Temporal Variation in Resource Use by Black-Throated Gray Warblers

The Condor 101:67-75 0 The Cooper Ornithological Society 1999

TEMPORAL VARIATION IN RESOURCE USE BY BLACK-THROATED GRAY WARBLERS

JOHN J. KEANE* AND MICHAEL L. MORRISON~ Departmentof EnvironmentalScience, Policy, and Management, Universityof California, Berkeley, CA 94720

Abstract. We studiedforaging behavior and habitatuse of male and female Black-throated Gray Warblers (Dendroica nigrescens)to quantify the effects of temporal variation on interpretationsof avian resourceuse. Overall, both sexes primarily foraged by gleaning in single-leaf pinyon pine (Pinus monopbyla);however, within-seasonand between-year variation in behavior and habitat use were found for both sexes. Warblers increased use of shrubs from May to mid-June, and decreaseduse of shrubs and increased use of pinyon pine from mid-June through August, during each year. Use of sagebrush(Artemisia tridentata) and Utah juniper (Juniper osteospemta)varied between years. Sex differences were found in proportional use of foraging maneuvers.Within-season shifts in plant speciesand habitat use correspondedto changesin arthropodnumbers on the plant speciesused by the warblers, but between-year shifts in behavior did not correspondas closely with changesin arthropodnumbers. Temporal variation in microhabitatuse resulted from shifts within sea- sons in the plant speciesused for foraging, which was associatedwith temporal changesin food abundance.Our results also demonstratethe importance of consideringthe effects of temporal scale in studiesof bird-resourceinteractions. Key words: Black-throated Gray Warbler, Dendroica nigrescens,foraging behavior, habitat use, Neotropical migrant, resourceuse, temporal variation.

INTRODUCTION strated variation on daily (Hutto 198 lb), intra- Data from studies of avian resource use have seasonal (Sakai and Noon 1990), interseasonal been used widely to describe the structure and (Hutto 1981a), and annual (Szaro et al. 1990) composition of bird assemblages (MacArthur time scales as well as on local (Collins 1981) 1958, Holmes et al. 1979, Rotenberry 1985), and regional (Wiens and Rotenberry 1981) spa- and have been used in the development of spe- tial scales. Furthermore, variation can result cies-habitat models designed for management from intersexual differences and factors such as applications (Vemer et al. 1986). Wiens (1986), age and reproductive status (Holmes 1986, however, showed that interpretations of assem- Grubb and Woodrey 1990). Few researchers, blage structure can be influenced by inherent however, have attempted to address the poten- temporal and spatial variation. Rice et al. (1986) tially interactive effects of between-sex, within- and Rotenberry (1986) demonstratedthat the ac- season, and between-year variation in resource curacy and predictability of species-habitatmod- use patterns on bird-centered habitat measure- els can vary depending upon the temporal and ments (Kelly and Wood 1996). spatial scalesover which they are developed and The distribution and abundance of food is a primary factor affecting the resource use pat- applied. Thus, temporal and spatial variation at terns of birds (Lack 1966, Martin 1987). For in- various scales need to be described and consid- sectivorous birds, changes in arthropod distri- ered because they can influence the results of bution and abundance may correspondto chang- ecological investigations and any subsequent es in bird resource use patterns at different tem- management applications (O’Neill et al. 1986, poral and spatial scales.For example, Brush and Block and Brennan 1993). Styles (1986) found that changes in bird abun- Studies of avian resource use have demon- dance between pine- and oak-dominated vegetation types correspondedto changes in arthro- ’ ’ Received 1 December 1997. Accepted23 July pod biomass. Similarly, seasonalvariation in the 1998. sizes and speciesof trees used by bark foraging 2 Presentaddress: Stanislaus National Forest, 19777 birds in western North American forests is as- GreenleyRd., Sonora,CA 95370. z Presentaddress: Department of BiologicalScienc- sociated with variation in prey distribution and es, California State University, Sacramento, CA abundance (Morrison et al. 1985, Lundquist and 95819. Manuwal 1990).

[671 68 JOHN J. KEANE AND MICHAEL L. MORRISON

Our objective was to study the resource use moved along transect lines to avoid repeatedob- patterns of Black-throated Gray Warblers (Den- servations of the same bird. When a warbler was droica nigrescens) during the breeding season. detected,the observer watched the bird for 5 set We investigated within-season, between-year, without recording data. This period was used to and between-sex variation, and their interac- minimize bias to the most conspicuousactivities. tions, in foraging behavior and the habitat char- The observer then recorded all activities and lo- acteristics of foraging sites. Additionally, we cations where the activities occurred over the measured arthropod abundances on the plant next 15 sec. Only observations where the bird species used by the warblers to determine was in full view for the 15 set were included whether variation in warbler foraging behavior for analysis. Variables recorded were: sex of the and habitat relationships correspondedto chang- bird, time of day, plant species where the activ- es in arthropod abundances. ities occurred, height of the bird, frequency of use of foraging maneuvers (glean, hover-glean METHODS [= sally-hover], flycatch [= sally-strike to aerial STUDY AREA prey], flush-chase [= flutter-chase], fly-glean [= We worked from late April to August of 1989 sally-strike at stationary substrate],probe, peck, and 1990 in the White and Inyo Mountains, Inyo as described by Remsen and Robinson 1990), County, California (37”50’N, 118”lO’W). The and substrate where each activity occurred (fo- White and Inyo Mountains range in elevation liage, twig [< 1 cm diam.], small branch [l-lo from 1,220 to 4,345 m and each forms half of a cm diam.], medium branch [ 1 l-30 cm diam.], fault block extending approximately 170 km large branch [> 30 cm diam.], trunk, air, ground, north to south (Nelson et al. 1991). Climate is and flower). All measurementswere visually es- typically one of cold, snowy winters and warm, timated. All foraging locations were flagged for dry summers with summer precipitation occur- later habitat analysis. ring from thunderstorms (Powell and Klieforth Data analysis. We split the data for each year 1991). into two periods for analysis. Early summer was All fieldwork was conducted in pinyon-juni- defined as occurring from late April to 10 June, per woodland between 1,970-2,875 m elevation. and late summer as 11 June to late August. We Woodland overstory was composedof singleleaf justified this cut-point date biologically because pinyon pine (Pinus monophyla) and Utah juniper it correspondedwith observed plant phenologi- (Juniperus osteosperma). Antelope bitterbrush cal patterns. Bitterbrush flowered primarily from (Purshia tridentata) and big sagebrush(Artemi- late April through mid-June. These periods also sia tridentata) were the dominant componentsof broadly overlapped with changes in warbler the understory. Other shrubs present with local- breeding phenology (Keane, unpubl. data). Nest ized distributions included Parry rabbitbrush building was observed in late May and nests (Chrysothamnusparryi) and cliffrose (Cowania with eggs were located from late May through mexicana). The herbaceouslayer was poorly de- mid-June. Nests with young were found from veloped and consisted of grassesand forbs in- mid-June through early July. Thus early summer cluding eriogonum (Eriogonum spp.) and pen- broadly overlapped with the territory establish- stemon (Penstemon spp.). Spira (1991) provides ment, nest building, and incubation phasesof the further descriptions of plant assemblagesin the nesting cycle, and late summer broadly over- White and Inyo Mountains. lapped with the nestling, fledgling, and post- breeding phases, although overlap undoubtedly ACTIVITY BUDGETS occurred as a result of renesting attempts. Data collection. We sampled activity behaviors The raw data for plant species use for each of Black-throated Gray Warblers within a lOO- foraging observation was transformed from a m band on either side of 10 transect lines. Tran- discrete to a continuous variable by calculating sect lines were 4.2 km long, about 0.5 km apart, the corresponding number of observation sec- and located to span the elevational distribution onds for each plant species per 15-set sample of the pinyon-juniper woodland (see Morrison et (Brennan and Morrison 1990). This resulted in al. 1993 for further details). Four observers col- a value of percent time use for each variable. lected data during the study. All observers were We used a multivariate analyses of variance trained to standardize data collection. Observers (MANOVA; Green 1978) to investigate the ef- TEMPORAL VARIATION AND RESOURCE USE 69 fects of year, period, and sex on plant species the center of a 15-m radius plot. Within each use by Black-throated Gray Warblers. Multicol- plot we counted the number of pinyon and ju- linearity between variables was reduced by elim- niper trees within three height classes(1.5-3 m, inating one of a pair of variables with a product > 3-6 m, > 6 m). We established a 30-m line moment correlation > 0.70 (Sokal and Rohlf transect bisecting the plot along a random com- 1981). The variable retained was the one with pass bearing. We measured the presence of pin- the higher F-value in the analysis (Morrison et yon, juniper, bitterbrush, and sagebrushin four al. 1987). Variables in MANOVA testing of height classes (< 1 m, 1.1-3 m, 3.1-6 m, > 6 plant species use were percent time use of pin- m) along a vertical line at l-m intervals along yon, juniper, bitterbrush, and sagebrush.Use of the transect line. These data were used to cal- other plant species was I 1% for either sex in culate a measure of percent cover of bitterbrush any period and therefore was discardedfrom the and sagebrushand to develop pinyon and juni- analysis. Variables were transformed using the per foliage indices. Foliage indices were defined square-root arcsine transformation for percent- as the total number of height classes that con- age data to better satisfy the assumption of nor- tained pinyon or juniper foliage. At 5-m inter- mality, although univariate normality does not vals along the transectline, for a total of 6 points infer multivariate normality. Box’s M statistic per transect,we used a meter stick and measured was used to test the assumption of homogeneity the height and crown diameter to the nearest 0.1 of variance-covariance matrices (Green 1978). m of the nearest bitterbrush and sagebrushalong Although the ramifications of failing to meet this the compassbearing of the transect line. assumption are poorly understood, apparently Data analysis. We used a MANOVA to in- they are of far less concern than performing the vestigate year, period, and sex effects on the se- analysis with inadequate sample sizes (Williams ries of habitat variables, as described above for and Titus 1988). foraging data. We used nonparametric log-linear analyses to examine the effects of year, period, and sex on ARTHROPOD ABUNDANCES the use of foraging maneuvers and substrateuse Data collection. Sampling the arthropods avail- separately (Bishop et al. 1975). The initial ma- able to birds is not a trivial problem (Smith and neuver and the substrateat which it was directed Rotenberry 1990). It is difficult to determine for each individual observation were used in the precisely from which spectrum of the available analyses (Bell et al. 1990). Because expected arthropodsa bird is selecting prey (Hutto 1990). values within cells should be > 1 and no more Furthermore, each arthropodsampling method is than 20% of the cells should have expected val- biased toward that spectrum of the arthropodas- ues < 5 (Co&an 1954), it was necessary to semblage it effectively samples (Cooper and pool infrequently used maneuvers and sub- Whitmore 1990). Because Black-throated Gray strates.This resulted in three categories for for- Warblers have been described as foliage glean- aging maneuvers (glean, hover-glean, aerial). ers (Grinnell and Miller 1944), we used branch- Aerial included flycatch (54% of pooled cate- clipping to provide a measure of relative arthro- gory), flush-chase (26%) and other (20%). pod abundancesin the primary locations and on Three categories also were used for substrates the plant species used by foraging warblers (foliage, air, other). The “other” category in- (Cooper and Whitmore 1990). However, not all cluded twigs and branches (41% of pooled cat- food items used by the warblers were adequately egory), flowers (26%), terminal buds (19%), and sampled using this technique. other substrates(14%). We used step-wise pro- We sampled arthropods over a 2-day period cedures to remove interaction terms not signifi- at approximately 2-3 week intervals throughout cantly different from zero and selected the sim- the breeding period to provide measures of rel- plest models that best fit the data (P > 0.05). ative changesin abundance on the primary plant species used by the warblers across time. Each MICROHABITAT USE day was divided into morning (06:00-11:00) Data collection. We measured habitat variables and afternoon (14:00-19:00) sampling periods. at 50 randomly chosen activity locations per sex We sampled 25 each of pinyon and juniper in per period per year for a total of 400 habitat the morning and 30 each of bitterbrush and sage- plots. The activity location of the bird served as brush in the afternoon of the first day. This pro- 70 JOHN J. KEANE AND MICHAEL L. MORRISON cedure was repeated on the second day in re- 0.001, MANOVA) and years (F,,,,, = 7.42, P < verse order (i.e., shrubs in morning and trees in O.OOl), and between periods within years (F, 897 afternoon) to account for within-day variation in = 7.45, P < 0.001). Overall, warblers primarily arthropod numbers between plant species. used pinyon pine in all years and periods, with We sampled plants beginning from a random- use varying between 62.5% and 88.9% (Table ly selected sampling point along a randomly se- 1). Use of pinyon was greater in late summer of lected transect line. At five sample points locat- each year @‘,,a,, = 30.36, P < 0.001, ANOVA) ed at 75-m intervals along the chosen transect and differed between periods between years line, we sampled five each of pinyon and jumper (F,,,,, = 6.39, P < 0.01). Use of bitterbrush was trees or six each of bitterbmsh and sagebrush. greater in early stmnner of each year (F,,897 = We chose the closest five trees or six shrubs of 16.64, P < O.OOl), ranging from 6.3% to 14.8% each species> 10 m from the center of the sam- in early stmrmer versus 0.7% to 4.6% in late pling point by selecting the first tree or shrub summer. Use of sagebrushwas greater in early closest to a compassbearing due north and sam- summer of each year (F,,,,, = 36.97, P < O.OOl), pling subsequent trees or shrubs located at ap- greater in 1990 (F,,,,, = 23.03, P < O.OOl), and proximately 60-70” angles in a circle around the differed between periods between years (F,,,,, = sampling point. We collected four samples from 23.45, P < O.OOl), ranging from 2.3% to 22.6% each tree at approximately 2 m in height in the in early summer versus 0% to 2.0% in late sum- four cardinal compass directions. We placed a mer. Use of juniper was greater in 1989 versus wire-rimmed, canvas insect net over approxi- 1990 (F, ,897= 7.03, P < 0.01, Table 1). Box’s mately 0.5 m of terminal foliage and clipped the M was significant (P < 0.001). branch into the net. Branches were vigorously Foraging maneuvers. Stepwise log-linear shaken to dislodge arthropods, which were col- analysis indicated that the best model for war- lected in a plastic cup in the bottom of the net. bler use of foraging maneuvers included the pe- The cup was then quickly sealed, the branch riod-by-maneuver interaction, and the year and placed in a plastic bag, and both labeled. Shrubs sex main effects (x216 = 14.95, P = 0.53). were sampled in a similar fashion along six Gleaning was the most frequently used foraging points located at 75-m intervals. Only one clip- maneuver by each sex, ranging from 61% to ping was collected per shrub. Arthropod samples 80.6% for males and 68.6% to 78.8% for fe- were frozen at the end of each day and branch males between periods and years (Table 1). Hov- clippings were weighed to the nearest 0.5 g. Ar- er-gleaning was the secondmost frequently used thropod samples were subsequently sorted and foraging maneuver by both sexesin all years and all arthropodspicked out. periods except for females during period 2 of Data analysis. We used a series of ANOVA 1989. Use of hover-gleaning ranged from 12.0% (Sokal and Rohlf 1981) to investigate the effect to 20.4% for males and 8.1% to 23.5% for fe- of date on the number of arthropodsper 100 g males between periods and years. Use of aerial of foliage for each of the plant species in each maneuvers increasedin late stmuner versus early of the years. The four samples from each tree summer of each year for both sexes and ranged were averaged to provide a single measure for from 7.4% to 18.6% for males and 6.1% to that individual tree. Data were transformed with 19.4% for females between periods and years. the log(x + 1) transformation to better satisfy Both sexes decreased use of gleaning in late the assumptions of normality and homogeneity summer of each year versus early summer. of variances, as measured by visual inspection Males increased use of both hover-gleaning and of residual plots and Cochran’s C, respectively. aerial maneuvers in late summer versus early Significant differences (P < 0.05) between the summer of each year. Females use of hover- number of arthropods within each year and on gleaning was lower and use of aerial maneuvers each plant species were determined by multiple higher in late summer versus early summer dur- comparisons using Scheffe’s test (SPSS 1986). ing 1989, whereas in 1990 females use of hover- gleaning was higher and use of aerial maneuvers RESULTS was approximately equal in proportion in late ACTIVITY BUDGETS summer versus early summer (Table 1). Plant species use. Warbler use of plant species Substrate use. Stepwise log-linear analysis in- differed between periods (F,,,,, = 14.83, P < dicated that the best model included the year-by-

72 JOHN J. KEANE AND MICHAEL L. MORRISON

TABLE 2. Arthropods(number/100 g of foliage)sampled on plant speciesin the White and Inyo Mountains, California,during 1989 and 1990.Values are means2 SD.”

Plant species

Bitterbush Sagebrush Pinyon Juniper 1989 18-19 May 8.3 2 0.9A 18.8 5 2.9A 0.6 2 0.2A 0.4 2 O.lA 9-10 June 5.4 ? O.lAB 22.5 ? 4.OA 0.6 ? O.lA 0.5 + O.lA l-2 July 2.2 e 0.3c 5.4 5 1.2B 1.3 !I O.lB 0.6 t O.lAB 18-19 July 2.9 2 0.3BC 4.5 5 0.7B 4.2 ? 0.6C 1.0 -c O.lC 15-l 6 August 1.7 +- 0.3c 2.7 ? 0.4B 1.0 2 O.lAB 0.8 + O.lBC 1990 24-25 April 1.2 ? 0.2A 3.6 ? 0.6AB 0.2 5 O.lA 0.3 + O.lA 16-17 May 5.8 5 0.2D 3.4 5 0.5ABC 0.5 5 O.lAB 0.4 ? O.lAB 30-31 May 11.1 2 1.4E 10.6 t 1.6EF 0.9 2 0.2BC 0.6 ? O.lB 14-15 June 2.7 t 0.3ABCD 11.7 2 2.1DEF 0.8 ? O.lBC 0.6 2 O.lB l-2 July 3.3 ? 0.4BCD 13.0 ? 3.4F 1.0 2 O.lBCD 0.5 2 O.lAB 15-16 July 3.7 5 0.6CD 6.2 2 0.8CDEF 1.2 ” 0.2CD 0.4 ? O.lAB 6-7 August 2.7 2 0.3ABCD 4.2 5 0.5BCD 1.4 L 0.2DE 0.5 ? O.IAB 20-21 August 3.3 2 0.4BCD 1.2 2 0.2A 1.7 + O.lE 0.5 5 O.lB

a WIthin each plant species and year, values with the same letter are not significantly different (I’ < 0.05).

0.001) in 1990. Arthropod numbers on bitter- of the breeding season.Greater use of pinyon in brush were greatest during May and early June late summer of each year corresponded with of both years (Table 2). Arthropod numbers both increasesin the number of arthropodssam- were greatest on sagebrushin May and June of pled on pinyon and declines in arthropod num- each year. Arthropod numbers on pinyon in- bers on bitterbmsh and sagebrush. creased over the sampling period in both years. The patterns in plant speciesuse and at least The large increase in the sample for 19 July some substrateuse (e.g., use of flowers in early 1989 was due to a large number of hemipteran summer) observed in our study are similar to nymphs (Keane 1991). Arthropod numbers on results from other studies of within-season var- juniper increased in July and August of 1989 but iation in resource use by birds that have dem- remained fairly constant throughout the season onstrated variation associated with changes in in 1990. Sagebrushalways had higher arthropod plant phenology and/or prey availability (Root numbers than pinyon or juniper, and usually had 1967, Hejl and Vemer 1990, Miles 1990). For substantially higher numbers than bitterbrush. In example, Hejl and Vemer (1990) reported that 1990, although numbers on sagebrushwere low- both Bushtits (Psaltriparus minimus) and Oak er than in May and early June 1989, they were Titmice (Baeolophus inornatus) used buckbmsh still about 70% higher than bitterbrush (vs. (Ceanothus cuneatus) more frequently during 300% higher in 1989). March and increased use of blue oak (Quercus douglasii) during April in a California oak DISCUSSION woodland. Increased use of buckbrush corre- Our results show within and between season sponded to its flowering phenology and both variation in foraging behavior and habitat use species of bird increased use of flowers as for- patterns by Black-throated Gray Warblers. aging substrates,whereas increased use of blue Greater use of shrubs in early summer of each oak correspondedto its leafing out period. Sim- year corresponded with greater relative abun- ilarly, Root (1967) demonstrated that seasonal dances of arthropods on both bitterbrush and changes in Blue-gray Gnatcatcher (Polioptila sagebrushduring the early part of the breeding caerula) foraging behavior and prey availability season.The greater number of arthropodson bit- were associatedwith changes in platit phenolo- terbrush was associated with its early season gy. flowering phenology. Although sagebrush did Annual shifts in the relative use of plant spe- not flower until late summer-early fall, arthropod cies did not correspondas closely to changes in numbers also were greater during the early part arthropod numbers as did the within-season TEMPORAL VARIATION AND RESOURCE USE 73 shifts. Greater use of juniper in 1989 versus iation in substrateuse, but no monthly nor an- 1990 correspondedwith higher arthropod num- nual differences in use of plant speciesand for- bers during July and August of 1989 versus aging maneuvers. Warblers gleaned 2 90% of 1990. However, arthropod numbers on sage- the time during each month of both years of the brush were lower during May through mid-June study (Miles 1990). Morrison (1982) also re- in 1990 versus 1989, although warblers used ported that male (82%) and female (88.2%) sagebrushmore during this period in 1990 ver- Black-throated Gray Warblers foraged predom- sus 1989. Sagebrushhad the highest numbers of inately by gleaning throughout the breeding sea- arthropods during most sampling periods, and son in oak-conifer forests in Oregon. Gleaning these numbers were substantially higher in 1989 also was the predominant foraging maneuver than 1990. Use of sagebrushby the birds was, used by warblers in the pinyon-juniper wood- however, higher during 1990, although use av- land within our study area, although we found eraged only about 8% higher for males and fe- significant within-season, annual, and intersex- males combined (Keane 199 1). Corresponding- ual variation in the proportion of maneuvers ly, warbler microhabitat had greater percent cov- used. Miles (1990) observed that although use er of sagebrushin 1990. Because sagebrushand of plant species did not differ between months bitterbrush are interspersed throughout the pin- or years, use of substratesvaried monthly, with yon-juniper region, and because sagebrush is warblers using leaves relatively more early in more abundant than bitterbrush (Keane 1991), the breeding seasonand twigs and small branch- we think our annual results were mostly a re- es more frequently later in the breeding season. flection of relatively minor variations in plant Black-throated Gray Warblers are able to ex- speciesuse by the birds, and likely small differ- ploit a number of vegetation types throughout ences in our sampling locations along transects western North America. Holmes and Schultz between years. (1988) proposed that food availability for forest Male and female warblers did not differ in birds is a function of: (1) the types and abun- their proportional use of plant species,although dances of prey present, which varies among they did differ in their relative use of foraging plant species, (2) the foliage structure and char- maneuvers and substrates.In addition, males had acteristicsof the plants, which influence prey de- greater activity heights than females in early tectability and accessibility, and (3) the morpho- summer, although these differences were re- logical and behavioral abilities of a species to duced in late summer (Keane 1991). Differences perceive and capture those prey. Although between the sexesmay be a result of intersexual Black-throated Gray Warblers forage predomi- competition (Petit et al. 1990, Kelly and Woods nately by gleaning from foliage in a number of 1996). However, males and females did not dif- these vegetation types, our results indicate that fer in plant species nor microhabitat use, pre- they can exhibit significant temporal variation in dominantly foraged by gleaning, and had more foraging behaviors. Furthermore, the specific be- similar activity heights in late summer, suggest- haviors that vary and the magnitude of these dif- ing that the sex differences we found were gen- ferences differ between vegetation types. These erally consistent with the proposal that repro- observations suggestthat this speciesis morpho- ductive activities influence foraging and activity logically and behaviorally adapted to exploit a patterns of birds, rather than direct resourcepar- number of forest and woodland vegetation types titioning between sexes (Holmes 1986, Kelly and that differences in foraging behaviors be- and Woods 1996). tween vegetation types is a result of unique re- Black-throated Gray Warblers in our study sponsesto the particular vegetation structureand area exhibited within-season and annual varia- prey availability specific to each vegetation type tion in plant species and substrate use. Black- (Holmes and Schultz 1988). throated Gray Warblers are distributed through- We propose that temporal variation in forag- out a number of forest and woodland vegetation ing behavior and microhabitat use by Black- types in western North America (Bent 1938). throated Gray Warblers in our study resulted Miles (1990) investigated monthly and annual from shifts within seasonsin the plant species variation in foraging behavior of breeding used for foraging, which in turn was associated Black-throated Gray Warblers in Arizona oak with temporal changesin food abundance.These woodlands and reported significant monthly var- results indicate that intraseasonal, interseasonal, 74 JOHN J. KEANE AND MICHAEL L. MORRISON and sexual differences in resource use can be GRINNELL,J., AND A. H. MILLER.1944. The distribution significant sourcesof variation in studies of bird of the birds of California. Pac. Coast Avifauna 27: 1-615. foraging behavior and habitat use. Thus, future GRUBB,T. C., JR., AND M. S. WOODREY.1990. Sex, studies should address sourcesof temporal var- age, intraspecificdominance status,and the use of iation in order to improve understanding of avi- food by birds wintering in temperate-deciduous an foraging behavior and habitat relationships. and cold-coniferous woodlands: a review. Stud. Avian Biol. 13:270-279. ACKNOWLEDGMENTS HEJL,S. J., AND J. VERNER.1990. Within-season and yearly variationsin avian foraging locations.Stud. William M. Block generouslyshared his knowledge in Avian Biol. 13:202-209. the design phase of this study. Jean Block, William HOLMES,R. T. 1986. Foraging patternsof forest birds: Block. Laura Ellison. Linnea Hall. Ellen Inson. Amv male-female differences. Wilson Bull. 98: 196- Kuenzi, Adrienne Mello, Lisa Nordstrom: Stephanie 213. Shoemaker, Chris Solek, Dan Su, and Maite Sureda HOLMES,R. T., R. E. BONNEY JR., AND S. W. PACALA. assisted with the field work. Irene Timossi provided 1979. Guild structureof the Hubbard Brook bird computer consultation,and Lori Merkle assistedwith community: a multivariate approach.Ecology 60: manuscript preparation. Dave Trydahl, Janet Rantz, 5 12-520. Tony Kerwin, Elizabeth Phillips, and the rest of the HOLMES,R. T, AND J. C. SCHULTZ.1988. Food avail- staff of the White Mountain Research Station facili- ability for forest birds: effects of prey distribution tated completion of this study. Richard T Holmes, Ri- and abundanceon bird foraging. Can. J. Zool. 66: chard L. Hutto, John P Kelly, and three anonymous 720-728. reviewers provided constructivereviews of earlier ver- Hurro, R. L. 1981 a. Seasonalvariation in the foraging sions of the manuscript.Funding for this project was behavior of some migratory western wood w& provided by the Frank M. Chapman Memorial Fund, blers. Auk 98:765-777. the White Mountain ResearchStation, and the USDA Htn-ro, R. L. 1981b. Temporal patterns of foraging Forest Service, Pacific Southwest Forest and Range activity in some wood warblers in relation to the Experiment Station. Computer time and facilities were availability of prey. Behav. Ecol. Sociobiol. 9: provided by the Department of Forestry and Resource 195-198. Management, University of California, Berkeley. Hurro, R. L. 1990. Measuring the availability of food resources.Stud. Avian Biol. 13:20-28. LITERATURE CITED KEANE, J. J. 1991. Resource use by Black-throated Gray Warblers (Dendroica nigrescens) in the BELL, G. W., S. J. HEJL, AND J. VERNER.1990. Pro- portional use of substratesby foraging birds: mod- White and Inyo Mountains of California. M.Sc. el considerationson first sightingsand subsequent thesis, Univ. California, Berkeley, CA. observations.Stud. Avian Biol. 13:161-165. KELLY. J. P.. AND C. WOOD. 1996. Diurnal. intrasea- BENT, A. C. 1938. Life histories of North American sonal, and intersexualvariation in foraging behav- wood warblers. Dover Publications,New York. ior of the Common Yellowthroat. Condor 98:491- BISHOP.Y. M. M.. S. E. FEINBERG.AND P. W. HOLLAND. 500. 1975. Discrete multivariate ‘analysis: theory and LACK,D. 1966. Populationstudies of birds. Clarendon practice. MIT Press, Cambridge, MA. Press, Oxford. BLOCK,W. M., AND L. A. BRENNAN.1993. The habitat LUNDQUIST,R. W., AND D. A. MANUWAL.1990. Sea- concept in ornithology: theory and applications. sonal differences in foraging habitat of cavity- Current Ornithol. 11:35-91. nesting birds in the southern Washington Cas- BRENNAN,L. A., AND M. L. MORRISON.1990. Influence cades. Stud. Avian Biol. 13:218-225. of sample size on interpretationsof foraging pat- MACARTHUR,R. H. 1958. Populationecology of some terns by Chestnut-backedChickadees. Stud. Avian warblers of northeasternconiferous forests. Ecol- Biol. 13:187-192. ogy 39:599-619. BRUSH,T., AND E. W. STILES.1986. Using food abun- MARTIN,T E. 1987. Food as a limit on breeding birds: dance to predict habitat use by birds-p. 57-63. In a life history perspective.Annu. Rev. Ecol. Syst. J. Vemer. M. L. Morrison. and C. J. Ralph leds.1, 18:453-487. Wildlife ‘2000: modeling ‘habitat relationshipsof MILES,D. B. 1990. The importance and consequences terrestrial vertebrates. Univ. Wisconsin Press, of temporal variation in avian foraging behavior. Madison, WI. Stud. Avian Biol. 13:210-217. COCHRAN,W. G. 1954. Some methods for strengthen- MORRISON, M. L. 1982. The structureof western war- ing the common x2 tests. Biometrics 10:417-441. bler assemblages:ecomorphological analysis of COLLINS,S. L. 1981. A comparison of nest-site and the Black-throated Gray and Hermit Warblers. perch-sitevegetation structure for seven speciesof Auk 99:503-513. warbler. Wil&n Bull. 93:542-547. _ MORRISON,M. L., L. S. HALL, J. J. KEANE, A. J. KUEN- COOPER.R. J.. AND R. C. WHITMORE.1990. Arthronod ZI, AND J. VERNER.1993. Distribution and abun- sampling methods in ornithology. Stud. A;ian dance of birds in the White Mountains, California. Biol. 13:29-37. Great Basin Nat. 53:246-258. GREEN,P E. 1978. Analyzing multivariate data. Dry- MORRISON,M. L., I. C. T~MOSSI,AND K. A. WITH. 1987. den Press, Hinsdale, IL. Development and testing of linear regression TEMPORAL VARIATION AND RESOURCE USE 7.5

models predicting bird-habitat relationships. J. steppe birds: even “good” models cannot predict Wildl. Manage. 51:247-253. the future, p. 217-221. In J. Verner, M. L. Mor- MORRISON, M. L., I. C. TIMOSSI, K. A. WITH, AND l? rison, and C. J. Ralph [eds.], Wildlife 2000: mod- N. MANLEY. 1985. Use of tree species by forest eling habitat relationships of terrestrial verte- birds during winter and summer. J. Wildl. Man- brates. Univ. Wisconsin Press, Madison, WI. age. 49: 1098-l 102. SAKAI,H. E, AND B. R. NOON. 1990. Variation in the NELSON,C. A., C. A. HALL JR., AND W. G. ERNST. foraging behavior of two flycatchers:associations 1991. Geologic history of the White-Inyo range, with stageof the breeding cycle. Stud. Avian Biol. p. 42-74. In C. A. Hall Jr. [ed.], Natural history 131237-244. of the White-Inyo range easternCalifornia. Univ. SMITH,K. G., AND J. T ROTENBERRY.1990. Quantify- Calif. Press, Berkeley, CA. ing food resourcesin avian studies:present prob- O’NEILL, R. V., D. L. DEANGELIS,J. B. WAIDE,AND T lems and future needs. Stud. Avian Biol. 13:3-5. E H. ALLEN.1986. A hierarchical concept of eco- SOKAL,R. R., ANDE J. ROHLF.1981. Biometry. W. H. systems.Princeton Univ. Press, Princeton, NJ. Freeman, New York. PETIT,L. J., D. R. PETIT.K. E. PETIT.AND J. W. FLEM- SPIRA,T. I? 1991. Plant zones, p. 77-86. In C. A. Hall ING. 1990. Intersexual and temporal variation in Jr. led.], Natural historv of the White-Invo range foraging ecology of ProthonotaryWarblers during eastern-California. Univ. Calif. Press, Berkeley, the breeding season.Auk 107:133-145. CA. POWELL,D. R., AND H. E. KLIEFORTH.1991. Weather SPSS. 1986. SPSS user’s guide. 2nd ed. McGraw-Hill, and climate, p. 3-26. In C. A. Hall Jr. [ed.], Nat- New York. ural history of the White-Inyo range eastern Cal- SZARO, R. C., J. D. BRAWN,AND R. I? BALDA. 1990. ifornia. Univ. Calif. Press, Berkeley, CA. Yearly variation in resource-usebehavior by pon- REMSEN,J. V., JR., AND S. K. ROBINSON.1990. A clas- derosa pine forest birds. Stud. Avian Biol. 13: sification schemefor foraging behavior of birds in 226-236. terrestrialhabitats. Stud. Avian Biol. 13:144-160. VERNER,J., M. L. MORRISON,AND C. J. RALPH.[EDS.] RICE, J. C., R. D. OHMART,AND B. W. ANDERSON. 1986. Wildlife 2000: modeling habitat relation- 1986. Limits in a data rich model: modeling ex- ships of terrestrial vertebrates. Univ. Wisconsin perience with habitat management on the Colo- Press, Madison, WI. rado river, p. 79-86. In J. Verner, M. L. Morrison, WIENS,J. A. 1986. Spatial scale and temporal variation and C. J. Ralph [eds.], Wildlife 2000: modeling in studies of shrubsteppebirds, p. 754-172. In J. habitat relationships of terrestrial vertebrates. Diamond and T. J. Case leds.1. Communitv ecol- _ 1, , Univ. Wisconsin Press, Madison, WI. ogy. Harper and Row, New York. ROOT,R. B. 1967. The niche exploitation pattern of WIENS, J. A., AND J. T. ROTENBERRY.1981. Habitat the Blue-gray Gnatcatcher.Ecol. Monogr. 37:317- associationsand community structureof birds in 350. shrubsteppeenvironments. Ecol. Monogr. 5 1:21- ROTENBERRY,J. T 1985. The role of habitat in avian 41. community composition: physiognomy or florist- WILLIAMS,B. K., AND K. TITUS. 1988. Assessmentof its? Oecologia 67:213-217. sampling stability in ecological applications of ROTENBERRY,J. T 1986. Habitat relationshipsof shrub- discriminant analysis.Ecology 69: 1275-1285.