage. Sl:ie Measured in degreeswith a cli- nometer.
aAlso usedto obtain relative cover values for eachspecies of woody plant along the 15-m intercept.