Breeding of the Golden-Crowned Kinglet in Northern Minnesota

J. FieldOrnithol., 56(1):28-40

BREEDING OF THE GOLDEN-CROWNED KINGLET IN NORTHERN MINNESOTA

BY BOB AND CARLYN B. GALAT1 Duringthe summerof 1954and 1957,and during spring and summer of 1955, 1956, and 1960, we studiedthe breedingbiology of the Golden- crownedKinglet (Regulus satrapa). This isapparently the firstlife history of this speciesever completed.About 3788 h were spentin observation at the Universityof Minnesota'sBiological Station at ItascaState Park, Minnesota. Two pairswere intensivelystudied from the beginningof their first nestingcycle through the end of their secondcycle (1955, 1956), two (1954, 1957) from incubationthrough the nestlingperiod, and one (1960) through its first nestingcycle and part of the second.Spot ob- servationswere conducteddaily at 12 other nests.

METHODS AND STUDY AREA. One of us generallyobserved activities at the nestfrom a blind, the other observedfrom the ground. We were in direct communication (talking or shoutingdistance) at each nest site during the period when kingletsstarted a secondnest and still had nestlingsin the first nest. Kingletsallowed us to part branchesand twigshiding the nestsfrom view to observethem from 30 cm away. All femalesat nestsunder intensivestudy allowed us to reachinto the nestand pickthem up. The female, and occasionallythe male, visitedus in the blind. When the adultsblocked view of the nestlingthey were feeding, we could push them asidewith a finger without undulydisturbing them. If they hap- penedto be at the nest site when we weighedand measuredeggs and nestlings,they flew into the blind, hoppedaround the scale,and even landed on us. Nestsunder intensivestudy were observedfrom blindson towersor tree platforms.Three nestswere lowered from their original heights becauseit wasnot feasibleto build towersor platformshigh enoughto observe them. (The distanceslowered were 12.2, 3.04 and 1.82 m.) These nestswere near the tree tops, so the tree wascut just below the nest site, and lowered to the desiredheight--as low at 2.7 m in one case. The first two nests were lowered at different increments over 3 days,but the third one waslowered only once.To retard dryingout of foliage,the stembase was put in a bucket containingsand and water. In no casedid nest-loweringdisrupt the nestingcycle. When the blind wason a neighboringtree, the tree top housingthe nest wastied to the blind and pulled to within 30 cm. At the end of each observationperiod, thesenests were untied. All eggs,adults, and nestlingsat nestsunder intensivestudy were marked with different colored acetate base paint. Adult femaleswere 28 Vol.56, No. • KingletBreeding Ecology [29 dabbedwith a paint brushwhile they incubatedor broodedthe young; malesas they fed matesor nestlings. Eggswere marked and weigheda few hoursafter theywere laid, and weighedagain before they hatchedto measureweight loss.Nestlings were marked in order to follow daily feather-tractdevelopment and weight gains. Red squirrels(Tamiasciurus hudsonicus) were killed near nestsunder intensivestudy. They were the only rodentsseen in the near vicinityof the nestingsites. Habitat.mAt Itasca,we foundGolden-crowned Kinglet nests in spruce- tamarackbogs (14 nests),and in mixed stands(5 nests)dominated by balsamfir (Abiesbalsamea), white spruce(Picea glauca), red pine (Pinus resinosa),with a few scatteredtamarack (Larix laricina), paper birch (Betulapapyrifera), aspen (Populus tremuloides), and jack pine (Pinusbank- siana).A total of 49 dayswas spent locating nests (55% of our efforts in spruce-tamarackbogs and the remainderin the fir-spruce-pinehabitat). Three of the intensivelystudied nests were in the park. Prevalentin the understorywere Bebb'swillow (Salix bebbiana);bog birch (Betula pumila)(var.); red osierdogwood (Comus stolonifera); and speckledalder (Alnusrugosa). The other three intensivelystudied nests were in the fir- spruce-pinehabitat. Ground cover includedblueberry (Vacciniuman- gustifolium);sourtop blueberry (V. myrtilloides);arrowwood (Viburnum rafinesquianum);serviceberry (Amelanchier humilus); spreading dog bane (Apocynumandroseamifolium); and the commonbracken fern (Pteridium aquilinumvar. latiusculum).Dominant shrubs were honeysuckle (Lonicera canadensis),and wild sasparilla(Aralia nudicaulis).Black spruce(Picea mariana)were dominantin the bog studyareas and rangedfrom 17- 20 m in height. Tamarackswere primarilyaround the edgesof the bog.

RESULTS Vocalizations.--Wefound 3 majorpatterns or typesof vocalizationsof the Golden-crownedKinglet during 5 seasonsof study: (1) Call notes A. Passive--1 to 5 "tse" noteson the samepitch. B. Active--same as "A" but louder, more rapid, and strident. (2) Simple song--Series of up to 14 notes, with the first 2-5 on the same pitch, and the remainder ascendingin less than half tone intervals;lasting up to 2 s. (3) Complex song A. Included the simplesong for the first part and a musicalwarble which dropped an octaveor more in pitch for the secondpart; lastedup to 3 s. B. Sameas "A," exceptthe lastpart of the warblebecame chattery and harsh. C. Like "A," but much more melodious. 30] B. andC. B. Galati j. FieldOrnithol. Winter 1985 Passivecall noteswere used by the matingpair in (1) communicating with one another,(2) alertingthe nestlingsof their presencewhen the youngfailed to gapefor food, (3) approachingthe nestwhen feeding, and (4) locatingand approachingfledglings for feedings.Active call noteswere used when either adultwas alarmed or distressed.They were usedin the presenceof avian and mammalianpredators or cowbirds. The simplesong was used when a malekinglet proclaimed his territory directlyto another male in a borderingterritory. The female usedthe songto answeror call the male while she wason the nest or while she approachedthe nest from a period off the nest. Complexsongs were usedprimarily for proclaimingterritory, but were alsoused when red squirrelsand Blue Jays(Cyanocitta cristata) were in the vicinity of the nestor near fledglings.Both sexesused the songduring nestconstruc- tion, especiallywhen one bird wasworking materialsinto place,while the other wasapproaching the nest,or while it wasawaiting its turn to get on the nest with nestingmaterial. Complexsong was often interspersedwith simplesong. Singingwas fairly constantduring the nestlingperiod and after the young fledged (first nesting).During late July and early August, less territorial singingtook place, and the singingthat was done seemed primarilyfor locatingand feedingfledglings. In comparingthe amountof vocalizingdone during different periods of the nestingcycle, the leastamount took placeduring egg-laying; the mostduring incubation. Territoriality.--Territorydimensions were determinedby following the singingmale and mappinghis songperches. Lines connecting the outermostpoints were measuredwith a steel tape. Eleven territories were measuredduring our study.These measurementswere then plot- ted on a map, and the enclosedarea measuredby dividingthe areas into trianglesand calculatingtheir sizes. Territories ranged in size from 0.9-2.5 ha, averaging1.6 ha. The pair that maintainedthe largestterritory (2.5 ha) had no competing kingletsbordering its territory. Defenseof the territory againstinvaders of its own specieswas per- formed primarily by the male, mainly through song.Physical combat betweenneighboring territorial maleswas observed only once. Interspecificaggressive behavior was noted with the Blackburnian Warbler (Dendroicafusca),Black-throated Green Warbler (Dendroicavi- rens),and Chipping Sparrow (Spizellaarborea). These birds had nests within the territoriesof kinglets.Black-capped Chickadees (Parus atrio capillus)and BorealChickadees (P. hudsonicus)were chased by both male and femalekinglets when they landedon the kinglet'snesting tree. Nestbuilding.--At Itasca,Golden-crowned Kinglets regularly built 2 nestsand raised2 broodsduring a breedingseason (we observed15 nestingpairs). The Goldcrest(Regulus regulus) of northernEurope, con- sideredconspecific with the Golden-crownedKinglet, alsohas 2 broods in Fennoscandia(Palmgren 1932) and Norway (Haftron 1978a). Vol.56, No. I KingletBreeding Ecology [31

The averageheight of 19 nestswe found in Itasca State Park was 15.3 m (range 8.2-19.6 m). Four were in white spruce,3 in balsamfir, and 12 in black spruce(Picea mariana). All nestsin black spruceand balsamfir were in the upper tree crowns,a few centimetersfrom the trunk. All but one of thesenests rested on radiatingtwigs supported by brancheswhich had been incorporatedinto the wallsnear the nestrim. The exceptionhad no basalsupport; it wassupported by twigson the rim. Three nestsin whitespruce were suspended by the rimson radiating twigs,with no basalsupport. If kingletsdisplayed a preferencetoward a compasspoint whenbuild- ing nests,it wasto the eastor southeast(11 nestseast; 6 southeast;1 southwest;1 northwest).Prevailing winds at Itascaare westerly.All east and southeastnests on the leeward side were high in tree crowns.The neston the exposednorthwest side was lowest, and wasonly 5 m east of a windbreak of tall balsam fir. All nestswere so well protected by overhangingfoliage that they weatheredwind and rain andwere never reached by sun.Haftron (1978b) found that during periodsof heavyand/or long-lastingrain, the Gold- crest'snest cup sometimesgets wet. This wasnot so with kinglet nests. Even with torrential rains accompaniedby strong winds,the nest re- maineddry. By nestinga few centimetersfrom the trunk in the black spruceand balsamfir, kinglet nestswere free from any rain that might trickledown the trunk. The foliagewas generally so dense around nests that they could not be seenfrom aboveor from nest level. We always had to part someof the twigsto observenest activity. The 3 nestsin white sprucetrees were on the undersideof longbranches near the tips. They, too, were completelyhidden from view from the top or at nest level, but couldbe partially seenfrom below. The earliestnest building we observedwas 18 May 1956. RobieTufts found 2 nestsjust startedin Nova Scotia, 10 April 1921, and Cordelia Stanwoodfound a half-finishednest near Ellsworth, Maine, 25 April 1912 (Bent 1949). We collected19 nestsand found them to be madeprimarily of com- mon locallyavailable materials. Most nestscontained lichens (Parmelia sulcata,P. phypodes,and Usneasp.) scattered through the inner, bottom, and outer walls. The interior bases of the nests were lined with fine stripsof paper birch bark, moss,lichen, deer hair, and feathers.The outerwalls consisted chiefly of mosses(Brachythecium solebrosum and Hyp- numsp.), spider web, cottonfrom cottongrass, parts of insectcocoons, and someunidentifiable brown-colored stringy-stretchable material. Kinglet nestsare deep cup-shaped,with the rims slightlyarched in- ward. The nestingcavity is almostcircular. Table 1 showsthe mean nestdimensions for 5 nestsbefore the eggshatched and after the nestlingsfledged. Five nestsaveraged 5 daysto complete(range 4-6 days). Contraryto Haftron (1978c),we observedcourtship feeding on sev- eral occasionsduring nest construction. The maleapproached with food while the female wason the partially-completednest; she responded by 32] B. andC. B. Galati J.Field Ornithol. Winter 1985

TABLE1. Mean dimensionsfor five Golden-crownedKinglet nestsshortly before the eggshatched and after the nestlingsfledged (cm).

Before After

Inside diameter 4.1 õ.5 Outside diameter 7.5 9.7 Insidedepth 4.0 3.2 Outsidedepth 7.7 5.1 Wall thickness 1.95 1.3

fluttering her wingsand beggingfor food. She alsopursued him in the neighboringtrees around the nest,fluttering and gapingfor food, while he gatheredit. He would then turn and feed her. Coition took place throughout nest building and the period of egg laying. The female instigatedit by flattening out in the partially con- structed nest or on twigs close to the nest while uttering call notes interspersedwith the complexsong. Her mate respondedby singinga seriesof complexsongs before mounting. Eggs.mSix eggsin one of our nestsaveraged .59 g the day before they hatched(range .48-.66). Six eggsin anotherclutch, weighed im- mediatelyafter the lastegg waslaid, averaged.77 g (range.59-.82 g). Five of theseeggs averaged .63 g (range.25-.78) 14 dayslater. There wasan averageweight loss of .077 g. (The sixthegg did not hatchand weighed.27 g.) Haftron (1978d)weighed 13 fresheggs of the Goldcrest and found a mean weight of .79 g. Eight clutches(which were known to be complete)contained 9 eggs in the first nest.Of 6 completesecond clutches, 4 had 9 eggsand 2 had 8 eggs.Haftron (1978c) concluded that 9.7 eggsis probablyfairly close to the meanclutch size for the Goldcrestin northern Europeas a whole. This compareswith 8.85 eggs(per 1st and 2nd clutch)for 14 nestsof the Golden-crownedKinglet at ItascaState Park. Initiationdates for first clutchesranged from 12 May 1955 to 27 May 1960. The earliestdate for the start of the secondclutch (amongthose that were successfulwith their first clutch) was 22 June 1960; latest 1 July 1956. Kingletslay eggsearly in the morning(between midnight and 0400 in 2 cases)on successivedays until the clutch is complete. Haftron (1978c) found Goldcresteggs were laid early in the morning on consecutivedays. He learnedthe exacttime of layingof 3 eggson 1-3 July 1971 was 0249, 0246, and 0258. Incubation.BIncubation wasperformed by the female only. All her nightswere spent on the nestafter the firstegg was laid. Haftron (1978b) also found this true for the Goldcrest. Through incubation,one female kinglet spentan averageof 71% of her time eachday on her first nest(nocturnal incubation not included). Shebegan spending time on her eggsin the secondnest after the seventh Vol.50, •qo. I KingletBreeding Ecology [3 3 egg was laid, and thereafter was on that nest 82% of the time. The averagelengths of time per attentiveperiod were 6.5 (range4-38 min) and 6.3 min (range 2-22 min) respectively. Weather conditions influenced attentivenessduring incubation. Greatestattentiveness (81%) occurredwhen it had rained throughout the observationperiod (temperature 18øC).Least attentiveness(47%) occurredduring the hottestday (temperature35øC). In 1955 a female beganspending time on the nest after her second egg of a 9-egg clutch waslaid. Five of the eggshatched 19 dayslater; the remainderhatched over the following3 days.A first-nestingfemale startedincubating after her lastegg (ninth) waslaid, and the first 3 eggs hatched15 dayslater; the remainderhatched over the next 2 days.A third female, secondnesting, began incubatingafter the seventhof 8 eggswas laid. No attentivenesswas observedduring either nestingcycle in 1956 until the last egg wasdeposited. At the first nestthe first egg hatched 15 June, and the last 16 June, givingan incubationperiod of from 14- 15 days.Using Nice's (1954) formula--the time from laying the last egg to the hatchingof the lastyoung--the incubationperiod would be 15 days.In her secondnest, this same female kinglet was attentive 70% of the time after her seventhegg waslaid 5 July. Her first egg hatched 20 July and the lastone 21 July. Again, the incubationperiod would be 15 days.Haftron (1978a) found the Goldcrest'sincubation period to be approximately16 daysfor 3 clutchesof eggs,and Palmgren(1959) states it is 15 days. Femalekinglets are very restlessduring incubation;changing direc- tionsevery few secondsby turning from quarter to full turns or more. However, they will remain still for severalminutes and even doze off to sleep. During the first nestingof a pair, the female came in with a load of insectsto "feed" her eggson the 12th day of incubation.She repeated that type of behavior on the 12th and 13th day during secondnest incubation.She shookthe nestby alternatelystamping her feet on the rim of the nest,as is typicalof birdsat nestswith young.She gave rapid call notes;finally she ate the food. Haftron (1978a) also observeda femaleGoldcrest, who was incubating eggs in onenest and intermittently feedingyoung in another, trying to get her eggsto respondto food. Hatching.--Hatchingdates ranged from 4Juneto 20July.Eggs hatched at any time during day or night. In one nest 7 eggshatched at 0805, 0837, 0939, 1250, 1355, 1505, and 1529. The eighthegg hatched the followingday at 1530. In anothernest all eggshatched during the night and before 0700 on two consecutivedays. Egg shellswere either eaten or removedby both parents.Unhatched eggswere never removedfrom the nest as far as we know. Eggsdis- appeared,but whetherthey hatched and the nestlingdied (and was then removed) was not known. Neststretching.--Nest stretching took place throughout incubation and 34] B. andC. B. Galati j. FieldOrnithol. Winter 1985

U3100, • --- FIRSTFEMALE • 90 SECONDFEMALE • 80 • 70 z A B 60 5ot;!, ß .-=,="'ß •ß.I, •, .•. /ßt •ß ,I, ß ..-_ ..z';-- .\- ,, ; _, .e"' ,e-- \ •.e•, ß •. • / \/•'• / • / ß 50 z , ./ ß \,•// • .-'-"'r/ n. 20 ß '\ /\,,\ / 'e....e...... J¾ e...... e/ t i i i i i ß i .i ß ,._,, i ß ß ,ß ß ß :/,i i

DAYS AFTER HATCHING F•cul• 1. Comparativefeeding by two first-nestingGolden-crowned Kinglet females. (A) indicatesstart of buildingof the secondnest by both females.(B) indicates1 st egg laid in second nest of both females. the nestlingperiod and was done entirely by the female. One female (1956) stretched her nest beyond its limits. By the 24th day a hole appearedin the bottom of the nest, large enoughfor nestlingsto fall through(oldest nestlings were 9 daysold). The nestwas 17.3 m from the ground, hanginglike a pendulumfrom the undersideof a black sprucethat had no basalsupport. Three nestlingsfell throughthe hole; 2 landed on limbs below the nest and the third on the ground. We patchedthe nest and returned the 3 that had fallen out. All nestlings survivedand eventuallyfledged. Feedingof nestlings.--Bothparents fed nestlings,with the male doing the greater shareearly in the first nestingcycle, while the female spent part of the time brooding(male 62.5 and female 37.5%). When the nestlingswere no longer brooded, her feeding visits increased.The femaleagain fell behindwhen she started building her secondnest (Fig. 1). Haftron (1978a) found that the male Goldcrestbrought far more food to the young throughoutthe nestingcycle than the female. We observedthis to be true of male kinglets. Spiderswere not acceptedfor food by the young.They were partially swallowed;then regurgitated.When this occurred,the parentstried to feed the spidersto other nestlings,who alsorejected them. The largestnumber of feedingvisits made by a femaleduring a 17-h period was329 when her oldestnestlings were 14 days.She was taking care of the young by herself. The smallestnumber waszero while she built a second nest. Two pairsstarted building a secondnest 8 d after eggsbegan hatching Vol.56, I•o. 1 KingletBreeding Ecology [3 5 in the first nest, and laid eggs6 d from the start of secondnest construction(second nests were completedin 5 d; malesdid little work). The first femalegreatly reduced her feedingvisits 2 d after shebegan her secondnest and againwhen she laid the first egg(Fig. 1). The second femalegreatly reduced her feedingvisits on the third, fourth (no visits), and fifth daysafter beginningsecond nest construction, and againwhen her first egg waslaid (Fig. 1). Sinceboth femalesleft mostof the feedingduties to their matesafter they beganbuilding their secondnests, overall low feedingpercentages were to be expected. Ambiguousbehavior while attending temporally overlapping nests.--After one female kinglet beganconstruction on her secondnest, she showed ambiguousbehavior between working on her new nestand feedingthe youngin the old nest.This wasfirst observedon the fifth day of con- structionwhen she was lining the newnest. On 3 visitsshe fed the young nestingmaterials (deer hair, moss,lichen). The followingday after she had depositedone eggin the newnest, she made 16 visitsto her nestlings at the firstnest, and on 4 of thoseshe fed them nestingmaterials (feather, moss,lichens). She alsoremoved deer hair, down, and mossfrom around the nestlingsand stuffedit into the mouthsof 2 nestlingswho swallowed it. This type of behaviorcontinued for 2 more days.Twice before her second nest was lined, she arrived with her bill loaded with insects and incorporatedthem into the nest wall. Food.--Duringthe nestingcycle, Golden-crowned Kinglets gathered food by foraging almostexclusively among the spruce,fir, and pine branches.We removed samplesof insectsfrom the bills of 2 pairs of adult kingletsbefore they fed nestlings.Alvah Petersonexamined the collected matter (Table 2). The food collectedfrom the kingletsduring the first nestingdiffered decidedlyfrom that of the secondbrood. The kingletsmust have col- lectedwhat wasreadily available. When one femaleshared feeding duties with her mate, shebrought in tiny insects(winged and wingless),larvae, and somemoths. After he disappeared,she brought primarily large Phalangida(daddy longlegs) and mothsuntil the youngfledged 5 dayslater, food itemsthat, appar- ently,were most abundant and readily available in the immediatenesting area,but whichprobably had little foodvalue since only 3 of 9 fledglings survived (Table 3). Whichnestling gets fed first. mBoth parentstended to feednestlings that were nearestto them. Of 294 feedingvisits made by both parentsto one nest over a period of 16 h, the nearestnestling was fed first 123 times; the farthest, 56 times; middle ones, 46; and others, 69 times. In one nest after the male disappearedand the female assumedthe exclusiverole of feeding9 nestlings,the "runt-of-the-litter"was crowd- ed over to the oppositeside of the nestfrom the approachingfemale and wasseldom fed. Once it receivedno food for over a 3•/2h period. It starved2 daysafter the male disappeared,and waspushed to the nest B. and C. B. Galati J. FieldOrnithol. Winter 1985

T^•L• 2. Examplesof foodcollected by twopairs of Golden-crownedKinglets.

Number Food Larvae Pupae Nymphs Adults Insects Aphididae 9 12 Caddisfly 2 Chrysopidae 2 Coccinellidae 6 1 Corrodentia 3 2 Culicidae 1 Geometridae 11 Hemiptera 2 Lepidoptera 1 Tipulidae 1 Lycaenidae 1 Membracidae 1 Midge (Chironomus) 1 Miridae 1 Pamphiliidae 1 Phalaenidae 10 Psocid 5 Pupa (unknown) 1 Syrphidae 15 Trichoptera Spiders(unknown) Arachnida(long-legged) (Tetraffnathaelongata) bottomand trampledon by the other nestlings.When the male wasstill helpingto feed nestlings,the femalewas observed digging the "runt" out from the bottomof the nesttwice when it was2 daysold; twiceat 5 days;once when it was8 days.

TABLE3. Numberof Phalangidsand insectsfed nestlingsby a femaleGolden-crowned Kinglet(male feeding included in parenthesesfor 18July only).

Numberof % of Phalan- gidsand moths Date Time Phalangids Moths Other insectsin total feedings 18July' 0430-2030 1(0) 18(12) 269(240) 6.5(4.8) 19July 0700-2000 121 55 134 56.7 20July 0700-1600 191 36 136 62.5 21 July 0400-2100 271 58 265 55.4 22July 0400-1200 141 22 143 53.3 23Julyb 0700-1130 58 11 116 37.3

Last day both parentsfed nestlings. Nestlingsfledged. Vol.r•6.1•o. 1 KingletBreeding Ecology [3 7

Nestlingdevelopment.--Kinglets are completelynaked when they hatch, exceptfor tufts of down abovethe eyesand on the crown.They are about 2.6 cm long and have flesh-coloredskin. When nestlingsare two daysold, they emit their first sound("tseek"), which is barely audible.At four daysthere is a trace of eye-opening. Sheathsbegin to form on the sixthday. The firstsigns of feathersappear on the primaries,secondaries, and rectriceson the ninth day. Eyesare completelyopen on the twelfth day. They begin crawlingin and out of the nest on the sixteenthday. Nestlingsweighed in 1956 gained from 5-7 times their hatching weightby the seventhday after hatching.The first nestlingmore than doubledits weight by the secondday after hatching.The lightestnestling on the first day of hatching(.73 g), surpassedthe weightof the heaviest one (1.03 g) on the seventhday after hatching. Fledglingbehavior.--About 2 daysbefore the nestlingsfledged, they crawled to and from the neighboringtwigs and branches30-60 cm away. After the first young fledged, within a matter of minutes,the remaining nestlingsleft the nest. In 6 nestswhere nestlingswere observedto fledge,the shortesttime for the nest to empty was 15 min; the longest,105 min. All young fledgedbefore noon (the earliestat 0640, the latest,1157). One nestling flew from a nest at 1639, just as the branch was moved to exposethe nestfor observation.However, the followingmorning, this nestling was back in the nest. The majority of Golden-crownedKinglets left the nestbetween 16- 19 daysafter hatching.Of 46 nestlingswhose ages were known, 2 fledged at 14days,2at 15,12at 16,2at 17,8at 18, and20at 19days.Atone nest where nestlingsfledged at 15-16 daysold, they had been taken from the nestdaily to checkweight and feather tract developmentuntil theywere 9 daysold, and at 14 daysfor experiments;disturbances which may havetriggered an earlier than normaldeparture. At anothernest where nestlingsfledged between 18-19 days,they were not handledas much nor as late in the nestlingperiod. Haftron (1978a) found the nestlingperiod for a pair of Goldcrestswho double-broodedto be 17- 22 days. When the youngleft the nest,they flew horizontallyto a tree 3-4 m from the nest, where they "tsipped" loudly and incessantly.Until all young had left the nest, both parentssplit feeding dutiesbetween the fledglingsand the youngstill remainingin the nest,but tendedto favor the nestlingswith more feedingvisits. Usually, within 1-2 h, the fledg- lingsfound one anotherby "tsipping,"and they then gatheredon the samelimb in a tight row, spendingmost of their time preeningbetween feedingvisits by their parents.They were not observedpreening one another. Bothadults fed fledglingsthe first dayout of the nest,but thereafter, with the first brood, feeding wasdone almostexclusively by the male becausethe female wasincubating on the secondnest. Haftron (1978a) 38] B. andC. B. Galati J. Field Ornithol. Winter 1985 foundthe sameto be true with the Goldcrest.Kinglet eggs in the second nest hatched about the time the fledglingsfrom the first nest became independent,enabling both parentsto care for the newlyhatched nestlingsin the secondnest. Twice for Goldcrestin Norway(Haftron 1978a) and oncein Finland(Palmgren 1932), the first eggwas laid in the second nest 3, 10, and 12 daysrespectively, before the first brood had left the nest. Goldcrest parents had to split their duties between the care of nestlingsand fledglings. Kingletfledglings remained close to eachother for the first 3-4 days after fledging.One family remainedwithin 120 m of the old nesting sitefor the first 11 daysout of the nest.They couldeasily be located within their territory, becausewhile they perched on the samelimb (about 10 m or more from the ground),they constantly"tsipped" while waiting to be fed. Periodically,the male left his fledglingsand fed his mate while she incubated at the secondnest. She seldomvisited the fledglings,but foragedfor food nearby.After the nestlingsfledged in the secondnest, both parentsshared almost equally in the feedingduties. On the fifth day out of the nest,fledglings made their first attempts to get their own food by peckingin the foliage. (It took longer when nestlingsfledged prematurely.) There wasan increasein calling,not only among the fledglings,but alsobetween them and the adults.The "tsipping"of the fledglingswas almost like an adult "tse," but not quite so clear. The first signsof warbling took place, but the "tsees" and warble were run together. Betweenthe eighth and eleventhday out of the nest,the amount of foraging done on their own increased;they spent more time accom- panyingthe adult while it foraged for their food. Fledglings,between 12-16 daysout of the nest,gathered a gooddeal of food for themselves,but still energeticallypursued the adults for food. Their tails were sufficientlydeveloped so that they were almost ascapable of hoveringin mid-airand maneuveringin flight asthe adults. No feeding of fledglingsby adultswas observedafter the young had been out of the nest 16-17 days. Territorial singinglargely ceased and territory boundariesfaded after the nestlingsin the secondnest fledged. Predation.--Predatorsmost likely to find kinglet nestsand destroy eggsor young were red squirrels,Blue Jays,and Gray Jays(Perisoreus canadensis).One raptor frequentlyseen in kingletterritories that might have been a threat to adult kingletswas the Sharp-shinnedHawk (Ac- cipiterstriatus). One pair lost its nestto a predator after we had loweredit from 17.8 m to 2.4 m abovethe ground. Two red squirrelswere feedingon cones 6 m awaywhen we approachedthe nestto start our daily observations. Another pair whosenest was lowered from its originalheight almost lost their secondbrood of 9 nestlingsto red squirrelson severalocca- sions.We shot4 squirrelsjust asthey were about to enter the nest.Two vol.56, •o. 1 KingletBreeding Ecology [39 other squirrelswere killed when they leapedfrom the nestingtree to a neighboringone. In anothernest, Gray Jays were the prime suspectsin killing 5 of 7 nestlings(the oldest being 16 days).As we were leaving our tree platform for the day, a flock of 7 GrayJays invaded the area near the nestand wereattacked by the adultkinglets. When we returnedto the blind the next day, a dead, partiallyeaten young was in the nest.On a limb 3 m belowthe nestwere two more partially eaten nestlings.A fourth dead nestlinghad been forced into a crack in the tree trunk .6 m below the nest. A fifth dead, partially eaten nestlingwas discovered in the moss 17.3 m belowthe nest.The lasttwo escapedand were perchedabout 13 m up in a spruce,where the parentswere feeding them. Nest desertion.--Of the three desertions out of the 19 nests under study,all were apparentlydue to our activities.Two were during nest constructionand one after the seventhegg waslaid. Nestsuccess.--From 12 kinglet nestswhich containeda total of 102 eggs,76 eggs(74%) hatched. Of 76 nestlingsin 10 nests,13 weretaken by predatorsand 1 starved--a nestlingsuccess of 83%. Of 62 nestlings that fledged,44 from 6 different familieswere followed;38 reachedthe age of independence--a fledglingsuccess of 86%. The mostfrequent causes of reproductivefailure amongkinglets were predation,starvation, and failure of eggsto hatch. Starvationappeared to be the causefor the deathsof 6 nestlingsin one nestwhere the female had to assumethe feeding dutiesalone. Predatorsraided two nestsand ate 13 of the 16 nestlingsin them. Femalesof 2 different nestsdisap- pearedduring the incubationperiod, leaving2 clutchesof 9 eggseach. Eight eggsin 6 different nestsfailed to hatch. One eggthat crackedcontained a deadyoung which was not removed duringour observationperiod, but the followingday it wasgone. King- lets removeddead nestlings,but not unhatchedeggs. However, when a 13-day-oldnestling starved in another nest, the female made no at- tempt to remove it (the male had disappeared).The remaining 8 nest- lingshad outgrownthe nest and spentmost of the time in 2 or 3 tiers, completelycovering the dead nestling. Five youngfrom one familydied on the first day they fledged.They couldnot fly well enoughto remainin the treesand were ignoredby the adult female when they becamegrounded. (The adult male had disappearedduring the latter part of the nestlingperiod.) Starvation was believed to be the cause of their deaths. Haftron (1978d) found that of 25 NorwegianGoldcrest nests, all containingeggs or young, 9 were robbed by predatorsor abandoned for unknown reasons. He did not record the clutch size nor the exact lossesof eggsand young.By assuminga mean clutchsize of 9.8 eggs (basedon 13 Norwegianclutches known to be complete)and supposing that even from successfulnests an average2 eggsor youngwere lost, Haftron cameup with a 50% fledgingsuccess. We suspectthat red squirrels,in Itasca, are the main predatorsof 40] B. andC. B. Galati J.Field Ornithol. Winter 1985 Golden-crownedKinglets and we strictly controlled the numbersof these rodents.We also savednestlings that had fallen through nest bottoms.The successof reproductionfor the Golden-crownedKinglet is thus not likely as high as we observedcompared to strictlynatural conditions.

SUMMARY We studiedGolden-crowned Kinglets at ItascaState Park, Minnesota, in the summerof 1954 and 1957, and during the springand summer months of 1955, 1956, and 1960. Nineteen nests were discovered, 18 of which containedeggs or young. It took about 5 d to build a nest. Nine eggswere laid in the firstnests and 8 or 9 in the second.Incubation took approximately15 d. Food consistedprimarily of insectsin various stagesof development.We observed44 of 62 nestlingsthat fledgedand found an 86% success rate.

ACKNOWLEDGMENTS We expressappreciation to William Marshalland the late T. Shantz- Hansen, who made facilitiesat the University of Minnesota Forestry and BiologicalStation availableto us, and to Walter Nelso.n, Mainte- nanceForeman, who directedconstruction of 4 towers,and suppliedus with tools, equipment, and supplieswhich enabled us to reach other nests.We thankJoe J. Hickey, under whosedirection and guidancewe initiatedour study,and S. CharlesKendeigh and John T. Emlen,Jr., who gaveus further directionand guidance.We alsogive sincere thanks to theseornithologists for critical advice.We thank the late Alvah Pe- terson for identifyinginsects which were collectedfrom the Golden- crowned Kinglets.Finally, we are grateful to William E. Dunton for typing and retypingthis manuscriptand for his constructivecriticism.

LITERATURE CITED BENT,A. C. 1949. Life historiesof North Americanthrushes, kinglets, and their allies. U.S. Natl. Mus. Bull. 196:382-399. H^FTRON,S. 1978a. Cooperationbetween the maleand femaleGoldcrest Regulus regulus whenrearing overlappingdouble broods. Ornis Scand.9.' 124-129. ß 1978b. Egg-layingand regulationof egg temperatureduring incubation in the GoldcrestRegulus regulus. Ornis Scand.9:2-21. . 1978c. Energeticsof incubationby the GoldcrestRegulus regulus in relationto ambient air temperaturesand the geographicaldistribution of the species.Ornis Scand. 9:22-30. ß 1978d. Weightincrease and feather development in the GoldcrestRegulus regulus. Ornis Scand. 9:117-123. NICE,M. M. 1954. Problemsof incubationperiod in North American birds. Condor 56: 173-197. P^LMCREN,P. 1932. Zue Biologievon Regulusr. regulus(1) und Parusatricapillus borealis Scly.Acta Zool. Fenn. 14:1-113. ß 1959. Fuglekonger.Pp. 215-218, in Nordensfugle i farver, Vol. 1. N. Bladel, ed. Munksgaard,Copenhagen. 1533 BeechwoodAvenue, Fullerton, California 92635. Received 15 Nov. 1983; accepted15 Nov. 1984.