The Condor 93:71-81 0 The Cooper Ornithological Society 1991

BREEDING BEHAVIOR OF EVENING ’

AEIBY C. SCOTT*AND MARC BEKOFF~ Department of Environmental, Population, and OrganismicBiology, Universityof Colorado, Boulder, Colorado 80309-0334 USA

Abstract. From 1983-1987 we studied the breeding behavior of Evening Grosbeaks (Coccothraustesvespertinus) living in the Front Range of the Rocky Mountains of Colorado. These were typically monogamous; at only one of 64 nests (1.6O’o)was polygyny documented. Reproductive behaviors performed by adult males and females were first analyzed across all mating seasonsand then with respect to five consecutive nest stages (build/egglaying, incubate, brood, brood/fledge, fledge)and the outcome (successor failure) of individual nests.Adult malesand femalescontributed unequally to the ten most frequently observedbehaviors; half were performed almost exclusivelyby one or the other sex. Males typically fed and protected their mates and what we presumed to be their young; whereas females usually built the nest, incubated, brooded, and also fed their young. Feeding, soft calling from the perch, and fecal removal were performed relatively equally by adult males and females. The successor failure of nests was not associatedwith patterns of parental behavior. Our results indicate that Evening Grosbeaksformed a partnershipin whichthey divided the effortinvolved in nest preparationand raisingthe youngat the high altitude wherethis populationwas observed. Key words: Breeding biology; reproductivebehavior; parental investment;monogamy; Rocky Mountain National Park; Evening Grosbeaks;Coccothraustes vespertinus.

INTRODUCTION outside of the breeding season(September-April) Evening Grosbeaks (Coccothraustes vespertinus; at elevations ranging from 1,500-2,700 m, but hereafter called grosbeaks)are common inhabit- during the mating season birds were generally ants of coniferous foreststhroughout the western less social and moved to higher elevations. United States.Early studies were limited to gen- In order to formulate a more comprehensive eral anecdotal descriptions (Speirs 1968), obser- picture of the social ecology of a species,its re- vations of flocking and feeding at artificial feed- productive habits must also be examined. Such ing stations (Shaub 1963; Balph 1976, 1977; a study entails analyzing the behaviors that are Balph and Lindahll978, Balph and Balph 1979) usedduring different stagesof breeding and iden- and the assessmentofgrosbeaks ’impact on spruce tifying the proximate and evolutionary factors budworm (Choristoneura fumifeana [Clem.]) that might be responsible for observed mating populations (Blais and Parks 1964, Takekawa et patterns. Until the 1970s most researchon avian al. 1982, Langelier 1983). Our research has ex- mating systems concentrated on species-typical amined special organization, aggres- patterns of reproductive behavior (Vemer and sion, and dominance relationships (Bekoff and Willson 1966,1969; Lack 1968). However, more Scott 1989) and ecologicalinfluences on nest site recently there has been a shift of emphasis to- selection (Bekoff et al. 1987) and nest success ward gaining a better understanding of individ- (Bekoff et al. 1989). We found grosbeaksto be ual variations in mating behavior, behavioral ad- nonterritorial, seasonallygregarious flockers, that aptations associatedwith reproducing in varying behave similarly to other (Newton 1972), habitats, and environmental factorsthat may in- especiallyduring nonbreeding seasonswhen they fluence the evolution of mating patterns (Oring live in large groups. For example, in our study 1982). areas,flocks of 30-100 grosbeaksoften were seen The purpose of this study was to describe the breeding behavior of adult male and female gros- beaks living in the Front Range of the Rocky ’ ’ Received 7 May 1990. Final acceptance2 October Mountains of Colorado, analyze sex differences 1990. 2 Presentaddress: The National Outdoor Leadership in nest building and care-giving activities, and School, P.O. Box AA, Lander, WY 82520. speculate which behavioral patterns were most 3 Send reprint requeststo M.B. likely to be influenced by environmental factors.

[711 12 ABBY C. SCOTT AND MARC BEKOFF

We also wanted to discern if reproductive be- BEHAVIOR OF REPRODUCTIVE ADULTS havior of grosbeaks is similar to that of other AND YOUNG GROSBEAKS avian taxa in which monogamy has either been Eighteen different behaviors were observed dur- exclusively or most commonly observed (Vemer ing courtship and nesting, with 10 actions ac- and Willson 1969, Wittenberger and Tilson 1980, counting for 96% of the total frequency (n = Oring 1982, Welty 1982). Since previous work 4,254) with which all behaviors were seen. Vo- showed that ecological factors could be used to calizations (calls, soft calls, trills, and soft trills) predict nest successor failure (Bekoff et al. 1989) were categorizedaccording to sound quality and we also determined if patterns of reproductive the location at which they were performed rel- behavior could be used to discriminate between ative to the nest. Although some behavioral pat- successfuland unsuccessfulnests. terns were actually composite actions, each was recorded as only one of the eighteen different METHODS behaviors. For example, begging typically in- cluded a postural component, wing flutter, and Examination of grosbeak breeding behavior be- vocalization, all of which were necessaryfor an gan in April 1983. Due to grosbeaks’ high mo- activity to be classified as a bout of begging. bility (Bekoff and Scott 1989) only six banded The ten main behavioral patterns were: 1) individuals were found in the two study sites fie&ng(FEED); 2) begging(BEG): adult females, where we found concentrations of nesting gros- nestlings, and fledglings stretchedtheir neck up- beaks (see Bekoff and Scott 1989 for details con- ward, gaped, fluttered their wings (see below), cerning trapping, handling, and banding; USFWS and emitted a vocalization consistingof one fun- federal permit 21627). Therefore, most of our damental and two harmonics with a frequency data are from observations of unbanded birds. range of 2.0-7.0 hKz; 3) carrying nest material We found two separatelocations that contained (CNM); 4) building nest (BN); 5) removingfecal breeding pairs. One was in Rocky Mountain Na- matter (FR); 6) calling from a perch (CP): (the tional Park (EstesPark, Colorado; 60 km north- perch was fewer than 3.5 m from the nest tree) west ofBoulder; elevation 2,530 m) and the other calls refer to single- or double-note vocalizations just west of Nederland, Colorado (33 km south- with ascendingfrequency from 2.0-5.0 kHz lead- west of Boulder; 2,420 m elevation; see Bekoff ing quickly into descendingfrequency from 5.0- et al. 1987 for descriptions of the study areas). 2.0 kHz with no harmonics; 7) soft calling from Systematic searchesfor grosbeaknesting activity a perch (SCP): soft calls consist of an ascending were conducted in river and perennial creek fundamental of 2.0 kHz with two harmonics, drainages throughout the study areas from late ranging from 2.0-7.0 kHz (perch was less than April through June in each of five years (1983- 3.5 m from the nest tree); 8) calling from nest 1987). (CN): see6 for description ofcalls; 9) callingfrom Sixty four nests were watched daily for 60-80 nest tree (CNT): see 6 for description of calls; consecutive min during the two time periods and 10) soft trilling from nest (STN): soft trill (05:00-l 1:30 and 17:00-20:30; ambient tem- consists of two notes, each with one harmonic peratures ranged between 15-22°C) when the ranging from 5.0-8.0 kHz. The soft trill and trill highest concentration of observable grosbeakac- (see below) are structurally similar but differ in tivity occurred. These time periods were chosen use. Soft trills are usually heard during feeding after observations made during four consecutive and preening, whereas trills seem to be used as 24-hr watchesindicated that 98% (57/58 instanc- contact vocalizations. es) of grosbeakactivity that we could see(or hear) Eight other behavioral patterns that were not occurred during these times. statistically analyzed because of the low fre- Observations were made using 20 x 45 spot- quency with which each was observed (less than ting scopesor 10 x 25 binoculars. Data were 2% n) were: wing droop (n = 37): wings are held written or spoken into a tape recorder for later close to the body and lowered while the tail is transcription. Grosbeak vocalizations were re- cocked upwards; performed by males and fe- corded using a Sony F-26s cardioid dynamic mi- males during courtship and nest building; wing crophone mounted on a Sony PBR-3300 para- jlutter (n = 84): wings are outstretched and flut- bolic reflector. Tape speed was 19 cm/set. tered at a rapid cadence; adult females flutter BREEDING BEHAVIOR OF EVENING GROSBEAKS 73 their wings during courtship and begging, and carrying nest material or seen arranging material nestlings and fledglings flutter their wings when in a tree where a nest subsequently was used by begging;courtship dance (n = 1): during precopu- a grosbeak pair. The building/egg laying stage latory behavior, an adult male bent his head was only quantified when the entire building pro- backwards, swivelled his body back and forth, cesscould be observed. Building/egg laying was and successivelyfluttered and droopedwith wings characterized by short (< 3 min) frequent visits in close proximity to his mate; headflufl(n = 3): to the nest. Egg laying was assumed to be com- head feathersare raised to varying degrees,often pleted at the onset of incubation. The incubation in conjunction with wing flutter (especially dur- stage began as soon as a female was observed ing male courtship); headfluff is also observed sitting on the nest for more than five consecutive during agonistic encounters; copulation(n = 12): minutes. Incubation was characterized by a fe- the male performs a wing flutter while facing the male sitting on the nest for longer periods of time female, the female moves in front of the male (usually 1Cl-50 min). Mean incubation time cal- and executesa wing flutter with her back to the culated from periods of known length was 26.6 male, and the male then mounts the female; cop- f 24.7 min. During incubation females left the ulation occurs for approximately 3-5 set after nest unattended only for short periods of time which the male dismounts while still fluttering (X = 2.4 f 3.4 min), but when they were gone his wings; bill touch (n = 10): mated pairs touch males usually remained near the nest tree. Brood- beaks with no food exchange;practice flight (n ing began as soon as the adult male or adult = 16); and trillingfrom nest(n = 2 1): trills consist female left the nest for short (3-5 min) excursions of one fundamental with a frequency range from and returned with food. On most occasionsnest- 2.5-5.0 kHz and one harmonic with a frequency lings were fed by the returning adult. Nestlings range from 5.5-7.0 kHz. were usually big enough to be observed two or four days after feeding by adults began, and nest- TIME BUDGET ANALYSES lings could be observed throughout brooding. Brooding ended as soon as one nestling fledged. The amount of time a female spent on or away Becausenestlings did not fledge simultaneously, from the nest was recorded for each observation we labeled a stage“ brood/fledge” to indicate the period. The following three measureswere used period when a female was still brooding some in time budget analyses: 1) total time within an nestlings after other young had fledged; her be- observation period that females spent on the nest havior was not a result of a second or double (TOTFON); 2) the time that females spent on clutch. The brood/fledge stage did not occur at their nest during each visit (FON); and 3) the all nests.Thefledgling stagebegan after all known time females spent off their nest each time they nestlings had fledged and lasted as long as the departed (FOFF). fledglings were being fed by at least one parent. We also calculated the frequency with which Nests also were classified as being successful males visited their nests(MVN). Frequency rath- or unsuccessful.Successful nests were those that er than mean time is reported here becausemales resulted in at least one known fledgling, whereas usually spent less than 2 min at their nest. unsuccessfulnests were those that failed during building, incubation, or brooding. Nest failure NEST STAGES and abandonment were obvious; adult birds did Nest stageswere assigned on the basis of adult not return to the nest and no grosbeaks were male and female behavior. At first, we climbed subsequently observed in the area. trees in order to observe the number of eggsand to assign hatching dates; however, this practice STATISTICAL ANALYSES was discontinued because of nest failure imme- The number of nests found each year was dif- diately after these efforts. Nests were otherwise ferent, consequently the total frequencies of be- inaccessible so other attempts at direct verifi- havioral patterns observed also varied among cation were unsuccessful.Mean duration of each years. However, because mean rates of perfor- nest stage was quantified only from nests with mance of behavioral patterns (in terms of bouts complete data. Building/egg laying was assigned per number of nestsobserved each year) and time as a stageas soon as an individual was observed budgets did not vary in pair-wise comparisons 74 ABBY C. SCOTT AND MARC BEKOFF between years or between nests, data from all most nests were begun in late May or early June. years and nests were combined. Sixty-four nests were found. Most were found SPSSstatistical programs (Nie et al. 1975) were during building (n = 36; 56%) and incubation (n used for univariate and multivariate analyses. = 22; 34%). Two nests were reused for second Means f standard deviations (SD) are reported. clutches and none were used in subsequent Behaviors were analyzed only for nests at which breeding seasons.We assumedthat the two gros- the duration of observed nest stages could be beaks observed feeding their young were their determined. Behavioral patterns were recorded parents;no third birds were observed.Sixty-three as individual “bouts” when there was a clear of 64 (98.4%) mated pairs were assumed to be transition between different activities; they were monogamous based on behavioral observations analyzed with respect to the age and sex (adult (see below). male, adult female, nestling, or fledgling) of the During courtship,adult males and femaleswere that performed (or initiated) them and where secretive.Neither sexperched conspicuously and appropriate, the age and sex of recipients. Chi- after an individual was observed, following it square analyses were used to test for differences usually did not result in our finding a nest. Vo- in the frequency distributions of each reproduc- calizations occurred sporadically and did not tive behavior acrossall years and of patterns of permit us to localize pairs near their nest. The behavior initiated and received by different in- most commonly observed behaviors during dividuals. Kruskal-Wallis analyses of variance courtship period were wing flutter, calls from were used to determine if there were differences perch, and wing droop (Table 1). Other behaviors among nest stagesin terms of 1) the mean fre- observed during courtship were begging,feeding, quencies of occurrence of the ten main behav- bill-touching, head-fluffing, soft-calling from a ioral patterns and 2) the variables described in perch, and calling from the nest tree. time budgets(TOTFON, FON, FOFF). For these When courtship behaviors were observed at analyses we assumed that there was indepen- the nesting sites, the male and female appeared dence among the different nests and also among already to have formed a pair bond. The one sequential nest stagesfor the same nest in a given courtship dance we observed occurredjust prior year becauseof the extensive time intervals over to copulation. Wing flutter performed by the adult which data were collected (see Hejl et al. 1990). female occurred most frequently during court- Successfuland unsuccessfulnests were com- ship in response to a male’s wing droop. Wing pared using Mann Whitney-U tests with respect flutters performed by adult females during in- to 1) the mean frequencies of occurrenceof each cubation appearedto be a form of food-soliciting of the ten main behavioral patterns, 2) the vari- and were most often seen in conjunction with ables described in the time budgets (TOTFON, begging. Wing flutter performed by an adult ap- FON, FOFF), and 3) the mean frequencies of pearsto be a “socio-infantile” behavioral pattern male visits to the nest (MVN). Stepwise logistic (Wickler 1968) commonly observed in feeding, regression(Kachigan 1986) was used to examine greeting, and submissive behavior. the possibility that a single behavior, or com- Nest construction was performed almost solely bination of behaviors, could be used to discrim- by females (141/151; 93.4%). Males helped min- inate between successfuland unsuccessfulnests. imally by occasionally carrying nest material to the nest (13/20 1; 6.5%). All four types of vocal- RESULTS izations (calls, soft calls, trills, soft trills) were The mean duration (days + SD) of each nesting heard during building. Lone females were ex- stage was: build/egg laying = 6.3 & 4.4; incu- tremely secretiveduring egg-laying. In fact, when bating = 13.4 + 4.0; brooding = 14.1 + 4; and females were observed during nest construction brood/fledge = 2.4 f 1.2 Grosbeaks arrived in and egg-laying, they were accompanied by a male their breeding areas during mid-to-late May as either at the nest site or while flying in and out a mated pair. Observations in both study areas of the nest tree. began in April and there was no grosbeakactivity During incubation, males often foragedwithin until pairs arrived. Almost immediately, pairs sight of the nest tree, but they also were observed began nest building; there was no indication that flying long distances (up to 1,000 m) during a they attempted to establish territories. The first foraging excursion. Females exclusively incu- nest we observed was on 18 May 1985, however, bated eggs,while males typically fed their mate BREEDING BEHAVIOR OF EVENING GROSBEAKS 75

TABLE 1. Frequencies of reproductive behaviors observed during courtship and different nest stagesat 64 evening grosbeaknests.

Build/egg BKWd/ Behavior courtship laying Incubation BEGil Fledgling fledge TOti

Courtship dance 100 - - - - - Bill touch 100 - - - - - 10 Head fluff 100 - - 3 Wing flutter 37 s ii 16 11 1 84 Wine. droon 62 22 13 - 3 - 37 Cop;lation 17 50 33 12 Begging 1 1 54 25 B 11 449 Feeding - 11 75 6 7 1,415 Carry nest material - 84 16 - - - 201 Build nest 18 1 - - 151 Soft calls from perch : 818 35 20 20 11 224 Calls from perch 5 I 63 20 2 3 680 Calls from nest 70 10 7 2 369 Calls from nest tree < 1117 56 9!: : 4310 213 Trills from nest - 9.5 38 21 Soft trills from nest - 16 60 243 Fecal removal - - - ;; -? 57 125 Practice flight - - - 38 6 56 16 Total 129 506 1,529 1,592 242 256 4,254

by regurgitation. After feeding their mate, males throughout nesting are presented in Table 1. then perched nearby, presumably defending the Feeding occurred most often, with the next most nest tree. Calls from a perch, the nest, or the nest common behaviors being vocalizations from a tree, were most commonly heard during incu- perch, vocalizations from the nest, and begging. bation. These calls appear to be location and/or The frequency of occurrence of each behavioral defensecalls (Ritchison 1980); we found that the pattern varied throughout the breeding season frequency of these calls increased during nest (Kruskal-Wallis analysis of variance, in all cases stagesin which eggs or nestlings were present. P < 0.0001). For example, during building/egg Both nestlingsand fledglingswere capable of pro- laying 57% of all observed behaviors involved ducing vocalizations which sounded similar to carrying nest material or nest building. The be- adult vocalizations. However, young birds only haviors most commonly observed during incu- vocalized when their presumed parents were ob- bation were: calls from perch (28%); calls from served near their nest tree. nest (17%); and begging(16%). Feeding occurred Fledglings were observed in the nesting area most during brooding(67%) with calls from perch for two to five days after leaving their nest. Their being the next most common behavior (9%). plumage was dull gray (similar to that of adult During thefledgling stage,feeding was again the females) and they were well-camouflaged in the most common behavior (36%), with soft calls habitat surrounding the nest. Most nestlings from a perch (19%) and begging (15%) the next fledged by late July. Fledglings were fed by both most commonly observed activities. The most adults at feeding stations in lower elevations commonly observed behaviors performed dur- throughout August and the first three weeks of ing brood/fledge were: feeding (37%); begging September (Bekoff and Scott 1989). During the (18%); calls from nest tree (8%); and calls from brood/fledge stage, one to two fledglings were a perch (7%). often observed near the nest tree; they perched Adult males and females contributed unequal- quietly within 20 m of the nest tree and vocalized ly to the ten main reproductive behaviors (Fig. only when adults were nearby. Nestlings begged 1; x2 = 1,022.79, df = 9, P = 0.0001). Half of and trilled more than did fledglings. these behaviors were performed almost exclu- Frequencies of occurrenceof the 18 behaviors sively by one or the other sex. The behaviors (n = 4,254) observed during courtship and performed relatively equally by adult males and 76 ABBY C. SCOTT AND MARC BEKOFF

20 Li f-l OVERALL MATE NESTLINGS FLEDGLINGS - EEG CM/I BN FEEDSCP CN CP CNT STN FFl RECIPIENTS BEHAVIOR FIGURE 2. Percentageof feeding bouts initiated by FIGURE 1. Percentagesof behavioral patterns per- adult males (cross-hatchedbars) and adult females(sol- formed by adult males and females for all nest stages id bars) and the recipients of food delivered by each, combined. BEG, begging;CNM, carrying nest mate- for all nest stagescombined. rial; BN, build nest; PEED, feeding;SCP, softcalling from perch;CN, calling from nest;CP callingfrom a perch;CNT, calling from nesttree; STN, softtrilling from nest;and PR, fecalremoval. quencieswith which adult males or adult females performed the ten main reproductive behaviors when comparing successful with unsuccessful females were soft calls from perch, feeding, and nests. Furthermore, stepwise logistic regression fecal removal. However, feeding was nonran- showed that no behavioral patterns were useful domly distributed among recipients (Fig. 2; x2 = for discriminating between successfulor unsuc- 230.33, df = 3, P = 0.0001). Females fed nest- cessful nests (P > 0.05); only 69% (n = 44) of lings more than males did; both adults fed fledg- all nests were correctly classified. For both suc- lings equally and only males fed their mates.Mean cessful and unsuccessful nests, the number of values for females’ time spent on and away from nestlings ranged from 1 to 4. The mean number the nest, and the frequency with which males’ of nestlings for successful(n = 35) and unsuc- visited nests, also varied throughout the nesting cessful (n = 29) nests was 2.9 (SD = 0.98) and stages(Table 2). 3.2 (SD = 1.0) respectively (P > 0.05). Patterns of change in the relative percentages of behaviors performed by adult males, adult DISCUSSION females, nestlings,and fledglingsduring: a) build- ing/egg laying, b) incubating, c) brooding, d) BEHAVIORAL CHANGES AND PARENTAL fledgling,and e) brood/fledgeare presentedin Fig- INVESTMENT DURING BREEDING ure 3. When both nestlings and fledglings were The present study demonstrated that the behav- present at the nest site (brood/fledge), the per- ior of adult grosbeakschanged at the beginning centagesof behaviors performed by adult males, of, and throughout, the breeding season.At the females, nestlings, and fledglings were different onset of the breeding seasonwe observed a shift from the patterns observed during either brood- in grosbeak social organization from one char- ing or fledging. During the brood/fledge stage, acterized by gregarious flocking to one charac- nestlings begged more than fledglings (34/46 terized by smaller groups, pairs, and singletons [74%]) and soft trilled from the nest more than (Bekoff and Scott 1989). This shift in social or- fledglings (9/l 2 [75%]). Neither nestlings nor ganization may be grosbeaks’ responseto an in- fledglingscalled from the nest tree. Calls and soft creasein food availability. Since food appearsto calls from fledglingsin percheswere rare (n = 6). be more abundant and more evenly distributed The most common vocalization by fledglingswas during the breeding season when compared to soft call from a perch (n = 38). other times of the year (Bekoff et al. 1989), gros- There were no differences in the mean fre- beaks may not benefit from group foraging while BREEDING BEHAVIOR OF EVENING GROSBEAKS 77

TABLE 2. Time budgetsfor mean time (f standard deviation) engagedin three activities during each of four nesting stages.The three activities were: (1) the total number of minutes females spent on their nest in each stageduring 60 min watches(TOTFON), (2) the number of minutes per hour females spent on their nest during each visit (FON), and (3) the number of minutes per hour females spent off their nest each time they departed (FOFF). The frequencyper hour with which males visited their nests(MVN) is also presentedin this table. The column labeled “Mean Frequency” indicatesthe mean frequencyof occurrenceof eachactivity performed during a 60-min observation period for all stagescombined.

Stage Variable Build/e~ laying IXlhtiOll Brood BKd/FQdg~ Mean frequency TOTFONA 15.0 50.0 28.0 4.3 36.0 n = 1,285 (k11.3) (e27.0) (k21.8) (k1.8) (+27.1) FONA (& 26.6 10.6 1.3 15.0 II = 2,568 (+24.7) (+ 14.8) (+ 14.3) (k20.2) FOFFA 6.2 10.1 11.6 15.3 10.9 n = 2,362 (+9.2) (-t 10.7) (k9.7) (+ 14.3) (k11.2) MVNA 8.2 3.1 4.0 n = 1,071 (k8.4) (Z) (+ 1;::) (? 1.6) (e5.7) * Kruskal-Wallis analysisof variance,comparing means for eachnest stage for eachvariable; P < 0.0000 I.

breeding (Bekoff and Scott 1989). While indi- males contribute about equally to care-giving viduals living in large flocks may be at an ad- (Sullivan 1990). vantage both with respect to the detection of predatorsand the protection that a group confers, BEHAVIOR AS A FACTOR IN it also may be that by foraging independently or NESTING SUCCESS in pairs or small groups, individuals reduce the There was little variation in the behavior of adults likelihood of attracting predators. attending successful nests or nests that failed be- Although we refer to the behavioral patterns fore fledglings emerged. Successful and unsuc- that we observed during nest building and care- cessful nests could not be differentiated using giving as “parental investment,” we do not know behavioral data. However, in a previous study that these actions actually “increase the off- specific habitat characteristics were predictably spring’s chance of surviving (and hence repro- associated with the success or failure of grosbeak ductive success)at the cost of the parent’s ability nests (Bekoff et al. 1989). The ecological features to invest in other offspring” (Trivers 1972, p. that were associated with successful nests ap- 139). Under natural conditions it is common for peared to be important in reducing predation by researchersto be unable to assesseither the out- terrestrial and aerial predators and also facili- come of different rearing strategieswith respect tated thermoregulation of the nest’s occupants. to offspring survival or if care provided to young Thus, our data indicate that the initial location is really costly to care-givers (Byers and Bekoff of the nest site influences the eventual outcome 1990). Reliable comparative measures of “pa- of the nest more than does the behavior of either rental investment” also are difficult to come by member of the mated pair after the nest site is (e.g., Cash and Johnson 1990, Gori 1990). established. Our assessmentsof parental investment reflect The dominance rank of a male also might in- the amount of time spent performing specificac- fluence nest placement, but we were unable to tivities or the frequency with which different be- assess this possibility. Hill (1988) found that havioral patterns were initiated during the breed- dominant male Black-headed Grosbeaks (Pheuc- ing season. Sex differences indicated that male ticus melanocephalus)tended to occupy more and female grosbeaksformed a partnership that heterogeneous habitat which contained fewer resulted in each parent providing a different type Steller’s Jays (Cyanocittastelleri), the most com- of care to each other and to their young. Some mon predator of Black-headed Grosbeak nests. other passerineshave been shown to have similar Time and energy expended during interspecific partnerships in which adult males and adult fe- aggression also may be reduced by 1) placing 78 ABBY C. SCOTT AND MARC BEKOFF

a b 100 100

$j 80 80 a 5 60 60 Y u 40 40

kt 20 20

0 0 . . UWBNFEEDC?SPCNCNTSlN - BEGCMABNRB)CPSYlXISTN C d 100

80

60

40

20 0

BEHAVIOR BEHAVIOR e 100 I I 80 Zl male n female 60 q nestling 40 q fledgling 20 n

BEHAVIOR FIGURE 3. Relative frequencieswith which adult males, adult females, nestlings, and fledglingsperformed behaviors during (a) building/egg laying, (b) incubation, (c) brooding, (d) fledgling, and (e) brood/fledge stages. CNM, carrying nest material; BN, building nest; BEG, begging;FEED, feeding; CP, calling from a perch; SCP, soft calling from a perch; CN, carrying nest material; STN, soft trilling from nest; and CNT, calling from nest tree.

nestsaway from areas inhabited by nonpredatory adees(Parusgambeli), Golden-crowned Ringlets competitors, a situation that we did not encoun- (Regulus satrapa), and Green-tailed or Rufous- ter on our study area, or by 2) establishing nest sided Towhees (Pipilo chlorurusand P. erythro- sites and feeding in areas where nonpredatory phthalmus) were not chased from the nest or noncompetitors live. Interspecific aggressivein- nearby trees. teractions rarely were observed between gros- beaks and similarly-sized nonpredatory non- REPRODUCTIVE LIFE HISTORY STRATEGIES competitors in the same area (Black-headed AND MATING PATTERNS Grosbeaks; Robins, Turdus migratorius). Small- Life history strategies, including reproductive er noncompeting nonpredatory speciessuch as behavioral patterns and associated mating sys- Pine Siskins (Carduelispinus), Mountain Chick- tems, evolve as a result of complex interactions BREEDING BEHAVIOR OF EVENING GROSBEAKS 79 between and their environment (Par- was seen only during the build/egg laying stage tridge and Harvey 1988). We speculatethat gros- and may be a way in which males “insured their beaks are adapted behaviorally to breed in hab- paternity” (Gowaty 1983, Fitch and Shugart itats that often have severe weather and short 1984, Gowaty and Plissner 1987); 2) there were growing seasonsbecause they: 1) devoted little very few instances of male-male aggressionnear or no time to prenesting activities compared to the nest after the nest was completed; 3) even the time spent performing other reproductive be- when two males were seen near the nest, only haviors, 2) did not establish breeding territories, one of them was tolerated by the female and the 3) spent only a short time building nests, and 4) other male would usually leave the nest tree; and produced fledglings that only remain in the 4) the male fed the female and defended the nest breeding area for a short period of time. White- tree during incubation and contributed to rearing crowned Sparrows (Zonotrichia Zeucophrys)liv- young throughout all nest stages. ing at high latitudes also show similar behavioral In many vertebrates, monogamy is thought to patterns when compared to conspecificsnesting evolve in responseto specificenvironmental and at lower latitudes and in warmer climates (Mor- social conditions (Kleiman 1977, Wittenberger ton 1976). and Tilson 1980). Although we did not directly Birds nesting at higher latitudes typically tend test Wittenberger and Tilson’s (1980, pp. 199- to produce larger clutches than those nesting at 200) hypotheses,our data support two of five of lower latitudes and in warmer climates where the their suggestions.First, male grosbeakscontrib- breeding seasonis longer (Van Tyne and Berger uted substantially to all nestsand most behaviors 197 1). Mean clutch size in this study was three, performed by the mated pair were displayed ex- which is smaller than the averagethat is typically clusively by one parent or the other. Second, reported for other (Austin 1968). becausethere were no caseswhere females raised However, no comparative data exist for gros- young without adult male assistance,it seemsas beaks breeding at different altitudes or latitudes. if participation by males in reproductive efforts Lastly, grosbeaksonly produced one brood per is necessary but not sufficient. Thus our data season, which is common for birds breeding at suggestthat “male parental care is both non- higher altitudes (Hubbard 1978, Carey 1988). shareableand indispensable to female reproduc- Life history analysesthat entail characterizing tive success”(Wittenberger and Tilson 1980, p. a species’ “mating system” (seeEmlen and Oring 199). However, comparative data are needed for 1977, p. 222) also are confronted with difficult nests at which males are absent to substantiate problems, especially of definition (Emlen and this claim and this information is presently un- Oring 1977, Harrington et al. 1982, Dewsbury available. Furthermore, in this nonterritorial 1987). Gathering unequivocal evidence of mat- species,males defended accessto individual fe- ing patterns also is problematic (Kleiman 1977, males. “[DIefending exclusive accessto a single Wittenberger and Tilson 1980). For example, female” (Wittenberger and Tilson, 1980, p. 200) Dewsbury (1987) discussedthe diversity of def- is also seen in other monogamous finches (New- initions of monogamous mating systemsby using ton 1972, Ricklefs 1977, Wittenberger and Til- the following criteria: 1) exclusivity of mating; son 1980). 2) joint parental care; and 3) association.He con- The one instance of polygynous behavior ob- cluded that none of these three criteria alone is served during this study (Fee and Bekoff 1986) sufficient for inferring monogamy. However, a indicates that although grosbeaksare usually mo- combination of two or more of these criteria re- nogamous, there may be some flexibility in their sults in a more reliable assignment of monogamy mating habits. Other studies also indicate that as the observed mating system. passerinesare not locked into a rigid monoga- We infer that grosbeaksare typically monog- mous mating system.However, polygyny hasonly amous based on our observations ofjoint paren- been documented in approximately 36 tal care and patterns of association between an species(Ford 1983; see also Fitch and Shugart adult male and an adult female during the breed- 1984, Gowaty 1987). ing season.We assumedthat the male associated Our data also indicate that adult grosbeaks with the female during and after nest building show male-biased sexratios throughout breeding began was both the father of her offspring and and nonbreeding seasons(Bekoff and Scott 1989). her mate throughout the season, based on the However, Balph and Balph (1976) observed fe- following criteria: 1) constantattendance by males male-biased sex ratios in winter flocks observed 80 ABBY C. SCOTT AND MARC BEKOFF at feeding stations in Utah. In general, monog- Richmondenathrough Pipilo (part). U.S. Natl. amous birds have been characterized as showing Mus. Bull. 237. higher female than male parental investment BALPH,D. F., AND M. H. BALPH. 1979. Behavioral flexibility of Pine Siskinsin mixed foraginggroups. (sensu Trivers 1972), a conclusion that stems Condor81:211-212. from observations of adult sex ratios favoring BALPH,M. H. 1976. Some physical characteristicsof males (Trivers 1972, Breitwisch 1989). How- Evening Grosbeakswintering in northernUtah. ever, Breitwisch (1989) showed that skewed sex N. Am. Bird Bander1:11&115. BALPH, M. H. 1977. Sex differences in alarm re- ratios favoring adult males may not necessarily sponsesof winteringEvening Grosbeaks. Wilson be a result of higher female parental investment, Bull. 89:325-326. and also that skewed sex ratios biased toward BALPH.M. H.. AND A. M. L~NDAHL. 1978. Winter adult males do not invariably favor the evolution philopatry of EveningGrosbeaks in northern Utah. or adoption of monogamy in all environments. N. Ani. Bird Bander 3:149-151. BEKOFF,M., AND A. C. Scorr. 1989. Aggression, Further, Murray (1984) and Lamprecht (1987) dominance, and social organization in Evening suggestedthat resource quality and availability, Grosbeaks.Ethology 83: 177-l 94. and protection of nests, mates, and young, are BEKOFF,M.. A. C. Scorr. AND D. A. CONNER. 1987. probably more influential in shaping mating sys- Nonrandom nest-site selection in Evening Gros- beaks. Condor 89:8 19-829. tems than are biased sex ratios. BEKOFF,M., A. C. Scorr, AND D. A. CONNER. 1989. In summary, our analyses show that although Ecologicalanalyses of nesting successin Evening adult males and females appear to play different Grosbeaks. Oecologia 81:67-74. reproductive roles during various parts of the Bws, J. R., AND G. H. PARKS. 1964. Interaction of breeding season,they both contribute to building Evening Grosbeak ( vespertina) and spruce budworm (Choristoneura fumiferana nests and care-giving throughout the breeding [Clem.]) in a localized outbreak treated with DDT season.Grosbeak parents, who are typically mo- in Quebec. Can. J. Zool. 42:1017-1024. nogamously bonded, continually divided the BREIIW~SCH,R. 1989. Mortality patterns, sex ratios, work of nest preparation and raising their young. and parentalinvestment in monogamousbirds, p. l-50. In D. M. Power[ed.], Currentornithology For example, we found that males typically fed Vol. 6. Plenum Press,New York. and protectedtheir mate and their young, where- BYERS,J. A., AND M. BEKOFF. 1990. Inference in as females primarily built the nest, incubated, socialevolution theory: a casestudy, p. 84-97. In brooded, and also fed their young. M. Bekoffand D. Jamieson[eds.], Explanation andinterpretation in thestudv of animalbehavior. Vol. II, Explanation, evolution, and adaptation: ACKNOWLEDGMENTS Westview Press. Boulder. CO. Financial support was provided by The Whitehall CAREY,C. 1988. Avian reproductionin cold climates, Foundation, The National Geographic Society, The p. 2708-27 15. In H. Ouellet [ed.], Acta XIX con- American PhilosophicalSociety, and a grant-in-aidfrom gressusintematinalis ornithologici. Univ. of Ot- The University ofColorado. Denise Arthur helpedwith tawa Press, . data collection and interpretation. Barb Fee, David CASH,K. J., ANDL. S. JOHNSON. 1990. Male parental Manry, Dan Blumstein, and Bay Roberts helped with investment and female competence in Yellow- data collection. Martha H. Balph, Louis Oring, Joseph headed Blackbirds. Auk 107:205-206. Ortena. Cate Orteaa. Thomas Daniels. Allen Moore. DEWSBURY,D. A. 11987. The comparative psychol- Pete; Landres, Margaret Dussault, David Armstrong; ogy of monogamy. Nebr. Sympl Motiv. Pap. 35: Alexander Cruz, and especiallySallie J. Hejl provided l-50. very helpful comments on a previous draft. Rob and EMLEN,S. T., ANDL. W. ORING. 1977. Ecology, sex- Janice Scott contributed constant support during the ual selection,and the evolution of mating systems. entire study. The administrators of Rocky Mountain Science 197:215-223. National Park allowed us to tag nest trees.Private land- FEE, B. A., ANDM. BEKOFF. 1986. Polygyny in the owners, in particular Dr. Henry Toll, graciously al- Evening Grosbeak. Wilson Bull. 98:308. lowed us accessto their land in and around Eldora and FITCH, M. A., AND G. W. SHUGART. 1984. Require- Nederland. Special thanks to Lee and Virginia Evens, ments for a mixed reproductive strategyin avian owners of Arapaho Ranch Wildlife Refuge and Bird species.Am. Nat. 124:116-126. Sanctuary,for allowing us to study grosbeakson their FORD,N. L. 1983. Variation in mate fidelity in mo- preserve. nogamous birds, p. 329-356. In R. F. Johnston [ed.], Current ornithology, Vol. 1. Plenum Press, LITERATURE CITED New York. GORI,D. F. 1990. Responseto Cashand Johnson. AUSTIN, E. O., JR. [ED.]. 1968. Life historiesof North Auk 107:206-208. Americancardinals, grosbeaks, buntings, towhees, GOWATY, P. A. 1983. Male parental care and appar- finches, sparrows, and their allies. Order Passeri- ent monogamy among Eastern Bluebirds (Siuliu formes: Family Fringillidae, Part one, Genera sialis). Am. Nat. 121:149-157. BREEDING BEHAVIOR OF EVENING GROSBEAKS 8 1

GOWAT~,P. A., AND J. H. PLISSNER.1987. Associ- tive strategiesin birds: reproductive effort. Am. ation of male and female Robins during the breed- Nat. 111:453478. ing season.Wilson Bull. 9956-62. RITCHISON, G. K. 1980. Singingbehavior of the Black- HARRINGTON,F. H., P. C. PAQUET,J. RYON,AND J. C. headedGrosbeak, PheucticusMelanocephalus, with FENTRESS.1982. Monogamy in wolves: A review emphasis on the function of singing by females. of the evidence, p. 209-237. In F. H. Harrington Ph.D.diss., Utah State Univ., Logan. and P. C. Paquet [eds.], Wolves of the world: per- SHAUB,M. S. 1963. Evening Grosbeak winter incur- spectivesof behavior, ecology, and conservation. sions-1958-59, 1959-60, 1960-6 1. Bird-Banding Noyes Publishing, NJ. 34:1-21. HEJL,S. J., J. VERNER,AND G. W. BELL. 1990. Se- SPEIRS,D. H. 1968. Hesperiphona vespertina brooksi quential versus initial observations in studies of Grinnell, Western Evening Grosbeak,p. 237-25 1. avian foraging.Studies in Avian Biol. No. 13:166- In 0. L. Austin, Jr. ted.], Life histories of North 173. American cardinals,grosbeaks, buntings, towhees, HUBBARD,J. D. 1978. Breeding biology and repro- finches. snarrows. and their allies. Order Passeri- ductive energetics of Mt. White Crowned Spar- formes: Family Fringillidae, Part one, Genera rows in Colorado. Ph.D.diss., Univ. of Colorado, Richmondena through Pipilo (part). U.S. Nat. Boulder. Mus. Bull. 237. KACHIGAN,S. K. 1986. Statistical analyses.Radius SULLIVAN,K. 1990. Daily time allocationamong adult Press, New York. and immature Yellow-eyes Juncosover the breed- KLEIMAN, D. G. 1977. Monogamy in mammals. ing season.Anim. Behav. 39:380-387. 0. Rev. Biol. 52:36-69. TAKEKAWA,J. Y., E. 0. GARTEN,AND L. A. LANGELIER. LACK; D. 1968. Ecological adaptations of breeding 1982. Biological control of forest insect out- in birds. Chapman and Hall, London. breaks:the use of avian predators.Trans. 47th N. LAMPRECHT,J. 1987. Female reproductive strategies Am. Wild. and Nat. Res. Conf. p. 393409. in Bar-headedGeese (Anser in&us): why are geese Tatvuas, R. L. 1972. Parental investment and sexual monogamous?Behav. Ecol. Sociobiol.21:297-305. selection,p. 136-l 79. In B. Campbell [ed.], Sexual LANGELIER.L. A. 1983. Habitat selectionofcommon selectionand the descentof man, 187l-l 97 1. Al- breeding bird speciesin western sprucebudworm dine, Chicago. outbreak areas.M.Sc.thesis, Univ. of Idaho, Mos- VAN TUNE,J., AND A. J. BERGER. 1971. Fundamen- cow. tals of ornithology.Dover Publications,New York. MORTON,M. L. 1976. Adaptive strategiesof Zono- VERNER,J., ANDM. F. WILLSON. 1966. The influence trichia breeding at high latitude or high altitude. of habitats on the mating systemsof North Amer- Proc. Int. Omithol. Congr. 16:323-336. ican passerinebirds. Ecology 47: 143-147. MURRAY, B. G. 1984. A demographic theory on the VERNER,J., AND M. F. WILLSON. 1969. Mating sys- evolution of mating systems as exemplified by tems, sexual dimorphism, and the role of male birds. Evol. Biol. 18:71-140. North American passerinebirds in the nestingcy- NEWTON,I. 1972. Finches. Collins, London. cle. Omithol. Monogr. 9: l-76. ORING, L. W. 1982. Avian mating systems,p. l-79. WELT, J. C. 1982. The life ofbirds. SaundersCollege In Famer, D. S., J. R. King, and C. P. Parkes[eds.], Publishing, New York. Avian biology,Vol. VI. AcademicPress, New York. WICKLER.W. 1968. Mimicrv in nlants and animals. PARTRIDGE,L., AND P. H. HARVEY. 1988. The eco- McGraw-Hill, New York: _ logical context of life history evolution. Science WITTENBERGER,J. F., AND R. L. TILSON. 1980. The 241:1449-1455. evolution of monogamy: hypothesesand predic- RICKLE~~,R. E. 1977. On the evolution of reproduc- tions. Annu. Rev. Ecol. Syst. 11:197-232.