FACTORS AFFECTING FEEDING AND BROODING OF GRAY NESTLINGS

ELLEN J. JOHNSONAND LOUIS B. BEST Departmentof Ecology,Iowa State University,Ames, Iowa 50011 USA

ABSTRACT.--GrayCatbirds (Dumetellacarolinensis) were studied to documentpatterns of feeding and brooding nestlingsand to relate theseto nestlingage, brood size, time of day, time of season,and weather. Major factors affecting feeding were nestling age and brood size. The averagenumber of food deliveries.brood-•. h-• increasedlinearly with the log of nestlingage for both male and femaleparents. Also, volumeof food delivered/feedingtrip increasedlinearly with nestling age. As brood size increased,feeding frequencies/brood increasedbut food deliveries/nestlingdecreased. Brooding and shadingwere affectedmainly by nestlingage and ambient temperature.Time spent brooding decreasedsubstantially as nestlingsgrew older. Shading was important when the sun shone directly on the nest. Femalesspent significantlymore time brooding nestlingswhen rain increased.As ambient temperature increased, time spent shading increased and time spent brooding declined. Received7 May 1980, accepted27 April 1981.

ONE of the greateststresses on during STUDY AREA AND METHODS the breeding seasonis caring for their nest- lings. Frequent feeding trips are required to This study was conductedin a shrubby and partly satisfythe food demandsof the rapidly devel- wooded pasture near Ames, Iowa during the springs and summers of 1977 and 1978. The area covered 16 oping young, and nestling body temperature ha of which approximately50% containedshrubs, must be maintained at a relatively constant primarily gooseberry(Ribes spp.) and multiflora rose level. Ratesof feeding and brooding by parent (Rosa multiflora), suitable as nesting sites for cat- birds can be related either to requirements of birds. A stream meandered through the study site. the nestlings or to environmental factors af- During 1976-1977, a drought occurred in central fecting the adults. Knowledgeof these rela- Iowa. Rainfall in Ames was 52 cm less than the nor- tionshipsis importantto understandbetter the mal of 116 cm during the 15-month period, May nestlingperiod of altricialbirds, but few work- 1976-July 1977 (Natl. Ocean. Atmos. Admin. 1976, ers have addressedthese topics. 1977, 1978). The drought ended in August 1977, and Although the Gray Catbird (Dumetellacaro- during April-July 1978,precipitation was 9 cm above the normal of 43 cm. linensis)is a commonspecies throughout much To locate all active nests, intensive nest searches of the United Statesand its nestingbiology has were conductedweekly over the entire study area. been extensivelydocumented (for a review see The status of active nests was checked at least once Johnsonand Best 1980), factors affecting feed- daily. Nearly all adult catbirdson the studyarea were ing and brooding frequencieshave not been capturedby using mist nets and marked with colored reported. Slack (1973) documented effects of leg bands.Additionally, a spot of paint was placed temperature,time of day, and day of incuba- on the head of one member of each pair. Nestlings tion on female attentivenessduring incubation were marked with a spot of paint on the top of the but reportedvery little concerningbrooding of . nestlings.Zimmerman (1963)collected limited Catbird nests were observed from blinds during information on attentive periodsof female cat- both 1977 and 1978. A mirror was positioned per- manently over each nest so that its contentscould be birds relative to nestling age. viewed. This was done prior to hatching; thus, birds The objectivesof this study were (1) to doc- became accustomed to the mirror before behavioral ument patterns of feeding and brooding nest- observationswere begun. Mirrors were positioned ling Gray and (2) to relate these to so that they did not shade the nest. Feeding fre- nestlingage, brood size, time of day, time of quenciesand time spentbrooding and shadingnest- season, and weather. lings by each parent were recorded.Brooding was

148 The Auk 99: 148-156. January 1982 January1982] Feedingand Brooding Catbird Nestlings 149 distinguished from shading becauseeach served a TABLE1. Total volume delivered/feeding trip and different function. Brooding occurredwhen the fe- estimated volume delivered.nestling-•.h -• for Gray Catbird young of different ages.a male settled down on the nestlings and provided heat to them. During shading, the stood over Volume the nestlings,with wings slightly spread,usually to (cm3) protect them from sunlight (sometimes birds as- Number Volume (cma) delivered. sumed this position when no direct sunlight was •e (x) of trips delivered/trip nestling-•. falling on the nest). Observations were made at each ays) sampled (•)b (• + SD) h •e nest from the time of hatching until the nestlings 0 11 0.10 + 0.08 0.09 fledged. Nests were observed an average of 92 h 1 28 0.11 + 0.07 0.24 2 19 0.14 + 0.11 0.41 each. Ages of nestlings were divided into 24-h 3 12 0.24 + 0.15 0.60 classesbeginning from the time of hatching (0-24 4 31 0.26 + 0.14 0.80 h = 0 day old, 24--48h = 1 day old, etc.). Five nests 5 23 0.27 + 0.12 1.00 were observed in 1977, with brood sizes of 2• 2, 3, 6 23 0.41 + 0.27 1.22 7 46 0.38 + 0.25 1.44 4, and 5, and four in 1978, with sizes of 2, 3, 4, and 8 11 0.35 + 0.35 1.67 5. Nestling growth rates were determined by Rick- lefs' (1967) method. a For the size distributions of individual food items, see Johnson et al. 1980. In 1978, food samples were collectedfrom all cat- b Statisticsare: • = 0.075 + 0.047x, r2 = 0.96. bird nestlings on the area except those for which c Seetext for detailson how this was computed. feeding frequency was being observed. Wire liga- tures were placed around the nestlings' necks, pre- venting them from swallowing food delivered by the size. Because many of the independent variables parents (Johnsonet al. 1980). Nestlings were watched were intercorrelated (Appendix), all variables also from a blind, and, after each feeding trip made by were consideredsimultaneously in a multiple regres- parents, the food was collectedfrom the nestlings' sion procedureto determine the relative importance throats. Food was sampledfrom nestlings0 through of each. The level of significancefor statisticaltests 8 days old; after 8 days, disturbances would cause was set at P •< 0.05 unless noted otherwise. them to fledge prematurely (young normally fledge at about 11 days old). The volume of food items was RESULTS AND DISCUSSION determined by water displacement. Nearly all factorsconsidered had someeffect Each day was divided into five 3-h time periods: dawn-0900, 0900-1200, 1200-1500, 1500-1800, and on feeding or brooding frequencies,and inter- action effects were common. 1800--dusk. Observations from blinds and food sam- pling were conducted during alternate intervals, Nestlingage.---The average number of food with the schedulereversed every other day. In this deliveries-brood-•.h-•, (P), increasedlinearly way, all intervals were representedfor each part of with the logx0of nestlingage, (x), for both male the study during every 2-day period. Temperature (P = 4.14+ 3.74logIx + 1], r2 = 0.76)and fe- (øC),relative humidity (%), barometricpressure (mm male ('•= 1.55+3.43 logIx+ 1], r2= 0.83) Hg), cloud cover.(classes 1 through 5), wind speed parents. The logarithmic relationship shows (classes1 through 7), and light intensity (foot-can- that most of the increasein feeding frequency dies) were recordedhourly. occurredearly in the nestlingperiod. A similar The relationshipsbetween eachindependent vari- able and nestling care were evaluated using either relationshipbetween nestlingage and feeding linear regression analysis or analysis of variance frequencyhas been reported for other species (ANOVA). Nestling age had a noticeable and con- (Morton et al. 1972, Westerterp 1973). sistent influence on both feeding and nest atten- The volume of food delivered to nestlings/ dance. Thus, in ANOVA tests for the effects of brood unit time depends not only on frequency of size, time of season,and weather on the dependent feedingsbut alsoon the amountof food in each variables, a two-way analysiswas used in which one delivery. Both volume delivered/feedingtrip of the factorsalways was nestling age. Values used and the variability in food volume delivered/ in the analysiswere meansof the dependentvariable feeding trip increasedwith nestlingage (Table calculatedfor each age categorywithin each classof 1). These trends occurred through 8 days of the other independentvariable (unweighted analysis of cell means; Snedecor and Cochran 1967: 475). age, after which food could not be sampled. Regression equations were determined by using Most of the increasedenergy requirements of means of the dependent variable for each value of older catbird nestlingswere met by enlarging the independent variable. Using means in both of the amount of food delivered/feeding trip, the above proceduresadjusted for unequal sample rather than increasing the feeding frequency. 150 JOH•SO• •,•v BEST [Auk, Vol. 99

1•,25r ProportionofFeeding 1'5 •_• / Y= 0.272 + O.014x . i' ' ß ß , • O I 2 3NESTLING 4 5 AGE6 (DAYS)7 8 9 IO II

Fig. 1. Nestattendance and proportion of nestlingfeedings by femalecatbirds in relationto nestlingage.

This was accomplishedby feeding largeritems lower than those of males (averaging about as the nestlingsincreased in sizeand were able one-half those of the males over the entire to consumethem (Johnsonet al. 1980). Volume nestling period), but females had almost sole delivered/feedingtrip alsoincreased with nest- responsibilityfor broodingand shadingnest- ling age in Great Tits, Parusmajor (Royama lings. 1966); Starlings,Sturnus vulgaris (Westerterp The averagetime spentbrooding and shad- 1973);and PurpleMartins, Prognesubis (Walsh ing by females decreasedlinearly until nest- 1978).In our study, the number of items deliv- lings were 7 daysold, after which little change ered/tripwas unrelatedto nestlingage (John- in nest attendanceoccurred (Fig. 1). Shading son and Best MS). usuallyoccupied only a smallpercentage of the The volume of food delivered.nestling ]. h-• females' time, but it was important when the for each age in 1978 was estimatedby multi- sun shone directly on the nest (Table 2). Shad- plying values from regressionequations of ing time increasedas the amountof direct sun- numberof feedingtrips.nestling -z-h -z (•' = light on the nest increasedfrom 0 to 50% but 1.16 + 2.67 log[x + 1], r2 = 0.90) and volume remained relatively constant at greater expo- delivered/feedingtrip ('• = 0.075+ 0.047x, sures.Nestlings usually moved away from di- r2= 0.96) versusage (Table 1). The food re- rectsunlight and occupieda shadedpart of the quirementsof the young increasedwith age nest when its exposure to sunlight was low; during the entire food-samplingperiod. Food with exposureabove 25%, however,they de- intake of Great Tit (Van Balen 1973), Starling pended on shading by the female. Brooding (Westerterp1973), and Purple Martin (Walsh decreasedsubstantially as nestlingsgrew old- 1978)nestlings initially increasedrapidly and er, but shading time declined only slightly then leveled off, but data could be collectedfor (Table3). As younggrew and beganto develop thesespecies further into the nestling-devel- plumage,they evidently could tolerate low tem- opmentperiod than for catbirds.Food intake peraturesbut still requiredprotection from ex- of catbirdyoung probablybegins to level off cessive heat from the sun. beyond8 days of age. Brood size.--Brood size was another impor- Theproportion offeeding trips made by fe • tant factoraffecting feeding frequencies.Feed- maleswas inverselyrelated to the amountof ing frequenciesincreased similarly with brood time spentbrooding and shadingthe nestlings size during both years of observation;thus, (r = -0.23, n = 2,075 half-hour intervals).As 1977 and 1978 data were combined (Fig. 2). the nestlings grew, females spent less time Feedingfrequencies/brood (F = 71.73, df = 3, brooding and shadingand, consequently,in- 44) and feedingfrequencies/nestling (F = 6.58, creased their contribution to the total food de- df = 3, 44) varied significantlyamong broods livery (Fig. 1, see also Pinkowski1978). The of different size. Newman-Keuls multiple feeding rates of female catbirdswere always rangetests were usedto determinewhere the January1982] Feedingand Brooding Catbird Nestlings 151

T^Br•E2. Shading time by female catbirdsin relation T^BLœ3. Time female Gray Catbirds spent brooding to the percentageof the nest exposedto direct sun- and shading nestlingsof different ages. light. Time spent Time spent Percentage Number Time spent Number brooding shading of nest of half-hour shading in Age of hours in min/h in min/h in direct intervals min/half hour (days) observed (• + SD)a (• + SD)b sunlight observed (• + SD) 0 100 34.5 + 17.9 7.7 + 19.0 0 1,609 1.3 + 4.9 1 86 27.5 + 18.1 5.3 + 14.6 1-25 55 7.6 + 10.4 2 78 25.0 + 17.2 4.3 + 12.2 26-50 76 12.1 + 12.6 3 92 22.8 + 16.0 5.8 + 11.3 51-75 28 11.6 + 12.0 4 97 14.9 + 16.2 6.6 + 14.5 76-100 15 13.1 + 12.9 5 90 15.6 + 17.8 3.8 + 12.1 6 89 4.2 + 11.1 7.0 + 16.0 7 98 3.5 + 7.3 4.0 + 10.5 differencesoccurred. Feeding frequencies/brood 8 106 5.2 + 10.9 2.1 + 7.5 were significantly different among all brood 9 103 4.4 + 11.3 4.6 + 14.0 10 93 2.0 + 5.3 1.4 + 4.8 sizes. The number of feeding trips.nestling-•' 11 26 1.3 + 1.2 3.3 + 9.0 h-• for broods of 2 or 3 was significantly a Statisticsare: predictedtime brooding= 30.3.9- 3.05 (Age), larger than that for broods of 4 or 5. The 0.90. volume of food delivered/feedingtrip, as de- b Statisticsare: predictedtime shading= 6.67 - 0.36 (Age), 0.48. termined from food samples collected from broods of 3 through 5 in 1978 (no broods of 2 were available for sampling), did not atively little effect on feeding frequencies(Fig. changesignificantly with brood size (two-way 3), although a significantdifference was found ANOVA); adults, therefore, did not adjust for (F = 5.57, df = 15, 1,712).Feeding rates for the different brood sizes in this way. Also, growth first hour after females left the nest in the rates of nestlings from different-sized broods morning were significantly greater than those were not significantly different (Johnsonand for all other times (Newman-Keuls multiple Best MS). Therefore,it appearsthat as brood range test). This was probably because nest- size increases,food requirements/nestlingde- lings were very hungry after a night of fasting, crease.This trend also has been reported for and thus they begged more energetically. many other species(Moreau 1947, Lack and Feeding rates also increased slightly at 1100 Silva 1949, Gibb 1950, Best 1977, Walsh 1978, and in early evening before females returned Pinkowski 1978) and could be because the ratio to the nest to brood for the night, but these of exposedsurface area to biomassdecreases values were statistically different only from with increasingbrood size, resulting in lower those at 1200. No other differences were found. thermoregulatorycosts/nestling (Royama 1966, Peak feeding times for catbirds in Kendeigh's Mertens 1969). (1952) study were at 1000, 1200, and 1800;low- Time spent broodingand shadingnestlings est rates occurred from 1300-1700, when tem- was not significantlyrelated to brood size ac- peratures were highest. Others have reported cording to a two-way ANOVA, contrary to re- highest feeding rates in early morning and suits reported for Purple Martins by Walsh lowest rates at midday (Kluyver 1950, Best (1978). The lack of uniformity in weather con- 1977, Pinkowski 1978). ditions when different nests were being ob- Brooding and shading by females were re- served and the variation in the exposureof lated indirectly to time of day (Fig. 3). Usually different nests to the sun could have masked brooding was highest in early morning and in brood-size effectsin this analysis. In the mul- evening and lowest in early afternoon, paral- tiple regressionresults, however, brooding de- leling diel changesin ambient temperature(see creasedsignificantly with greater brood size. Weather). Shading by females, which peaked This could be because females are forced to in early afternoon,was closelycorrelated with sacrificesome brooding time in order to satisfy changesin light intensity throughout the day. the greaterfood demandsof a larger brood or Shading by males was minimal, never aver- becauseless brooding is required as the ratio aging more than 0.8 min for any half-hour pe- of exposedsurface area to biomassof nestlings riod in the day. decreases. Timeof season.--Feedingfrequencies changed Timeof day.--In general, time of day had rel- little with time of season. Royama (1966) re- 152 Jou•so• A•O BEST [Auk, Vol. 99

•320 -- BroodSize=2 •--'•" -- -- BroodSize=4 .•.-•' • / ---Brood-- ß - Brood Size=3Size=5 .,.•' .-•"

/ ..-.----"- ----

0 I 2 3 4 5 6 7 8 9 DII NESTLI• AGE (DAYS) F•. 2. •eed•n••equenc•es.b•ood-•'h-• (•)• •o•different b•ood s•zes (• = •estli• a•eJn days). Sta- tistics•o• the •especfive b•ood sizes a•e: b•ood size 2• • = 5.05+ 2.43]o•(• + •), r• = 0.52;3, • = 4.85 + &87]o•(• + •), r• = 0.8•;4• • = 5.57+ 7.•5]o•(• + •), r• = 0.83;5, • = 4.•0 + •5.22]o•(• + •), r• = O.gO.The •e•ession Hne •o• the b•oodo• 5 has beenextended below • day old• but no obse•afionswere made •o• that b•ood s•ze at 0 day old. portedthat GreatTits mademore frequent food on broodingbut not on shading.Average day- deliveries to early broods because food size time temperatureswere significantlygreater was smaller than that delivered to late broods, (two-wayANOVA, F = 9.74,df = 4, 4) during attributable to a seasonalchange in nestling Julythan duringJune or Augustfor both years diet. In our study, time of seasonalso affected of the study (an average of 2.9ø and 2.7øC prey specieschosen (Johnson and BestMS; e.g. higher than the averagefor the rest of the sea- caterpillars were eaten more frequently early son for 1977 and 1978, respectively).The av- in the season, moths and grasshoppersmore erageamount of time spentbrooding was least commonlylater), but sizes of individual food (but not significantly,two-way ANOVA) in itemsdelivered were not significantlydifferent July. (two-way ANOVA) throughout the season. Weather.--Precipitation influenced feeding Thus, food delivery rate/nestlingwas not af- frequenciesof both malesand females(Fig. 4). fectedby the type of food delivered. As rain increasedin intensity and duration, Time of seasonhad a slight, indirect effect feeding frequencies declined for both males

• 4 24

•'< I X • x Feeding•'1 •'o •

/

• 0600o•oo ,ooo ,2• E •OODA Y,6oo ,•oo 2000 Fig. 3. Feedingfrequency by bothcatbird parents and time spent brooding and sharing by femalesin relation to time of day. January1982] Feedingand Brooding Catbird Nestlings 153

30 •. - • / • -....?Brooding * -,.,I

•' (7'Feeding

IO q3

L! I I I I 0 I 2 3, 4 5 6 7 RAIN CLASS

Fig. 4. Feeding frequenciesand nest attendancein relation to precipitation. Rain increasesin intensity and duration from 0 (no precipitation)to 7 (heavy rain >50% of the time). and females(two-way ANOVA), althoughthe unimportant) resulted in a linear decrease in relationship was significant only for males time spent brooding by the females; converse- (F = 5.86, df = 4, 12). Femalesspent signifi- ly, temperature increases above 18øC (below cantly more time brooding nestlingswhen rain this level shading ceased)caused an exponen- increased (F = 6.60, df = 4, 12); males never tial increasein time spent shading (Fig. 6). broodednestlings in the rain. Feedingrates of At higher light intensities, brooding de- males declinedduring rain, probablybecause creased(• = 8.00 - 0.49x, re= 0.61) while they spentk•s time foraging in the rain. shadingincreased ('• = -1.00 + 1.04x,r e= Temperature affectedfeeding ratesmostly at 0.88). The reversewas true as relative humidity temperature extremes (Fig. 5). Feeding by fe- becamegreater (brooding: '• = -0.28 + 1.52x, males dropped essentiallyto zero at the ex- re = 0.96; shading:'• = 6.69 - 0.90x,r e= tremes, correspondingto maximal time spent 0.72). The relationships of these two weather either brooding or shading. Males' feeding parameters to nest attendance, however, may rates stayed relatively constantat all tempera- be attributed largely to their correlation with tures (major variations probably were the re- temperature (Appendix, see also Relative im- suit of small sample size); thus, males did not portanceof variables).The multiple regression compensatefor the lowered feeding rates of analysis(Table 4) verified three of the four sig- females at temperatureextremes. Regression nificant relationshipsobtained when the vari- analysesshowed that an increasein tempera- ables were analyzed individually. The seem- ture up to 32øC(above this level brooding was

E?I- '7 .-3/ '

AM91ENT TEMPERATURE (eC)

4 8 12 16 20 24 28 32 ;56 40 AMBIENT TEMPERATURE (øC) b•e•t tempeTatuTe.Statistics •oT bToodin• (calculated •om •o through32øC) a•e: • = 25.70- 0.8•, r• = Fig. 5. Feedingfrequencies in relation to ambient 0.%. Those •o• shad• (calculated•om Z8ø through temperature. •0øC)ate: ]o•(• + Z) = -1.1• + 0.06•,r • = 0.%. 154 JOHNSONAND BEST [Auk, Vol. 99

TABLE4. Standardizedregression coefficients for factorspotentially affecting feeding, brooding, and shading of Gray Catbird nestlings.

Feeding trips. nestling-• .half hour-• Brooding Shading by female by female Variable Male Female (min/half hour) (min/half hour) Brood size -0.06 *a -0.29'* -0.11'* -0.01 Nestling age 0.20** 0.30** -0.64'* 0.01 Ambient temperature -0.17'* -0.15'* -0.36** 0.42** Relative humidity -0.04 0.02 0.19'* 0.05 Wind speed 0.02 0.08** 0.06** -0.06* Cloud cover -0.05 -0.03 0.01 -0.02 Barometricpressure -0.18'* -0.03 0.00 0.14'* Light intensity 0.01 -0.00 0.14'* 0.14' Rain -0.14'* -0.11'* 0.19'* 0.01 Multiple R2 0.18 0.21 0.61 0.30 a * = significantat P • 0.05, ** = significantat P • 0.01. ingly positive relationship between light by both male and femalecatbirds. This would intensity and brooding in the multiple regres- be expectedas energy demandsof the young sion analysis probably is spurious, resulting increasewith age. Although brood size was from high intercorrelation among variables. second in importance for females, it was of The relationshipsof feeding frequencieswith lesser importance for males. Females must relative humidity were the inverse of those spend a certainamount of time broodingand with temperature. Changes in shading and shadingregardless of broodsize. Thus, feed- brooding were complementaryas light inten- ing frequencies/broodare more constant for fe- sity increased;thus, the sum of the time spent males than for males, and, consequently,feed- in nest attendance remained relatively con- ing frequencies/nestlingdecrease to a much stant. Correspondingly, feeding frequencies greater extent for femalesthan for males as were similar at all light intensities. brood size increases(see also Hails and Bryant When considered individually, the other 1979). Barometric pressure was the second weatherparameters (wind speed,cloud cover, most important variable predicting feeding and barometricpressure) were not significant- frequencies by males. Ambient temperature ly relatedto either feedingfrequencies or nest and rain were third and fourth, respectively, attendance.In the multiple regressionanalysis, for both males and females, with feeding de- however,only cloudcover lacked some signif- creasingas these variablesincreased (as dis- icant relationship(Table 4). We have no expla- cussed in Weather). nation for the apparent relationshipsof wind Brooding by femaleswas influenced mainly speedand barometricpressure to feedingrates by nestling age (Table 4). As nestlings grow or of barometricpressure to shading.Brooding older, they developplumage and the ability to increasedat greaterwind speeds,while shad- thermoregulate, dramatically reducing the ing decreased,probably becausethe dissipa- need for brooding. The secondmost important tion of nestlingbody heat increaseswith great- variable affectingbrooding was ambient tem- er convectionalcooling. perature; rain and relative humidity were of Relativeimportance of variables.--Theabso- lesser importance. Of the factors affecting lute values of the standardized regressioncoef- shading, ambient temperaturewas most im- ficients (Table 4) were used to evaluate the rel- portant, followed by light intensity and baro- ative importanceof eachindependent variable metric pressure. in determining feeding frequenciesand nest The predictiveability of the combinationof attendance. (Time of day and time of season variables measured (see R 2, Table 4) was quite were nonlinear variables in the regression low for both male and female feeding frequen- analysis;consequently, they are not included cies. The R 2 values could be low because fac- in the list of standardized regression coeffi- tors not measuredlargely accountfor variation cients.) Nestling age was the most important in feeding frequency.But it is also possible variable affecting feeding frequency/nestling that feeding, particularly over short time in- January1982] Feedingand Brooding Catbird Nestlings 155 tervals (0.5 h in our study), varies unpredict- MERTENS,J. A. L. 1969. The influence of brood size ably and that feeding rhythms are adjusted on the energy metabolism and water loss of over broader time periods to assure sufficient nestling Great Tits, Parusmajor major. Ibis 111: food delivery for normal nestling growth. Reg- 11-16. ulation of nestling body temperature, how- MOREAV, R. E. 1947. Relations between number in brood, feeding-rate and nestling period in nine ever, requires more immediate responsesto species of birds in Tanganyika Territory. J. changesin the variables measured;thus, their Anim. Ecol. 16: 205-209. predictive ability for brooding and shading MORTON, M. L., J. E. ORIJVELA, & S. M. BVDD. 1972. was greater than that for feeding frequencies The biology of immature Mountain White- (Table 4). crowned Sparrows( Zonotrichialeucophrys orian- tha) on the breeding ground. Condor 74: 423- ACKNOWLEDGMENTS 430. NATIONAL OCEANIC AND ATMOSPHERIC ADMINIS- We thank Patricia Heagy for assisting with the TRATION. 1976. Climatological data--Iowa, vol. field work. David Cox and Paul Hinz gave advice on 87. Asheville, North Carolina, Nat. Climatic statistical analyses, and Robert Shaw provided data Center. on light intensity. James Dinsmore, Douglas John- 1977. Climatological data--Iowa, vol. 88. son, John Rotenberry, and two anonymous review- Asheville, North Carolina, Nat. Climatic Center. ers provided helpful suggestionson an earlier draft 1978. Climatological data--Iowa, vol. 89. of the manuscript. Financial support for this study Asheville, North Carolina, Nat. Climatic Center. was received from the Iowa Agriculture and Home PINKOWSKI,B.C. 1978. Feeding of nestling and EconomicsExperiment Station, Sigma Xi, the Frank fledgling Eastern Bluebirds. Wilson Bull. 90: 84- M. Chapman Memorial Fund, and the JosselynVan 98. Tyne Memorial Fund. This is Journal Paper No. RICKLEES,R. E. 1967. A graphical method of fitting J-9849of the Iowa Agriculture and Home Economics equations to growth curves. Ecology 48: 978- Experiment Station, Ames, Iowa, Project No. 2168. 983.

LITERATURE CITED ROYAMA,T. 1966. Factors governing feeding rate, food requirement and brood size of nestling BEST,L. B. 1977. Nestling biology of the Field Spar- Great Tits Parusmajor. Ibis 108: 313-347. row. Auk 94: 308-319. SLACK,R. D. 1973. Breeding biology and behavior GISS, J. A. 1950. The breeding biology of the Great of the catbird (Dumetella carolinensis).Unpub- and Blue titmice. Ibis 92: 507-539. lished Ph.D. dissertation, Columbus, Ohio, HAILS, C. J., & D. M. BRYANT. 1979. Reproductive Ohio State Univ. energetics of a free-living bird. J. Anim. Ecol. SNEDECOR,G. W., & W. G. COCH•½AN.1967. Statis- 48: 471-482. tical methods, 6th ed. Ames, Iowa, Iowa State JOHNSON,E. J., & L. B. BEST.1980. Breeding biology Univ. of the Gray Catbird in Iowa. Iowa State J. Res. VAN BALEN,J. H. 1973. A comparative study of the 55: 171-183. breeding ecologyof the Great Tit Parusmajor in --, & P. A. HEAG¾. 1980. Food sam- different habitats. Ardea 61: 1-93. pling biasesassociated with the "ligature meth- WALSH, H. 1978. Food of nestling Purple Martins. od." Condor 82: 186-192. Wilson Bull. 90: 248-260. KENDEIGH, S.C. 1952. Parental care and its evolu- WESTERTERp,K. 1973. The energy budget of the tion in birds. III. Biol. Monogr. 22: 1-356. nestlingStarling Sturnusvulgaris, a field study. KLVYVER,H. N. 1950. Daily routines of the Great Ardea 61: 137-158. Tit, Parusm. major L. Ardea 38: 99-135. ZIMMERMAN,J. L. 1963. A nesting study of the cat- LACR,D., & E. T. SILVA. 1949. The weight of nest- bird in southern Michigan. Jack-Pine Warbler ling robins. Ibis 91: 64-78. 41: 142-160. 156 JOHNSONAND BEST [Auk, Vol. 99

APPENmX.Correlation coefficientsof independent variables measuredeach hour during feeding frequency observations (includes only those where P • 0.01).a

Ambient Baro- Time of Time tempera- Relative Wind Cloud metric Light season of day ture humidity speed cover pressure intensity Time of day Ambient temperature 0.38 Relative humidity -0.43 -0.58 Wind speed -0.17 0.13 0.28 -0.37 Cloud cover -0.14 -0.26 0.52 -0.16 Barometricpressure -0.35 -0.13 -0.18 Light intensity 0.32 -0.45 0.26 -0.28 0.22 Rain -0.12 0.25 -0.09 0.30 -0.11 -0.20

a Samplesizes for the calculatedvalues range from 703 to 969.