The Nesting Biology of the House Finch in Honolulu, Hawaii

WESTERN BIRDS

Volume 6, Number 2, 1975

THE NESTING BIOLOGY OF THE HOUSE FINCH IN HONOLULU, HAWAII

LAWRENCE T. HIRAI, Department of Zoology, University of Hawaii, Honolulu, Hawaii 96822

The House Finch, ½arpodacusmexicanus frontalis (Say), is a small seed-eatingpassefine found throughout western North America. It was introduced to Hawaii, probably from San Franciscoas an escapedcage- bird, prior to 1870 (Grinnell 1911). Today it is found on all the main Hawaiian islands. This was a field study of the nesting biology of the House Finch from January 1972 through July 1974, coveringthree nestingseasons. Some of the findingsare discussedin relation to what is known about HouseFinch populationsin North America.

METHODS AND STUDY SITE

Most of the study consistedof field observations,with almost daily checksof nests. Measurementswere madeusing a caliper,ruler, or tape measure.Weights were obtainedusing Pesola scales. To facilitate future identification, eggswere marked with a black "Shatpie" pen; nestlings were marked in 1972 with red nail polish, but in 1973 and 1974 with a red Shatpie pen, which provedmore satisfactory. Nestlingsnear fledging and adults were bandedwith U.S. Fish and Wildlife Servicemetal bandsand one or two plasticbands in different color combinations.Un- less otherwise indicated calculations are in the form mean + one standard deviation. The 0.05 value is consideredthe level of rejection. The study site was confined to the University of Hawaii 0.4 km' main campus,located in Manoa Valley, a residentialarea in Honolulu. WesternBirds 6:33-44, 1975 33 HOUSE FINCH

The House Finch is one of about 15 bird speciesfound on or near the campus. All are introducedexcept the migratoryand nonnestingGold- en Plover (Pluvialisdominica). From 25 to 50 pairs of HouseFinches nest on campusat any one time, the nestingseason covering about six months from mid-February through August. The breedingseason is followed from late July-Augustthrough October by a molting period.

THE NEST

In 1972 I found 91 nests, in 1973 103 nests, and in 1974 63 nests, under construction,with eggsor young. Nestsdo not usuallypersist from one year to the next or eventhrough one nestingseason; House Finches build new nests for each effort. Because House Finches do not defend very large territories, usuallyat the most only the smallareas around the active nests,nests are often in closeproximity to nestsof other House Finchesand bird species. The 257 neststhat I found were built in 26 different types of vege- tation. Pandanus(Pandanus) and palms (Palmae) were the most com- mon nestingtrees, with 41.2% and 28.0% of the total numberof nests, respectively. Monkeypod (Samaneasaman) with 7.4% and Fiddlewood (Citbarexylumspinosum) with 6.6% were other frequently usedtrees. Nestsusually were constructedon sitesproviding cover and shade,such as in clustersof pandanusleaves, on stemsor axesof palm leaves,or in foliagelocated in the outer portionsof trees. The height of a nest was measuredfrom the bottom of the nest to the groundbelow if the nestwas lower than 4.6 m or estimatedas close- ly as possibleif higher. Heightswere variable,ranging from 1.8 to 15.0 m, with an averagein 1972 of 4.3 + 1.5 m, in 1973 of 5.0 + 2.4 m, and in 1974 of 4.6 + 1.6 m. Nest constructiontakes from 6 to 22 days, averaging11.8 + 4.7 days for 15 nests. Nest buildingis done almostexclusively by the female. Only materialbrought to the nestby the femaleis used,and the male aids the female in molding the nest only in the first half of the nest-buildingperiod. Thereafteronly the female workson the nest,the male accompanyingthe female to and from the site, singingfrom a nearby perch. Preparationof the site, suchas enlargingit by nibbling away adjacent leaves,takes the first one or two days of nest building. Then the foundation is laid down and the nest is molded into shape. At this stage larger-sizedmaterial is used, such as old Fiddlewood inflo- rescences,fibrous strips of Coconut (Cocos nucifera) and pandanus, freshand dried grassesand leaves,string, paper strips, and feathers. The cup lining is finally added, consistingof fine and soft coconutfibers, hair, and frayed cigarettefilters. HOUSE FINCH

The finished cup-shapednest variesfrom a rather loose to a very compact structure. Nest dimensionsof 25 nestsmeasured in 1973 were (in cm): outer width x length, 9.5 ñ 1.3 x 12.0 ñ 3.3; inner width x length, 5.5 + 0.6 x 6.2 + 0.6; outer cup depth, 7.3 ñ 1.7; inner cup depth, 4.2 ñ 0.8; and rim thicknessfrom 1.0 ñ 0.6 to 4.7 + 2.8.

EGGS AND CLUTCH SIZE

Copulation occursat the invitation of the female and is seen,at the earliest,a few daysbefore nest building is completed.I havenever seen malecourtship display lead immediately to a copulatoryattempt. Egglaying usuallystarts the day after nestbuilding ends, although I observeda female adding material to a 1974 nest that containedat leastthree eggs. The first egg,and most probablythe entire clutch,is laid in the early morninghours, before 08:00. Usuallyone egg per day is laid until the clutchis completed.Out of a total of 44 nests,only five had a day or moreskipped in the layingof the clutch. In no case wasmore than one egglaid per day. House Finch eggsare colored light blue, with black or brown-black specksor lines concentratedat the roundedends. Eggmeasurements were taken only from neststhat were found on or before the day the first eggswere laid and where the clutcheswere completed. Eggswere measuredand weighedthe day they were laid. Basedon 197 eggsfrom 48 nestsfrom the threestudy years, greatest width x lengthand weight averaged13.5 -+ 0.6 x 19.1 + 0.9 mm and 1.89 ñ 0.15 g, respectively. Eggswithin a clutchdo not showsignificant differences in weightwheth- er they were laid first, second, third, fourth, or fifth in a clutch, in small or large clutches,or early or late in the nestingseason (one-factor anovas, P)O.05). Bergtold(1913) gavemeasurements of House Finch eggsfrom Den- ver, Colorado,that averaged13.7 x 19.6 mm, somewhatlarger than my measurements.The largereggs from Denveralso were heavier in weight. Basedon two clutchescontaining four and five eggs,an eggweighed 2.25 ñ 0.08 g. There is a significantdifference in the weightsof the eggsfrom Denverand Honolulu (two-tailedt-test, t -- 7.030, d.f. -- 204, P•O.00'l). Without other information, such as the availability of food for the HouseFinch populationson campusfor the pastthree yearsand in Denverduring the early 1900s,no adequateexplanation is possible. Clutchsize on campusaveraged about four eggsand doesnot differ significantlyfrom those reported from Arizona and California (two- tailed t-tests, P)O.05; Table 1). Only papersthat have information amenablefor statistical comparisonare used in Table 1. Hensley's (1959) data were slightlymodified in that 1 usedn: 11, insteadof his 35 HOUSE FINCH

Table 1. Clutch sizes for House Finches

CALI- CALI- ARIZONA FORNIA FORNIA HAWAII RANGE 3-5 3-6 4-6 2-5 MEAN 4.0 4.3 4.4 4.2 STANDARD DEVIATION 0.4 0.9 0.6 0.6 NO. OF NESTS 11 18 25 127 SOURCE Hensley Grinnell Evenden This (1959) and (1957) Study Linsdale (1936) n: 12, because one of the clutch sizes could not be determined from the'paper. Also Grinnell and Linsdale (1936) caution that the four neststhat they found with 3-eggclutches may havebeen incomplete.

THE INCUBATION PERIOD The incubationperiod, the length of time betweenthe layi•ngof the last egg in the clutch until that egg hatcheswhen all the eggshatch, rangedfrom 11.5 to 13.5 days,averaging 12.8 + 0.6 daysfor 29 nests. Clutch sizesof the nestsused to calculatethe incubationperiod ranged from two to five, but did not affect the lengthof the incubationperiod (one-factor anova, F = 3.37, m/n = 3/25, P = 0.1-0.05). The incubation period in Hawaii does not differ from the length of 13.3 + 1.1 days (based on 11 nests) determined by Evenden (1957) at Sacramento, California (two-tailed t-test, t = 1.887, d.f. = 38, P = 0.1-0.05). Because of water evaporation,a HouseFinch eggloses 15.9 --+6.1% of its origin- al weightbefore hatching (based on 58 eggsfrom 22 nests). An egg pips, at the earliest,less than 24 hoursbefore the young hatches. Hatchingmay occur at any hour of the day or night, and the eggsof a clutch hatch out overa periodof daysand not all on one day. In rare instancesone young may hatcheach day until all the younghave emerged. Usuallytwo younghatch on the first day of hatchingand one youngper day thereafteruntil all the younghave hatched. I havenever found all the eggsof a clutch hatching on the same day. Eggsalso hatch in the order that they are laid. In six caseswhere only one young hatchedon the first day of hatchingand wherethe entire clutcheventu- ally hatched,the first egglaid was the one that hatchedfirst. In nine other cases,where more than one young hatched on the first day, the remainingeggs in the clutcheshatched in the order in which they were laid. The female gets rid of the egg pieceseither by flying off with them or eatingthem at the nest. 36 HOUSE FINCH

Basedon the hatchinginformation and on temperature-proberead- ingsat two 1973 nests,incubation, the applicationof heat to the eggs by the adult HouseFinch, possiblystarts by the night the first eggof the clutchis laid andalmost certainly by the secondnight, and not when the penultimateor last eggis laid, as in someother bird species.

THE NESTLING PERIOD

At hatching,a HouseFinch is helplessand blind, with somefluffy whitish down on the head and body. A chick fledges,or leavesthe nest, 14 to 19 days after hatching,averaging in 1972 16.8 + 1.1 days (basedon 12 youngfrom six nests)and in 1973 18.5 + 1.0 days(based on 24 youngfrom eightnests). By this time the youngis ratherfully feathered,with a short,stubby tail, onlya few strandsof downadhering to the headfeathers, and distinctivelyswollen beak corners. Nestlingsare fed regurgitatedseeds. Nestlingsfrom 15 nestsin 1972 and28 nestsin 1973 wereweighed daily or almostdaily. Figure1 shows the meanweight + two standarddeviations of a nestlingfrom day 0 (theday the young hatches) until day13. Thereis no significant dif- ferencebetween 1972 and 1973 in the weightsfor the respectivedays (two-tailedt-tests, P>0.05). Thereis a steadyincrease in body weight of a chickfrom lessthan two g on the day of hatchingto about17 g, 13 dayslater. AlthoughI couldnot obtainweights beyond day 13 without frighteningthe young prematurelyfrom the nest, in two casesin 1973 a just-fledgedyoung, 19 daysold, and a chick,18 daysold and a day away from fiedging,both weighed19.0 g, suggestingthat a House Finch nestlingprobably increases in weightto approximatelythat of an adult, 19.5 + 1.6 g (basedon weights,ranging from 16.9 to 23.8 g, of 43 adultstrapped in 1973, with no differencein maleand female body weights;two-tailed t-test, t = 1.575, d.f. = 41, P = 0.4-0.3). At Denver,Colorado, Bergtold (1913) weighedeight young from two nests. At one nest four young hatchedon the sameday; at the other nest two nestlingshatched on the first day, one eachon the secondand the third days. Sevenyoung took from 14 to 16 daysto fledge;it is unclearif the eighthchick fledged. Figure 1 showsthe averageweight of a Denver nestling. The Coloradochicks weigh significantlymore at hatchingand throughday 9 than the Hawaiiannestlings, but from day 10 to 13 there are no differencesin the weightsof the young from either localities(two-tailed t-tests). The fact that the nestlingsfrom Denverand from Honoluluweighed the sameby nestlingday 13 would seemto indicatethat the different growth curveswere not affectedby the food supplyfor the nestlings.The Denveryoung, however, hatched from eggsthat were significantlyheavier than those from which the Honolulu chickshatched, and it may be that the rate of growth of a 37 HOUSE FINCH

_c)

(SW¾•10) iHOl::lM ONI'I.LS::IN

38 HOUSE FINCH nestlingis affectedby the weightof the egg. Schifferli(1973) hadsimi- lar growth patternswhen he comparedweight developmentsof Great Tit (Parusmajor) nestlingshatched from heavyeggs with thoseyoung hatchedfrom light eggs,concluding that eggweight canaffect the rate of growth. BecauseHouse Fincia young hatch over a periodof days and not all on the same day, there are differencesin chick weightsearly (days 2 and 3)and later (days 9, 10, and 11) in the nestlingperiod; older young, thosethat hatchedon the first and seconddays of the hatchingperiod, weigh more than their youngersiblings, those that hatch on the third and fourth days (one-factoranovas, P•0.05). Early in the nestlingperiod, the eyes of the older young open before their still sightlessand youngerbroodmates, giving the older chicksan advantagein beingable to direct their gapestoward the feedingadult and possiblygetting more of the food. Later in the nestling period, the older young are larger, more vigorousat feedingthan their youngerchickmates, and thus probably get more of the food. Althoughthere seemsto be somedegree of siblingcompetition for food broughtby the adults,in most casesHouse Finch nestlingsthat hatchedlater usuallyalso survived to fledge.

THE FLEDGLING PERIOD

The young fly well when they leave the nest and seemto moveout of the nestingarea soon thereafter. A brood may fledgeall in one day or over a numberof daysand seemto do soin the morning hours. Based on limited data, fledglingsare fed by both adultsfor at leasttwo weeks and probably are independentafter three weeks. The adult pair then renests. Even with a six month nestingseason in Hawaii, most likely only two broods can be raised successfullyin one seasonby a pair be- causeof the longnesting period (about two months,including 20 days before renesting)and the low nestingsuccess that I found. In North America two broods in one seasonseem to be averagefor the House Finch (Evenden 1957, Gill and Lanyon 1965, Hensley 1954).

ADULT NESTING BEHAVIOR Daily or almost daily observationsof 60-minute duration in 1972 and 1973 at 9 nests during the egg-layingperiod, 23 nestsduring the incubationperiod, and 25 nestsduring the nestling period indicate that only the female House Finch incubatesthe eggsand broodsthe young. She is attentive at the nest during the daylight hoursabout 50% of the time on the day the first eggis laid and 90% of the time after the clutch is completedand until the eggshatch. Nestattentiveness then declines 39 HOUSE FINCH

until 10 days after the young hatch, whereafter day brooding is in- frequently observed. Night brooding ends about 12 to 14 daysafter the young hatch. The male is seen at or near the nest about once an hour during the incubationperiod, returningto courtship-feedthe female regurgitated seeds. Except in one case,males were not observedat the nestsat sun- set or night, and it may be that they roost togetherelsewhere during the nestingperiod, as reported by Evenden(1957). House Finch nestlingsare fed by both the male and the female;also during the first one-thirdof the nestlingperiod the male feedsthe female at the nest and she,in turn, feedsthe nestlings.It is rare when no feedingsoccur during a 60-minute period and as many as four feeding trips may be made, averagingabout two feedingsper hour. The usual pattern in feedingthe nestlingsis for the adultsto alternatein returning to the nest. Out of 30 feedingsnoted at 13 1973 nestswhere the young were more than sevendays old, only four consecutiveones were by the same adult and only three feedingswere within one minute of each other;times between feedings ranged from 1 to 38 minutesand averaged 16.7 __+12.0 minutes.

NEST SANITATION The House Finch belongsto the Carduelinae,a subfamilyin which someof the speciesdo not removethe fecal sacsof the youngfrom the nest, a behavioraltrait unusualfor passefinebirds. Duringthe early part of the nestlingperiod, both the male and the female removethe sacs,usually by eatingthem or, very rarely,by carryingthem away. The first sacsappear on the rim from four to ninedays after the first young ha.tch,averaging 6.1 & 1.3 days(based on 29 nestsinspected daily in 1972 and 1973). Especiallywhen three or four youngfledge, the nest rim may be coveredwith fecal matter, althoughthe cup itself is often clean,partly becausethe older nestlingsraise their cloacalregions over the rim when passingfecal sacs. This fecal matter makesfor an ideal "home" for many other animals, especiallyarthropods. Usingthe Tullgren-funnelmethod, the nestfau- nas from three 1972 nestswere extractedand identified, revealinga combinedtotal of 17 different types of animals. The mostinteresting animal found is a blood-suckingmite (Mesostigmata:Dermanyssidae). It was presentat everynest I found, rangingfrom beingscarcely notice- able to literally coveringthe nest and the nestlings.Because the chicks gainedweight and fledged evenat those nestswhere the mite populations were in the thousands, I do not believe that this mite is a direct causeof nestlingmortality. The mites, though, may so weaken the 40 HOUSE FINCH young that its chance of survivalis greatly reducedafter fledging. Bergtold(1913) collecteda similarmite.

NESTING .SUCCESS

Nesting successmay be defined in a number of ways. In this paper nestingsuccess is defined as the proportion of the number of young that fledge in relation to the numberof eggsthat are laid; hatchingsuc- cessas the proportion of the number of young that hatch in relation to the number of eggsthat are laid; and nestlingsuccess as the proportion of the number of young that fledgein relationto the numberof young that hatch. No nests were consideredin computing nesting success where even one egg or young was lost, or believed lost, by my actions or by others. Also, only neststhat were found before hatchingbegan are included. Table 2 listsnesting success for House Finch populationsin Honolulu and North America, primarily urban and suburban environments. A number of the studieswere done for more than one year, but for ease of computation a combined nesting successis given for each locale. Nestingsuccess in Hawaii for the three seasonsis 21.7% and compares similarlywith the finding of West (1972) in New Mexico,but is signifi- cantly lower than in the other locations(2x2 X' tests). The highernest- ing successesof Evenden's(1957) and Hensley's(1954) House Finch populationsresult from both higherhatching and nestling successes. The HouseFinch in Hawaii hasa muchlower nestingsuccess than mostother populations. In' fact, nestingsuccess for the HawaiianHouse Finch is far lower than the 46% calculatedby Nice (1957) for temperatealtricial species. It is closerto the 30% found by Skutch(1966) for Central Americanspecies that build openor roofednests. Although the nesting successof the Hawaiian House Finch is low, it is an abundant bird. A total of 403 eggsor youngwere lost in the three study years. Nest losseswere due to: strongwinds knocking eggs and young from nests (34.7% of the total losses);predation, probably by a rat species,on eggs, young, and in six casesthe nestingfemales (27.5%); failure of eggsto hatchbecause they were infertile, containeddead embryos, or were de- serted (17.4%); and nestlingsdying becausethey were inherently weak at hatching,starved to death, or becameso entangledin the nestmateri- al, especiallythe hair used to line the cup, that they could not free themselvesand were left behind when the rest of the brood fledged (20.3%). To an unknown extent, House Sparrows(Passer domesticus) interfere in House Finch nestingby stealingmaterial from active nests, incorporatingthe material into their own, and probably even killing nestlingsby peckingthem to death and throwingthem out of the nests. 41 HOUSE FINCH

Z m

Z •

z• z o•

42 HOUSE FINCH

Similar House Sparrow interferencesare reported from Colorado and California (Bergtold 1913, Evenden1957).

SUMMARY

The nestingbiology of the HouseFinch, Carpodacus mexicanus frontalis (Say), was studied for three nesting seasonson the Universityof Hawaii Manoa campus. From 25 to 50 pairs of House Finchesnest in the study area from February through August, with nest construction taking approximately12 days, the incubation period 13 days,the nestling period 17 or 18 days, and the fledgling period from two to three weeks. Clutch size averagesfour eggs,and a pair mostlikely successful- ly raisestwo broods per nesting season. Nesting successin Hawaii is low, 17% in 1972 and 1974 and 30% in 1973. Findings indicate that the nesting biology of the House Finch in Hawaii doesnot differ greatly from populationsin North America. Sig- nificantdifferences seem to exist,though, in eggweight, nesting success, and possiblyin the nestlinggrowth-pattern.

ACKNOWLEDGMENTS

My appreciationto Dr. Andrew J. Bergerfor his guidanceduring the entire period of the study and for his editi_ngof this manuscript;Dr. John S. Stimsonand Charlesvan Riper, III, for their constructivecom- mentsduring the latter part of the study; SusanTakemoto for her assist- ancein preparingthe graphs;and Darwin S. Yoshiokafor his kind assist- ancein identifying the animalsextracted from the 1972 nests. For their assistance,support, advice, encouragement,and friendshipduring this study, my deepestmahalo to William Y. Brown, Patrick Conant, Sandra J. Guest, Leighton M. Hirai, Kyong Nan Kim, H. Eddie Smith, and my parents. This study was supported,in part, through Dr. A. J. Berger,by NSF Grant No. GB 23230 of the Island EcosystemsIRP under the US/ InternationalBiological Program, and was preparedas partial fulfillment of the requirementsfor a Masterof Sciencedegree in Zoology. A fuller report of the study is found in Hawaii IBP TechnicalReport No. 47.

LITERATURE CITED

Bergtold,W. H. 1913. A study of the House Finch. Auk 30:40-73. Evenden, F.G. 1957. Observationson nesting behavior of the House Finch. Condor 59:112-117. 43 HOUSE FINCH

Gill, D. E. and W. E. Lanyon. 1965. Establishment,growth, and behaviorof an extralimital population of House Finchesat Huntington, New York. Bird- Banding 36:1-14. Grinnell, J. 1911. The Linnet of the Hawaiian Islands:a problem in speciation. Univ. Calif. Publ. Zool. 7:179-195. Grinnell, J. and J. M. Linsdale. 1936. Vertebrate animals of Point Lobos Re- serve,1934-35. CarnegieInst. WashingtonPubl. 481. Hensley,M.M. 1954. Ecologicalrelations of the breedingbird population of the desertbiome in Arizona. Ecol. Monogr. 24:185-207. Hensley, M. M. 1959. Notes on the nestingof selectedspecies of birds of the Sonoran Desert. Wilson Bull. 71:86-92. Nice, M. M. 1957. Nesting successin altricial birds. Auk 74: 305-321. Schifferli, L. 1973. The effect of eggweight on the subsequentgrowth of nestling Great Tits Parusmajor. Ibis 115:549-558. Skutch, A. F. 1966. A breedingbird censusand nestingsuccess in Central Amer- ica. Ibis 108:1-16. West, S. 1972. House Finch nesting in New Mexico. Elepaio 33:52-53.

Sketcb by Err Deis