Oecologia. Reproductive responses to varying food supply in a population of Darwin's finches: Clutch size, growth rates and hatching synchrony

Transcription

1 Oeclgia (Berlin) (1985) 66:411~416 Oeclgia 9 Springer-Verlag 1985 Reprductive respnses t varying fd supply in a ppulatin f Darwin's finches: Clutch size, grwth rates and hatching synchrny Trevr Price Divisin f Bilgical Sciences, The University f Michigan, Ann Arbr, MI , USA Summary. I shw hw fd shrtage affects reprductin in a ppulatin f Darwin's Medium Grund Finches, Gespiza frtis. Despite the cmmn ccurrence f starvatin and absence f nest predatin, hatching is typically highly synchrnus and adaptive brd reductin appears t be absent. Variatin in bth grwth rates and clutch size in assciatin with the varying cnditins is dcumented. This variatin is interpreted as being a direct respnse t envirnmental cnditins rather than adaptive phentypic plasticity. I cnclude that selectin pressures t raise ne r tw chicks during times f fd shrtage, r t delay grwth rates, are weak r absent. Variatin in clutch size, in grwth rates and in the degree f hatching synchrny have ften been discussed in terms f their pssible adaptive significance in enabling ne r mre members f a brd t survive a perid f fd shrtage during which the whle brd might therwise starve (Ricklefs 1965, 1968; Lack 1968; O'Cnnr 1977). In cntrast t these adaptive interpretatins, variatin in clutch size and in grwth rates may be simple, nn-adaptive, cnsequences f envirnmental cnditins (Bryant 1978; Hwe 1978; Hrsfall 1984). Several alternative adaptive explanatins, apart frm respnses t fd supply, have been prpsed fr the presence f hatching asynchrny (Clark and Wilsn 1981 ; Hahn 1981). In this paper I cnsider the way in which these characteristics change in respnse t variatins in fd supply in a ppulatin f Darwin's Medium Grund finches, Gespiza frtis, n I. Daphne Majr in the Galfipags Islands. The ppulatin shws extreme variatin in reprductive success. This is a cnsequence f the variable rainfall which in turn affects the availability f the fd that finches eat (Grant and Bag 1980; Bag and Grant 1984). In 1978 nly 24 mm f rain fell, and breeding cmpletely failed (Bag and Grant 1984). At the ther extreme, in 1983, 1,359 mm f rain fell, and breeding was very successful, with sme pairs raising at least seven brds (Gibbs et al. 1984). The bserved multiple brds and their verlap in time (Bag and Grant 1984) are clearly adaptive respnses enabling pairs t raise many ffspring when fd is abundant (cf. Grant and Grant 1980; Burley 1980). O'Cnnr (1977) has discussed the way in which the timing and predictability f fd shrtage might give rise t selectin fr clutch size r grwth rate variatin, r hatching asynchrny. First, if cnditins at the time f egg-frmatin prvide a gd indicatin f thse at the time when nestlings are being fed, clutch size may be adjusted s that the ptimum number f nestlings are prduced. Secnd, if cnditins are largely unpredictable at the time f egg frmatin, but nt highly variable during the nestling perid, it is predicted that hatching will be asynchrnus and brds will be reduced thrugh the death f the smaller, later-hatched chicks. Finally, if cnditins are bth unpredictable and variable in the shrt term, a temprary slwing f grwth preceded by fat strage may ccur. These alternative strategies are nt necessarily exclusive, and a cmbinatin may be favred by selectin. Fr example, a mdificatin f grwth rates culd allw a finer adjustment f the energy needs f a brd t the energy supply than wuld the lss r additin f whle chicks (Ricklefs 1968, 1976). During my study, which lasted frm , several pairs attempted t raise mre than ne brd nly nce (in late 1981, Price 1984a). In cntrast, during fur perids many pairs failed t raise any yung. I shw hw grwth rates and clutch sizes varied in assciatin with the different cnditins, whereas hatching synchrny remained high. I use the results t argue fr the absence f any strng selectin pressure n parents t raise less than a full brd during times f fd shrtage. Methds Mst f the methds used in this study, and a descriptin f the island and its fauna and flra, can be fund in Millingtn and Grant (1983, 1984) and Bag and Grant (1984). I spent mst f the first six mnths f each year n I. Daphne, and lcated active nests f G. frtis as they were being cnstructed. Clutch size, and hatching and fledging success, were recrded fr all nests (Price 1984b). The incubatin perid fr these finches is typically twelve days frm the laying f the last egg, and fledging ccurs days after hatching (Bag and Grant 1984). The number f fledglings was usually knwn frm direct bservatins, but in sme instances in late 1979 and in 1981 these bservatins were lacking, and then the number f fledglings was taken t be the same as the number f 8 day ld chicks in the nest, unless a chick was greatly underweight. Chicks were weighed (in g), measured (in ram), and banded at age 7-9 days. Between 50 and 100 chicks each

2 - 412.] ' 9 $$ 1981 e ~_ d w i.~.~ ~ FEll MAR APR MAY dun MONTHS FEB MA~ APR MAY MONTHS JUN FEB MAR AI:~ MAY MONTHS JIJN, [ 5,~.a z '~ 9:, ~ ~ ~: t ~ 0.6 m ~ g = ~,=,..... m I l ~ FEB MAR APR MAY JUN FEll MAR APR MAY JOH FEll MAR APR MAY,KIN MONTHS MONTHS MONTHS Fig. l. Fd supply, fraging and clutch initiatins fr each year Abve: bimass f caterpillars (dry weight) cllected in 20 min searches f Bursera gravelens trees (... ) and Prtulaca hwellii plants ( --), and the number f Prtulaca flwers cunted n 50 plants (... ). Numbers f Prtulaca flwers are frm Millingtn and Grant (1983). Belw: first egg dates fr first clutches in each breeding but (g/z/a) and the prprtin f feeding bservatins that were n fd that appeared in respnse t rainfall f that year ( ). Symbls abve each plt indicate rainfall : mm, ram, mm year were weighed and measured at tw day r lnger intervals frm day 0 (hatching), 1, r 2, until 8 t 10 days ld, r until they died. The measurements used here are beak depth, (measured at right angles t the beak cmmisure abve the nares), and length f the furth primary (including sheath, frm the feather base t tip). Lgistic grwth equatins, y = A/(I + exp (- k (x- t))), where y is the character, x the age in days, and k, A and t cnstants (e.g. Ricklefs 1979) were fitted t the grwth data. I used the iterative prgram BMDPAR in BMDP t d this. Mst f the parents f the chicks had been uniquely clur banded and measured (Price 1984b). Flwer pllen, caterpillars and small seeds are imprtant items in the diet f nestlings and fledglings (Bag and Grant 1984). The availability f these fds was quantified by cllecting caterpillars in randm 20 rain searches f vegetatin types, by cunting flwers f Prtulaca hwellii n 50 specific plants every day, and by cunting seeds n plants in five 4 m 2 plts every ten days. Caterpillars were saved in alchl and dried and weighed in the labratry. Observatins f feeding behavir were made n a regular mrning walk cnducted at 10 day intervals ver apprximately tw-thirds f the island; the first item f fd cnsumed by each bird was recrded9 Results Fd supply and breeding respnses The Galfipags receive highly unpredictable amunts f rain, falling at lw elevatins almst entirely during the first six mnths f the year (Grant and Bag 1980). Finches 0 z c9 ~0 J z 500- r "T- I O O...I L.U I-" < r', v ~ 4- ~ LU ~) 3 6'1 (.9 ~) EARLY 1979 LATE 1979 EARLY 1980 LATE BREEDING PERIOD Fig. 2. The number f eggs ( [] ), hatchlings (=) and fledglings ( 9 ) in each breeding perid. The few nests in each perid (~ 3) fr which either eggs, r hatchlings r fledglings were nt knwn are excluded. The number f fledglings was directly bserved except fr sme nests in late 1979 and I981 when it was taken t be the same as the number f 8 day ld chicks, unless a chick was substantially underweight start breeding within tw weeks f the first rains (Fig. 1). During rainfall was generally lw, with annual ttals f 69, 54 and 80 mm respectively (Millingtn and Grant 1983; Fig. 1). It stimulated egg laying in five separate perids (Fig. 1); and there was additinally a perid f singing and nest-building, but n egg-laying, in respnse t rain in February, Substantial numbers f chicks fledged frm nly tw f these five perids (Fig. 2). c

3 413 Table 1. Clutch sizes fr first clutches in each breeding perid. Means, standard deviatins and sample sizes (in parentheses) are given, separately fr females brn befre 1976, and in N 1978-brn females bred in early 1979 Early 1979 Late 1979 Early 1980 Late Pre-1976 females (16) 3.4 _ (23) (18) (25) (20) 1978 females - 2.8_+0.80 (22) 2.4_+0.92 (11) 2.9_+0.65 (28) (36) Table 2. Hatching asynchrny. The prprtin f nests visited when chicks were 0-1 days ld in which an egg had still t hatch (with the number f nests in parentheses), and variance in weight amng hatched nestlings at 0-1 days is given Year Prprtin fnests with Variancein an egg which later hatched Weight, lg g (29) (16) (18) Nest predatin has nt been knwn t ccur n the island. Therefre, breeding success can be related dearly t fd supply (Fig. 1). In bth successful breeding episdes ver 90% f bserved feedings were n fd that was available as a direct result f the year's rainfall (Fig. 1), and nestling crps were full f caterpillars, Prtulaca pllen and small seeds, all prduced in respnse t the rain. Variatin in the relative abundance f caterpillars and Prtulaca flwers thrugh time is shwn in Fig. 1. Small seed prductin is nt shwn: f the mst cmmn plants, Amaranthus sclerantides and Berhaavia erecta reached peak seed abundance days after rain, Prtulaca at least 45 days after rain. The imprtant bservatins fr the purpse f this paper are (1) at the time f egg frmatin individuals are feeding substantially n rain-respnse fds, (2) there is a delay f frm several days up t ne mnth (in the case f the seeds f sme plant species) befre the different rainrespnse fds becme abundant and (3) during times f unsuccessful reprductin, fd supply and fraging n rain respnse fds decline while chicks are still in the nest (Fig. 1). Figure 2 shws the number f eggs, nestlings and fledglings in each perid. Large numbers f clutches were deserted in all perids in 1979 and The majrity f abandned eggs cntained chick embrys. The timing f desertins may be clsely matched t a deteriratin in the fd supply. In late 1979 clutches were apparently deserted as caterpillars disappeared but befre Prtulaca flwers appeared. All the deserting females subsequently laid a secnd clutch, presumably in respnse t the resurgent fd supply. In late 1979 mst chicks fledged, but in late 1980 there was substantial chick mrtality in the nest (Fig. 2). Variatin in reprductive respnses Clutch size. Mst females were knwn t have been brn befre 1976, r in 1978, s age-chrt membership can be identified and cntrlled fr in cmparisns f clutch sizes amng years. Nearly all clutches were f three r fur eggs. Therefre, I used Z 2 tests and cmpared clutches f three Table 3. Brd reductin amng different-aged nestlings frm ages 0 t 8 days. Nests included in the analysis are thse in which mre than tw chicks hatched, and which were visited when the chicks were 0-1 days ld, apprximately 4 days ld, and again at apprximately 8 days ld. Nests which failed cmpletely befre age 8 days are ttaled separately Number f brds Partial brd N brd Cmplete Ttal reductin reductin nest failure < 4 days > 4 days u_ ~ O 8 s " 9 "'" V N 9 NUMBER OF CHICKS HATCHED Fig. 3. Number f chicks fledged in relatin t the number hatched frm all brds fledging at least ne yung in early 1979 and in Dashed line indicates 100% fledging success Table 4. Lgistic grwth curve parameters. K (in days 1) is a measure f grwth rate, A (in g r ram) is the asymptte. Only chicks first measured at 0 t 2 days ld are included. Sample sizes are fr the number f chicks measured until at least age day 6, with the ttal number f measurements in parentheses Late 1979 Late Adult size Mean+s.e. Mean_+s.e. Mean_+s.e. (apprximate) Weight K 0.42_ _ _+0.03 A 15.84_ _ _ g Beak depth K 0.22_ _ _0.02 A 6.33_ mm Furth primary K !0A0 - A 23.28_ _ mm Sample size 71 (323) 22 (116) 75 (341)

4 414 2 l i 0i 77 i 2i AGE IN DAYS AGE IN DAYS 12 Fig. 4. Grwth curves fr the characters bdy weight and beak depth in late 1979 and late The curves fr 1981 are very similar t thse fr 1979 r fewer eggs with thse with fur r mre t test fr differences amng breeding episdes. These tests gave Z ] = 26.2, P<0.001 fr pre-1976 females and )~2=44.3, P<0.001 fr 1978 females), indicating significant variatin in clutch size amng perids (Table 1). This variatin is nt a functin f different grups f females breeding at different times. Fr example, amng the 19 females brn befre 1976 which laid eggs in late 1979, in late 1980, and als in 1981, mean first clutch sizes were 3.33, 3.27 and 3.74 respectively 0f22 =9.8, P<0.01). Clutch size varies in assciatin with rainfall and fd supply. The largest clutches (including the nly clutches with 5 eggs) were assciated with the perids f high rainfall, large fd supply, and successful breeding, in late 1979 and The ppulatin was als studied in 1978 by Bag and Grant (1984) when there were three breeding perids. If I include these episdes in an analysis f clutch size variatin amng the pre-1976 females I find a significant crrelatin between the cumulative rainfall in each year and the mean clutch size in the breeding perid (rs=0.96, N=8, P < 0.05). Hatching synchrny. Abut half the nests visited when the ldest chick was 0-1 days ld had at least ne egg which later hatched, and this prprtin did nt vary amng years (Z22 = 0.28, P > 0.9, Table 2). This suggests that hatching is typically spread ver abut 24 h, which is highly synchrnus (Clark and Wilsn 1981). There is almst n evidence fr facultative asynchrny in pr years. In nly ne nest, in late 1980, was an egg knwn t have hatched mre than 48 h after the first egg hatched. Variatin in weight amng chicks at first weighing actually appeared t be lwer in the pr year f 1980 (Table 2) when facultative asychrny might have been mst expected. Brd reductin. Every nest but ne that yielded fledglings in the perids f lw reprductive success (early 1979 and all f 1980) fledged fewer yung then were hatched, i.e. there was brd reductin (Fig. 3). Nne f the yung fledged at these times survived fr mre than ne mnth after fledging. Sme brd reductin was als bserved in 1981, when fledging success was high (Table 3). Brd reductin ccurred mre cmmnly amng lder nestlings, when fd requirements wuld be greatest. Only a few entire brds died befre the chicks were 4 days ld. Amng 13 nests suffering partial brd reductin when the chicks were at least 4 days ld, 12 lst the chick which was smallest at hatching, and nly ne lst a larger chick (P< 0.01, binmial test). Hwever, lss f very yung nestlings (1-3 days ld) did nt appear t be assciated with size at hatching (frm tw nests the smallest at hatching died, frm 4 thers ne f the larger nes died). Grwth. Chicks grew mre slwly in the pr year f 1980 (Table 4; Fig. 4) than in the ther tw years. In 1980 bdy weight at day 7-8 suffered a greater prprtinal decrease than beak depth, as shwn by the differences f the lgarithmic values (Table 5). The primaries shwed the greatest prprtinal decrease; they grw late in develpment (Bag 1984), and are therefre mre greatly affected by malnutritin in lder nestlings. Chick grwth rates were much mre variable in 1980 than in the ther years (e.g. Fig. 4). Sme chicks were able t maintain clse-t-nrmal grwth rates in Others survived lng perids with stunted grwth (Fig. 4) but nne f these subsequently increased their grwth rates. All chicks grwing abnrmally failed t survive fr lng, althugh a few did fledge in late Grwth in healthy chicks has a strng genetic cmpnent (Price and Grant 1985). This can be shwn by calculating the cvariance between chick and mid-parent, which prvides an estimate f half the genetic cvariance between the adult and juvenile character (Falcner 1981). The regressin f chick values n mid-parent values, as the rati

5 415 Table 5, Measurement means_+ standard deviatins fr chicks at age 7-8 days, and the difference between the average values fr 1979 and 1981, and F statistic is based n a/-way ANOVA amng years. **P < 0.01 Late Late 1981 Differ- F ence 3 9 ~/Welght lg g +_0.05 _+0.12 _+0.07 Beak depth, lg mm _+0.06 _+0.tl -t th Primary, lg mm Sample size ** ** "* Table 6. Regressins f ffspring measurements (family means) at age 7-8 days n mid-parent measurements. Sample size gives the number f chicks, with the number f families in parentheses. * P < 0.05, **P< /~_+s.e. /~+s.e. f+s.e. Weight, 0.83_+0.33* _+0.26"* lg g Beak depth, 0.51 _+0.16"* 0.34_ _+0.14"* lg mm Sample size 52 (32) 34 (13) 160 (50) f the cvariance t the variance f the mid-parent values, prvides a simple way f visualizing changes in the cvariance between ffspring and parents in assciatin with starvatin (Table 6). In particular, a negative cvariance wuld indicate a negative gentype-envirnraent cvariance during perids f heavy mrtality, i.e, that the genetically larger chicks were actually smaller than the genetically smaller chicks. In 1980 the cvariances fr bth beak depth and weight were apparently reduced, and fr weight it was (nnsignificantly) negative (Table 6). Hwever, increased variance and reduced sample sizes in 1980 led t large standard errrs fr the estimates f the regressin cefficients, and differences in the regressin cefficients amng years are nt significant (ANCOVA, F (2,90)=0.59, P=0.56 fr beak depth; F(2,90)= 1.69, P=0.19 fr weight). Discussin This ppulatin and Darwin's Finches in general, have reduced grwth rates (Bag 1984), and lnger nestling perids (Grant and Grant 1980; Bag and Grant 1984) than finches in seasnal temperate envirnments. Tlais suite f characters may be adaptively related t a predatr-free envirnment in the Galfipags (Bag and Grant 1984). Parents are able t raise mre, slwly-grwing chicks by feeding each chick less fd per day and, in the absence f nest predatin the resulting lnger nestling perid is nt detrimental. The quantity f fd individual chicks receive may have given rise t the slwer grwth in Darwin's Finches, but this des nt imply that grwth rates wuld immediately respnd t changing fd abundance. Rates f tissue maturatin will be affected by persistent selectin pressures fr slwer grwth, and this may limit any flexibitility in the grwth rates f sme characters (Ricklefs 1979). Selectin pressures arising frm sib-sib cmpetitin may als affect grwth rates (Werschkul and Jacksn 1981; Ricklefs 1982). Sibs that grw t a genetically large size, and hence large in the nest als under gd cnditins (Table 6) may be at a metablic disadvantage when fd is shrt, prviding parents d nt actively discriminate in their favr when feeding the brd, as seems t be generally the case in birds (Richter 1984). If the reduced cvariance between chick and adult weight in 1980 was real (Table 6), the genetically larger chicks were starving fiirst. There is ther evidence fr selectin against large chick size, particularly after fledging (Price and Grant 1984). Genetically larger chicks seem t shw sme lack f flexibility in their grwth schedule, at least fr skeletal characters such as beak depth (Tables 4-6) and their apparent starvatin befre smaller chicks may be a cnsequence f such unaltered grwth. A failure f larger chicks t reduce their grwth rate and thereby reduce energy requirements culd arise ut f strng selectin fr rapid maturatin at ther times, which may be as imprtant as relative size in sib-sib cmpetitin. A reductin in the rate f increase in mass is an inevitable cnsequence f fd shrtage. In sme species which suffer large and temprary fd shrtages, such as Huse Martins, Delichn urbica, grwth delays have been shwn t be accmpanied by metablism f previusly stred fat (Bryant 1975; O'Cnnr 1977). Darwin's finches d nt stre fat. As fd supplies are steadily declining during times f starvatin, with any increase due t rain likely t be delayed by several days, selectin pressures arising ut f temprary fd shrtages are prbably absent. Rainfall and/r fd abundance at the time f egg-frmatin may be a reasnable indicatr f cnditins fr raising nestlings, and this is the situatin under which clutch size is expected t vary adaptively (O'Cnnr 1977). A crrelatin between rainfall and clutch size is bserved bth in this species, and its sympatric cngener, the Cactus Finch, Gespiza scandens (Millingtn and Grant 1984). Hwever, judging frm the number f chicks raised during pr cnditins, the ptimum clutch size wuld have ften been less than tw; and there was n advantage t laying eggs at all in three f the breeding perids, when nne f the raised yung survived fr mre than a mnth. In cntrast the actual clutch size, at least fr experienced females, never fell belw three eggs (Table 1). The large disparity between bserved and ptimal clutch sizes suggests that the assciatin f clutch size with envirnmental cnditins may simply be a cnsequence f the cnditins at the time f egg frmatin. Females presumable lay clutches even when cnditins are pr because smetimes rain falls again and fd cnditins imprve prir t the hatching and fledging f chicks. A given level f hatching asynchrny culd always be present in a ppulatin, with eliminatin f late hatching chicks nly ccurring if insufficient fd is being brught t the nest (Lack 1968; Hwe 1976, 1978; Hahn 1981; Zach 1982; Richer 1984). Hwever, the degree f synchrny can als vary with envirnmental cnditins. Tits, Parus spp., ften shw hatching asynchrny f secnd brds at a time when fd supply is declining (Gibb 1950). Since their first brds hatch synchrnusly, it is prbably disadvantageus fr parents t hatch brds asynchrnusly

6 416 when fd is abundant: smaller chicks may suffer frm crwding by their larger sibs (Clark and Wilsn 1981). Similarly, the high degree f synchrny in Darwin's finches may prevent lss f late hatching chicks. Facultative asynchrny under pr cnditins may nt have evlved because there is a very restrictive set f cnditins under which less than a cmplete brd culd be successfully raised, and survive t breed. Such cnditins, such as additinal rain sn after fledging, have yet t be bserved n Daphne (e.g. Fig. 1). Indeed, there was n advantage t raising even a single chick during three perids f fd shrtage in this study, since the few fledglings which reached independence died sn after. In these cnditins hatching synchrny culd be selectively favred as a means f ensuring rapid whle-brd lss. T summarise, in the Gal/tpags envirnment, selectin pressures fr the reprductive flexibility t raise ne r a few yung in pr cnditins may be small r absent. Further, any such selectin, particularly as it affects grwth rates and hatching synchrny, may be ppsed by selectin pressures perating when reprductive success is high. Recruits int the breeding ppulatin have nly cme frm perids f relatively high verall breeding success. Acknwedgements. I thank Ayse Unal and Spike Millingtn fr help in the field, and Dave Andersn, Lisle Gibbs, Peter Grant and Dlph Schluter fr cmments n the manuscript. The research was funded by NSF grants DEB and DEB t P.R. Grant, with sme additinal supprt frm the Chapman Fund. I thank the Direcci6n General de Desarrll Frestal, Quit, and the Natinal Park Service, Galfipags fr permissin t wrk n Daphne, and the Charles Darwin Research Statin fr lgistic supprt. References Bag PT (1984) Grwth and allmetry f Darwin's finches (Gespiza) n Isla Daphne Majr, Galfipags. J Zl (Lndn) 204: Bag PT, Grant PR (1984) Darwin's Finches (Gespiza) n Isla Daphne Majr, Galfipags: breeding and feeding eclgy in a climatically variable envirnment. Ecl Mngr 54: Bryant DM (1975) Breeding bilgy f Huse martins Delichn urbica in relatin t aerial insect abundance. Ibis 117 : Bryant DM (1978) Establishment f weight hierarchies in the brds f huse martins Delichn urbica. Ibis 120:16-26 Burley N (1980) Clutch size and clutch verlap: alternative and cmplementary reprductive tactics. Am Nat 115 : Clark AB, Wilsn DS (1981) Avian breeding adaptatins: hatching asynchrny, brd reductin and nest failure. Quart Rev Bil 56: Falcner DS (198i) Intrductin t quantitative genetics. 2nd editin. Lngman, New Yrk Gibb J (1950) The breeding bilgy f the Great and Blue titmice. Ibis 92 : Gibbs HL, Grant PR, Weiland J (1984) Breeding f Darwin's finches at an unusually early age in an E1 Nin year. Auk 101 : Grant PR, Bag PT (1980) Rainfall n the Gal/tpags and the demgraphy f Darwin's finches. Auk 97: Grant PR, Grant BR (1980) The breeding and feeding characteristics f Darwin's finches n Isla Genvesa, Galfipags. Ecl Mngr 50: Hahn DC (1981) Asynchrnus hatching in the Laughing Gull: cutting lsses and reducing rivalry. Anita Behav 29: Hrsfall JA (1984) Fd supply and egg mass variatin in the Eurpean Ct. Eclgy 65: Hwe HF (1976) Egg size, hatching asynchrny, sex, and brd reductin in the Cmmn Grackle. Eclgy 57: Hwe HF (1978) Initial investment, dutch size, and brd reductin in the Cmmn Grackle (Quiscalus quiscula L.). Eclgy 59: Lack D (1968) Eclgical adaptatin fr breeding in birds. Methuen, Lndn, England Millingtn SJ, Grant PR (1983) Feeding eclgy and territriality f the Cactus Finch Gespiza scandens n Isla Daphne Majr, Galfipags. Oeclgia (Berlin) 58:76-83 Millingtn S J, Grant PR (1984) The breeding eclgy f the Cactus Finch Gespiza scandens n Isla Daphne Majr, Galfipags. Ardea 72: O'Cnnr RJ (1977) Grwth strategies in nestling passerines. Living Bird 16 : Price TD (1984 a) The evlutin f sexual size dimrphism in Darwin's fisnches. Am Nat 123: Price TD (1984b) Sexual selectin n bdy size, plumage, and territry variables in a ppulatn f Darwin's finches. Evlutin 38: Price TD, Grant PR (1984) Life histry traits and natural selectin fr small bdy size in a ppulatin f Darwin's finches. Evlutin 38: Price TD, Grant PR (1985) The evlutin f ntgeny in Darwin's finches: a quantitative genetic apprach. Am Nat 125: Richter WE (1984) Nestling survival and grwth in the Yellwheaded blackbird Xanthcephalus xanthcephalus. Eclgy 65 : Ricklefs RE (1965) Brd reductin in the Curve-billed Thrasher. Cndr 67: Ricklefs RE (1968) Patterns f grwth in birds. 110 : Rickl&s RE (1976) Grwth rates f birds in the humid New Wrld trpics. Ibis 118: Ricklefs RE (1979) Adaptatin, cnstraint, and cmprmise in avian pstnatal develpment. Bil Rev 54: Ricklefs RE (1982) Sme cnsideratins n sibling cmpetitin and avian grwth rates. Auk 99: Werschkul DB, Jacksn JA (1979) Sibling cmpetitin and avian grwth rates. Ibis 121: Zach R (1982) Hatching asynchrny, egg size, grwth and fledging in tree swallws. Auk 99 : Received August 1, 1984