does prior breeding experience improve reproductive success in collared flycatcher females?
TRANSCRIPT
Oecologia (2003) 134:78–81DOI 10.1007/s00442-002-1099-x
P O P U L A T I O N E C O L O G Y
Mariusz Cicho�
Does prior breeding experience improve reproductive successin collared flycatcher females?
Received: 11 March 2002 / Accepted: 6 October 2002 / Published online: 29 October 2002� Springer-Verlag 2002
Abstract In order to assess the importance of breedingexperience in shaping age-specific patterns of reproduc-tion an experiment on collared flycatchers (Ficedulaalbicollis) was performed. Experimental females wererestrained from reproduction by destroying their nestsafter prolonged incubation. Unmanipulated females wereused as controls. Only young, first time breeders wereconsidered. A year after manipulation, experimental(inexperienced) and control (experienced) females werecompared for a number of breeding traits. They did notdiffer with respect to laying date, clutch size, number offledglings and number of local recruits. However,fledglings from broods of inexperienced females weresignificantly lighter and had shorter tarsi. This suggeststhat prior experience of feeding young may indeedinfluence the future reproductive performance in altricialbird species such as the collared flycatcher.
Keywords Breeding performance · Reproduction costs ·Age-specific reproduction · Foraging skills · Ficedulaalbicollis
Introduction
In iteroparous organisms reproductive performance oftenincreases with age in early reproductive life and decreasesat old age (e.g. Charlesworth 1994; Cicho� 2001). Manyhypotheses, involving factors such as differential survival,delayed breeding, breeding experience, foraging abilityand reproductive effort, have been proposed to explainthis pattern in birds, but experimental tests have yieldedequivocal results [see Forslund and P�rt (1995) forreview]. One of the hypotheses most frequently proposedto explain initial improvement in reproductive perfor-
mance posits that the improvement results from increas-ing experience, making breeding activities like incubationand brood care more efficient. Previous studies attempt-ing to test this hypothesis were mainly observational andin fact have failed to reveal any consistent trends.Although these studies controlled for age effect theyhave generally neglected individual variation in intrinsicquality and the potentially confounding effect of pastreproduction (Forslund and P�rt 1995; P�rt 1995; Camand Monnat 2000). For example, the study in whichyoung females of collared flycatcher chose not to breedwhile some bred successfully after experimental distur-bance did not support the hypothesis of breeding expe-rience. It was suggested instead that the costs of pastreproduction or differences in individual quality, ratherthan prior experience, play major roles in explainingobserved variation in future reproductive success (P�rt1995).
Since both reproductive performance and experienceare related to age, the benefits of breeding experiencecannot be reaped without paying the cost of performingreproduction. It is generally accepted that one of the mostimportant factors affecting reproductive performance isthe cost of reproduction (sensu Williams 1966), usuallypaid in terms of subsequent survival probability andfecundity [see Lind�n and Møller (1989); Stearns (1992)for reviews]. Thus, breeding experience and costs ofreproduction inevitably interact to mould reproductivepatterns: breeding experience is expected to increasefuture reproductive performance, while reproductive costsshould decrease it. Therefore, in order to investigate thebreeding experience hypothesis, one will have to look forpotential confounding effects of reproductive costs andtheir relative importance for the observed age-specificpatterns of reproduction.
To date, no experimental evidence exists showing thatbreeding experience improves reproductive performance.As the previous studies were mainly non-manipulativeany potential effects of breeding experience could beobscured by differences in individual quality and moreimportantly by differences in previous reproductive effort.
M. Cicho� ())Institute of Environmental Sciences, Jagiellonian University,Gronostajowa 3, 30–387 Krak�w, Polande-mail: [email protected].: +48-12-2690944 ext. 123Fax: +48-12-2690927
In contrast the present paper was designed to control forboth these confounding factors in order to reveal thepotential remaining effect of breeding experience per se.To my knowledge the current paper is the first in whichthe amount of breeding experience gained by individualswas manipulated, with random assignment of females tothe experimental groups. In order to prevent birds frombeing experienced, here some females were exemptedfrom parental care by removing the whole nest afterprolonged incubation. Then, a year after the experiment, Icompared subsequent reproductive performance of fe-males that had been allowed to breed successfully(experienced, unaffected females) with those that hadbeen experimentally prevented from completing repro-duction (inexperienced, restrained females). If breedingexperience is an important factor affecting future perfor-mance, experienced birds should enjoy higher reproduc-tive success in the year after the experiment.
Materials and methodsStudy area and methods
The study was conducted in a nest-box population of collaredflycatchers on the island of Gotland, Sweden (for details about thestudy area see Gustafsson 1989). The experimental manipulation ofbreeding experience was performed in 2 years, 1993 and 1994, withthe experimental birds being followed for 2 subsequent years aftermanipulation. The collared flycatcher is a small (»13 g) migratorypasserine bird species breeding mainly in eastern and centralEurope and wintering in southern and central Africa. It nests innatural tree cavities, but prefers nest boxes when provided. TheGotland population is characterized by fairly restricted dispersal ofadults and young (P�rt and Gustafsson 1989; P�rt 1991). Thus, it ispossible to follow individuals and measure their reproductivesuccess throughout their lifetime. Many individuals are ringed asnestlings or yearlings and therefore are of known age.
In the present study, nest boxes were inspected every 5 days tolook for new nests. In order to determine age, females werecaptured in the middle of the incubation period when they were alsoweighed. Age was determined on the basis of ring number when thefemale was already ringed or classified as yearling or older forunringed females, according to the shape of the primary coverts andthe colour of the upper mandible [Karlsson et al. (1986); ageingtechnique is highly reliable for experienced person]. Since theeffect of breeding experience should be primarily expected amongfirst time breeding birds, only yearling females were included.Females were assigned to experimental (n=26 in 1993 and n=27 in1994) or control (n=18 in 1993 and n=15 in 1994) groups, matchedin terms of egg laying date. (The experimental group was over-represented because I expected unsuccessful birds to disperse forlonger distances, away from the nest box area). In order to preventexperimental females from becoming experienced in rearing young,their nests were destroyed after prolonged incubation. Prolongationof incubation served two reasons; first to reduce the probability ofrenesting (no renesting attempt was in fact observed) and second toreduce the difference in reproductive costs encountered byexperienced and inexperienced birds that might be essential indetecting any effects of breeding experience as emphasized in theintroduction section. In collared flycatchers, only females incubate.They usually start incubation after completing the clutch and theincubation period is 14 days. In the present experiment theincubation period was extended by 7 days. Eggs in the experimentalnests were exchanged with the same number of artificial clay eggsafter the seventh day of incubation, and after 21 days from clutchcompletion the whole nest was removed from the nestbox.
To assess future reproductive performance, both experimentaland control females were monitored in the next year following theexperiment. The year after the experiment the adults were trappedin the nest boxes while feeding young. Tarsus length and body massof their fledglings were recorded 13 days after hatching, i.e. justbefore fledgling. Out of 86 females involved in the experiment inboth years 36 were recorded 1 year after the experiment. Out ofthese 36 nests three were deserted during the early stage of breedingand in three other nests nestlings were not measured.
The breeding variables were analysed by means of two-wayANOVA to account for year effects (the manipulation wasperformed in 2 years), after assuring assumptions of parametrictests (Sokal and Rohlf 1995). Analyses of differences in fledglingbody weight and tarsus length were based on measurements ofindividual nestlings. Thus, in order to account for within-nestvariation individual nests were nested in experimental treatment inthe three-way nested ANOVA (year and experimental treatment asindependent variables and nest ID as factor nested in experimentaltreatment). In all analyses, except for clutch size and fledglingnumber, the yearly differences in breeding variables were signif-icant (P<0.01), but no interactions appeared significant (P>0.3), sothey were not presented in the Results section. Differences in thenumber of recruits and dispersal distances were assessed with aMann-Whitney test with both years pulled together after standard-ization by subtraction of the mean value for each year.
Results
The analysis of breeding data from the year aftermanipulation showed that females experimentally ex-empted from reproduction (inexperienced) raised fledg-lings of lower body mass (three-way nested ANOVA:F1,150=7.71, P=0.006, Fig. 1) and with smaller tarsi (three-way nested ANOVA: F1,150=31.09, P<0.001, Fig. 1) thanthe control (experienced) females. There was no signif-icant difference between experimental and control fe-males with respect to laying date (two-way ANOVA:F1,31=1.96, P=0.17, power=0.18, Fig. 1), clutch size (two-way ANOVA: F1,31=0.59, P=0.45, power=0.1, Fig. 1),number of fledged young (two-way ANOVA: F1,29=0.39,P=0.54, power=0.15, Fig. 1) or number of recruits (Mann-Whitney, data pooled between years: z=0.92, P=0.36,power=0.38, Fig. 1).
Brood failure as simulated by the present experimentmay be considered also as some kind of experience andpotentially may make birds more inclined to disperse tosome more favourable area. This should become reflectedin their probability of returning to the breeding grounds,which in this case includes both mortality and emigrationfrom the study area. I did not find any significantdifferences in return probability (c2-test: c2=1.58, df=1,P=0.21, 38% vs. 0.52%, experimental and controlfemales, respectively). Moreover, experimental femalesdid not disperse over significantly longer distances withinthe study area as compared to control females(1,469 m€1,776 SD vs. 1,041 m€1,582 SD; data pooledbetween years; Mann-Witney z=–0.83, P=0.41, pow-er=0.25).
Discussion
In collared flycatchers, fecundity is age-related. Itincreases during the first 2 years and slightly decreases
79
later in life (Gustafsson and P�rt 1990; Cicho� M,unpublished data). The present results suggest that theincrease in reproductive performance early in life may bedue to experience successively obtained from eachbreeding episode. In first time breeding females, thoserestrained from reproduction did not gain the reproductiveexperience granted those allowed to breed successfully
(experienced controls). A comparison of reproductiveparameters between these two groups next year aftermanipulation revealed a positive effect of breedingexperience on fledgling body weight and tarsus length.Inexperienced birds raised smaller fledglings than expe-rienced birds. Differences in other reproductive traitsappeared non-significant. Thus, the effect of breeding
Fig. 1 Breeding performance of experimental inexperienced andexperienced control collared flycatcher females in a year followingthe experiment (least square means€SE obtained from two-way
nested ANOVA). Body weight and tarsus length were measured innestlings at 12 days after hatching. Numbers above bars denotenumber of nest in each group
80
experience was only shown for the breeding component(namely nestling rearing) that had actually been manip-ulated. (Note, however, that the power of the statisticalcomparisons which appeared non-significant was low,possibly due to relatively small sample size and largevariance observed in the considered traits).
The previous studies testing the breeding experiencehypothesis were mainly observational. Their results areinconsistent and generally imply that breeding experiencein fact had little or no effect on reproductive performance(Forslund and P�rt 1995; P�rt 1995; Cam and Monnat2000). The differences in individual quality and con-founding effects of reproductive costs has been proposedto account for difficulties in revealing the potential effectsof breeding experience. For example, in P�rt’s (1995)study inexperienced females were those which afterdisturbance during mating and early nest building stagedecided not breed at all in their first year of life. Thus,inexperienced females in this study were not paying anycosts of performing reproduction, and because of that theymight actually do better than experienced ones next year.In contrast, females in my study were only partiallyexempted from reproductive costs since they had to investin egg production and prolonged incubation of the clutch.This may explain the discrepancy between P�rt’s (1995)and the present study.
The findings of the present study could be alternativelyinterpreted as a result of high costs of prolongedincubation in inexperienced females. For this to occur,costs of prolonging incubation by 1 week must have beenlarger than the costs of 2 weeks of feeding young. Thisseems very unlikely. Although the costs of incubationmay not be negligible in this species (e.g. Cicho� 2000),the main cost of breeding in altricial bird species isusually attributed to chick rearing effort (e.g. Monaghanand Nager 1997). In fact, prolongation of the incubationperiod may possibly have a positive effect when testingthe breeding experience hypothesis: the differences inreproductive costs between experimental and controlbirds became smaller allowing effects of breeding expe-rience coming into play.
The approach employed in the present experiment doesnot allow one to address directly the question of whichcomponents of the individual behaviour were affected.Age-related improvements may include, for example,foraging skills, selection of territory, nest site and mate,predator avoidance, and social dominance (Forslund andP�rt 1995). I believe that the manipulation performed inthe present study might affect foraging skills, as theexperimental birds were restrained from feeding theiryoung. Tarsus length and body weight of nestlings reflectsnutritional condition during growth in collared flycatchers(e.g. Meril� 1997). Nestlings from an experimentallyenlarged brood usually have considerably smaller tarsiand body weight in comparison with nestlings from non-manipulated broods, despite relatively high heritability ofthese traits (Meril� 1997). Lower weight and shorter tarsiof nestlings of inexperienced females suggest that theirnestlings were underfed. This means that experienced and
inexperienced females probably differed in their foragingefficiency. This leads to the conclusion that foraging forprovisioning offspring probably cannot be improvedwhile females forage only for themselves.
In conclusion, the present results suggest that breedingexperience may indeed be an important factor shapingpatterns of reproduction. This study seems to be the firstto provide experimental evidence supporting this hypoth-esis. Among other things, it may explain the successiveincrease in fecundity in early age classes, which haspuzzled ecologists for a long time. It also contributes toour understanding of why old partners seem to bepreferred in many birds. One important general implica-tion of this finding is that learning may have life historyconsequences in terms of lifetime reproductive success.
Acknowledgements I thank A. Cicho� and P. Olejniczak for helpduring the field work. Tomas P�rt, Ben Sheldon, Henk van derJeugd and Staffan Ulfstrand provided valuable comments on theearlier version of the manuscript. Financial support was providedby the Swedish Institute. The experiment complied with the ethicallaws in Sweden.
ReferencesCam E, Monnat J-Y (2000) Apparent inferiority of first-time
breeders in the kittiwake: the role of heterogeneity among ageclasses. J Anim Ecol 69:380–394
Charlesworth B (1994) Evolution in age-structured populations,2nd edn. Cambridge University Press, Cambidge
Cicho� M (2000) Costs of incubation and immunocompetence inthe collared flycatcher. Oecologia 125:453-457
Cicho� M (2001) Diversity of age-specific reproductive rates mayresult from ageing and optimal resource allocation. J Evol Biol14:180-185
Forslund P, P�rt T (1995) Age and reproduction in birds –hypotheses and tests. Trends Ecol Evol 10:374–378
Gustafsson L (1989) The collared flycatcher. In: Newton I (ed)Lifetime reproduction in birds. Academic Press, London, pp 75–88
Gustafsson L, P�rt T (1990) Acceleration of senescence in thecollared flycatcher Ficedula albicolis by reproductive costs.Nature 347:279–281
Karlsson L, Persson K, Walinder G (1986) Ageing and sexing inpied flycatchers, Ficedula hypoleuca. Var Fagelvarld 45:131–146
Lind�n M, Møller AP (1989) Cost of reproduction and covariationof life history trait in birds. Trends Ecol Evol 4:367–371
Meril� J (1997) Expression of genetic variation in body size of thecollared flycatcher under different environmental conditions.Evolution 51:526–536
Monaghan P, Nager RG (1997) Why don’t birds lay more eggs?Trends Ecol Evol 12:270–274
P�rt T (1991) Philopatry and age as a factor influencing reproduc-tive success in the collared flycatcher (Ficedula albicollis).PhD dissertation. Acta Univestatis Upsaliensis. Uppsala Uni-versity, Uppsala
P�rt T (1995) Does breeding experience explain increased repro-ductive success with age? An experiment. Proc R Soc Lond B360:113–117
P�rt T, Gustafsson L (1989) Breeding dispersal in the collaredflycatcher (Ficedula albicollis): possible causes and reproduc-tive consequences. J Anim Ecol 58:305–320
Stearns SC (1992) The evolution of life histories. Oxford Univer-sity Press, Oxford
Sokal RR, Rohlf FJ (1995) Biometry, 3rd edn. Freeman, New YorkWilliams GC (1966) Natural selection, the cost of reproduction, and
a refinement of Lack’s principle. Am Nat 100:687–690
81