The Condor9318 17-824 0 The Cooper OrnithologicIll Society 1991

EFFECTS OF BROOD SIZE AND AGE ON SURVIVAL OF FEMALE WOOD

FRANK C. ROHWER School of Forestry, Wildlife and Fisheries,Louisiana State University,Baton Rouge, LA 70803

H. W. HEUSMANN Division of Fisheriesand Wildhfe, Westborough,MA 01581

Abstract. During a 13-year study of Wood Ducks (Aix sponsa)nesting in eastern Mas- sachusetts,43.3% of nests were parasitized by conspecifics.Early seasonnests were more frequently parasitized than were late nests. First-nesting females were less frequently par- asitized than experienced breeders, partially becausefirst-nesting females nest later in the seasonwhen parasitism declines. Nest parasitism created clutch size variation that allowed us to investigatethe effect of clutch size and brood size on survival of Wood Ducks. Females that hatched large broods had a minor delay in nesting the following year. Brood size had no influence on the survival rate of female Wood Ducks. Survival rates averaged 52.8% over all year and ageclasses. Survival ratesdeclined by 6.1% per year of breedingexperience. Key words: Age-speclfi survival; brood size; clutch size; intraspecificnest parasitism; reproductiveeffort; Wood : Aix sponsa.

INTRODUCTION ments did not increase the number of fledglings, A great deal of life-history theory is devoted to so parental effort may not have been substan- relationships between patterns of mortality and tially increased. reproductive effort. Two tenets concerning pat- Patterns of age-specificsurvival rates also are terns of mortality and fecundity are central. First, expected to influence reproductive effort. De- parental survival should be inversely related to clining survival should increase reproductive ef- reproductive effort (Williams 1966, Chamov and fort as age, because residual reproduc- Krebs 1974), though the relationship may be non- tive value is declining (Pianka and Parker 1975, linear (Pianka and Parker 1975). Second, ani- Steams 1976). Studies of age-related mortality mals should show senescence(Hamilton 1966), patterns among have produced a variety of expressedas declining survival or reproductive results. Some analyses suggestthat old-aged in- efficiency in old-age classes. Both tenets have dividuals show reduced likelihood of survival important ramifications on optimal patterns of (e.g., Coulson 1984, Loery et al. 1987, Bradley reproductive effort, yet they have not been ad- et al. 1989) whereas others find little evidence equately tested, particularly for precocial birds. that survival probabilities decline with age (e.g., The relationship between parental effort and Buckland 1982, Parkin and White-Robinson parental survival in birds is best studied by ma- 1985). If senescenceis real, it may partially ex- nipulating brood size and recording effects on plain the increasein clutch size and reproductive parents’ survival and future reproduction (Al- output of older birds (Curio 1983, Pugesek 1983, erstam and Hijgstedt 1984). Brood enlargements Winkler and Walters 1983). To help understand decreasedparental return rates in some studies the factors influencing optimal reproductive ef- (Askenmo 1979, Nur 1984, Reid 1987, Dijkstra fort we examined the effect of age and prior re- et al. 1990), but not in others (Harris 1970, De productive effort on survival of female Wood Steven 1980, Rsskaft 1985, Lessells 1986, Heg- Ducks (Aix sponsa). ner and Winglield 1987, Gustafsson and Suth- erland 1988,Korpim&i 1988,Pettiforetal. 1988, METHODS Wiggins 1990). In two of the above studies (As- Data were derived from a 13-year study of fe- kenmo 1979, Korpim&i 1988), brood enlarge- males nesting in artificial nest boxes at 25 sites in eastern Massachusetts (see Heusmann and ’ ’ Received 15 February 1991. Final acceptance5 Bellville 1982). Females were captured at the August199 1. nest during incubation and were banded or had

F171 818 FRANK C. ROHWER AND H. W. HEUSMANN

prior band numbers recorded. A few birds were viva1 rate for a cohort if we had at least three initially banded during late summer. Most fe- years to observe the return of its members. males were captured as breeding adults, so anal- Nesting dates were recorded as Julian dates yses of survival are based on years since first coded so that 1 April equals day 1. To control breeding. Breeding philopatry is strong in female for variation in nesting phenology, we adjusted waterfowl (Johnsonand Grier 1988, Rohwer and each years’ hatch dates so that they matched the Anderson 1988) including Wood Ducks (Grice pattern for the data pooled over all years. Nest and Rogers 1965, Hepp et al. 1987). Only 11 dates for each year were scaled so the date by marked females (4.0%) were recorded nesting at which 10% of nests had hatched was day 45. We two different sites.This is important becauselarge matched yearly data at the tenth percentile be- scale movements would cause experienced fe- cause that is reasonably good indicator of the males to be assigned to the first-year age . initiation of nesting, but, unlike first nest dates, The sedentary nature of females means that re- is not overly influenced by sample size and the turn rates are reliable indicies of survival (Hepp earliest nesting individuals. Scaling by median et al. 1987), though they may represent under- hatch date would have produced practically the estimates of survival. same effect. We examined the influence of brood size on Statistical analyseswere based on the SAS sys- parental survival by comparing return rates of tem (SAS Institute Inc. 1985). Two and three femaleswith normal-sized broods to femaleswith way contingency tables were analyzed using log- enlarged broods resulting from the frequent oc- linear models (Proc Catmod). Analyses of cate- currence of intraspecific nest parasitism (Rohwer gorical and continuous data were analyzed using and Freeman 1989). Clutches of more than 13 logistic regression (Proc Catmod). eggshave almost always been enlarged by par- asitism (Morse and Wight 1969, Semel and Sher- RESULTS man 1986) so we used that clutch size as our criterion for identifying parasitized nests. We as- EFFECTS OF REPRODUCTIVE EFFORT sumed that the occurrence and intensity of par- The frequency distribution of returning versus asitism would be independent of host attributes nonreturning female parents was independent of (i.e., inherent survival probability). For 256 fe- brood size (Table 1). That result was consistent males we recorded clutch size for two or more for any grouping of brood sizes. We also ex- years, so we could examine the frequency of re- panded the contingency table analysesto include peatedparasitism for individual females.We used an agedimension, with classesof 1-year, 2-years, observed rates of nest parasitism in a binomial and 2 3-years of breeding experience. The inter- expansion to generate expected frequencies of action of brood size, breeding age, and survival repeated parasitism over the years of observa- was not significant (x2 = 4.2, P > 0.10). If in- tion. cubation or brood care were costly, then we might Our measure of brood size in all analyses was expect female survival to systematically decline the number of young leaving the nest, which pos- with increased brood size. Contrary to this pre- itively correlated with the number of young that diction, females did not show a decline in sur- survived (Heusmann 1972). Initial brood size vival associatedwith increased brood size (rang- was correlated with clutch size (r = 0.81, n = ing from 8 to 22 ducklings); in fact, the slope of 90 1, P -c 0.0001); thus, our analysesevaluate the the survival vs. brood size relationship was pos- combined effect on parents of incubating itive, though non-significant (ANCOVA weight- and caring for broods. However, we cannot as- ed by brood size samples, F = 0.06, P > 0.10). sessthe effectson parents of producing eggs. The interaction term in the above analysis was Our survival analysesused a methodology out- also nonsignificant (F = 0.79, P > O.lO), indi- lined in Seber (1982) and Loery et al. (1987). cating that neither yearling nor experienced fe- Analyses were cohort-based, and involved 1,02 1 males were negatively influenced by brood size. captures and recaptures of 471 female Wood Experienced females nested earlier than youn- Ducks. We marked birds for two years before we ger females.Nests of second-yearfemales hatched assigned new captures to the first-breeding age 13.8 days earlier than first-year females; simi- category. We only calculated an age-specificsur- larly, as birds entered the 3-year and ?4-year WOOD DUCK SURVIVAL 819

TABLE 1. Fate of female Wood Ducks attending to broods hatching from parasitized nests (> 13 broodsof differentsizes. eggs)resulted in a non-significant regressionbe- tween brood size and change in hatch date the Initial brood size following year (Fig. 1, hatch date in year X + 1 113 14-16 17-19 >20 = 0.5 (brood size in year x) - 3.9 days, F,,9, = Disappeared 299 74 1.5, P > 0.10). Returned 419 81 :I ::

G = 4.2 P > 0.10. NEST PARASITISM Nest parasitism occurred in 43.3% of nests, but classestheir hatch dates were 2.4 and 1.4 days was not equally distributed among females. First- earlier than each prior year. These hatch dates year females were parasitized less frequently were significantly different (F3,574= 52.6, P c (30.8%) than were older females (56.0%) (G = O.OOOl),but the Tukey-Kramer multiple range 60.8, P < 0.0001). Parasitism rates among ex- test distinguished only two classesof birds with perienced birds did not vary with age since first respect to hatch date, namely first-year and ex- breeding (G = 0.6, P > 0.10). Logistic regression perienced birds. indicated that nest date had a strong influence We also examined the relationship between on rates of parasitism, with early nests more fre- brood size and nesting date the next year. We quently parasitized than late-season nests (x2 = controlled for individual variation in nest dates 4 1.7, P < 0.0001). First-year and experienced by examining change in hatch dates for individ- females had similar rates of seasonaldeclines in ual females between years X and X + 1. We cor- parasitism (hatch date.age interaction, x2 = 2.2, rected for the age-related changesin nesting time P > 0.10). However, the age classterm was sig- by subtracting the mean advance in hatch date nificant (x2 = 4.0, P < 0.05) which means that for that age class (given above) from the indi- experienced breedershad a higher probability of vidual hatch date. Increasing brood size by one being parasitized than first-year females, even duckling causedless than a one day delay in hatch after controlling for the earlier nesting of expe- date the following year (Fig. 1, hatch date in year rienced breeders. X + 1 = 0.9 (brood size in year X) - 11.4 days, Parasitism was nearly equally distributed F 1.224= 11.5, P -c 0.001). Restricting the analysis among females with multiple nest records(Table

;I” . l . I ...... l l -. * . 25-

O-

. -25- . i i * . . . . -50 ’ . . . 1 ’ . ’ . 1 . ’ ’ ’ 1 . * ’ ’ ’ . ’ ’ ’ ’ . . ’ - 0 5 10 15 20 25 30

Brood Size in Year X FIGURE 1. The relationship between brood size and hatch date in the following year. Change in hatch date equals hatch date in year X + 1, minus hatch date in year X, plus a correction for age effects (see text). A negative value for change in hatch date indicates earlier nesting, positive means later nesting. 820 FRANK C. ROHWER ANDH. W. HEUSMANN

TABLE 2. Actual and expectednumbers of femaleWood Ducksshowing multiple parasitismevents.

Frequencyof parasiticevents 2 nests/female 3 nests/female 4 festslfemale per individual Actual ExDected Actual ExDected Actual Exvected

0 46 43.5 16 8.2 4 1.7 : :: 23.774.8 17 24.422.5 68 11.27.3

3 - - 11 5.9 7.2 4 - - - - 1.6 x2= 4.14 x2 = 15.6 x2 = 5.6 P> 0.1 P < 0.05 P> 0.1

2). Only the set of females where we had records uals over the same one year interval. Such anal- of three nests showed deviations from the ex- yses, of necessity, combine data from different pected distribution of nests being parasitized. In cohorts. Complete data were available for seven that case, greater than expected numbers of fe- yearly comparisons;the data run diagonally from males were repeatedly parasitized or consistently top-right to bottom-left in Table 3. For example, escapedparasitism. one comparison would be the survival of novice females in 1976, second-year females from the SURVIVAL 1975 cohort, 3-year-females from 1974, and so Wood Duck survival declined with age and on. Yearly regressionsof survival produced five breeding experience (Table 3). Regressions of negative and two positive slopes,with two of the survival upon years of breeding experience negative slopes significant (both P < 0.05, re- (weighted by numbers of birds) produced four gressionsweighted by numbers of birds). negative slopesand three positive slopes (Table 3). The regression of average survival versus DISCUSSION breeding experience projected that yearly sur- Our data on return rates suggestthat incubation vival rate drops by 6.1% each year (P < 0.05). of enlarged clutches and care of enlarged broods We calculated survival for five breeding years, do not have negative consequencesfor female because there were few survivors past that age Wood Ducks. Similar results have been obtained and survival estimates became highly variable for other waterfowl (Dow and Fredga 1984, Les- in the older age classes(Table 3). sells 1986, Rockwell et al. 1987, Hepp et al. 1990). Cohort-based analyses can be faulted because Such results are not unexpected, because the they compare survival estimates obtained under young of waterfowl find their own food and are different conditions. Survival estimates for wa- quite independent at an early age. Descriptions terfowl are known to fluctuate on a yearly basis of waterfowl behavior show that parental in- (Nichols et al. 1984). One way to avoid those vestment increaseswith brood size in some spe- problems is to compare different aged individ- cies (Afton 1983, Lessells 1987, Schindler and

TABLE 3. Age-specificsurvival ratesfor cohortsof Wood Ducks. Agesare from the first known breeding attempt.

Yearof first breeding wnodAF 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 z;g

l-2 .50 .61 .81 .54 .53 .67 .66 .52 .69 .67 .73 .63 * 0.09 2-3 .53 .53 .73 .54 .39 .77 .59 .50 .48 .43 - .55 + 0.12 3-4 .56 .78 .44 .43 .77 .50 .41 .56 .40 - - .54 + 0.15 4-5 0.0 .57 .71 1.0 .80 .30 .45 .40 - - - .53 * 0.31 5-6 0.0 .75 .60 .33 .25 .33 0.0 - - - - .32 + 0.28 Slope’ -.081 .026 -.074 .OOl .009 -.106 -.118 - - - - -.061 P >O.l >O.l >O.l >O.l >O.l >O.l co.05 - - - - co.05 ’ Slopesfrom regressionweighted by numberof femalesor numberof survival estimates. WOOD DUCK SURVIVAL 821

Lamprecht 1987, Sedinger and Raveling 1990), of parasitic laying early in the season,and is not but not in others (Lazarus and Inglis 1978, Scott simply due to a greater time of exposure (i.e., 1980, Guinn and Batt 1985). It is unclear, how- both laying and incubation) of early nests (Lank ever, whether changesin the time spent in alert et al. 1989, Hepp et al. 1990). After controlling posturesor other such inactive and nonrisky pa- for date effects,we found a weak, but significant, rental behaviors affect parental survival in wa- relationship between host age (categorizedas ex- terfowl and thereby influence optimal fecundity perienced vs. inexperienced breeders) and the levels. frequency of parasitism. Older females may ex- Increased effort in one year could manifest it- perience greaternest parasitism due to natal phil- self in poorer reproduction in later years (e.g., opatry of their offspring (Andersson 1984). Al- Roskaft 1985). Lessells(1986) noted that Canada ternatively, the age effect may simply be a Geese (Branta canadensis)with enlarged broods dominance effect, such that when two females delayed laying the following spring. Such delays are both using the same nest box the older (dom- could negatively influence clutch size and sur- inant) female eventually incubates the jointly vival of young (Rohwer, in press). After con- produced clutch. trolling for individual variation (Batt and Prince Repeated observations on individuals show 1979) and parental age (Grice and Rogers 1965, that parasitism is relatively equally distributed Afton 1984, Dow and Fredga 1984), we found a among females. Only in the set of females for minor delay in nesting associated with larger which we had three nest records was there a de- brood sizes. This effect was largely the result of viation from the expected distribution of mul- females with small clutchesin one year initiating tiple parasitisms. In that case, a greater than ex- nesting substantially earlier in the following year. pected number of females experienced no Such females may lose or abandon their brood, parasitism or frequent parasitism. This may be so they nest earlier because of reduced parental a nest site effect. Females show consistency in effort. A more parsimonious explanation is that nest site, and some sites are much more suscep- the large advance in nesting dates was a result of tible to parasitism than other sites (Semel et al. comparing hatch dates of replacement nests (re- 1988). nests have small clutches) with hatch dates of Much ornithological literature suggeststhat first-nestsin the subsequentyear. Nest dateswere survival rates are not significantly related to age not influenced by prior brood sizesfor the sample (Richdale 1957,Ludwig 1967,KadlecandDrury of birds that had parasitically enlarged clutches. 1968, Henny and Wight 1969, Potts 1969, Bul- That is the sample of birds where parental quality mer and Perrins 1973, Richdale and Warham and brood size were probably uncoupled by par- 1973, Coulson and Horobin 1976, Buckland asitism. Wood Ducks, unlike Canada Geese, do 1982, Dow and Fredga 1984, Woolfendon and not extend parental care past fledging, so a strong Fitzpatrick 1984, Parkin and White-Robinson association between brood size and subsequent 1985, No1 and Smith 1987, Gibbs and Grant nesting date would be surprising. 1987, Curry and Grant 1989). However, our data Natural and experimental manipulations of on Wood Ducks suggestthat survival declines in brood size in waterfowl show that brood size older ageclasses, as reported in some recent stud- does not influence duckling survival (Morse and ies (Coulson 1984, Dunnet and Ollason 1978, Wight 1969, Heusmann 1972, Clawson et al. Ainley and DeMaster 1980, Loery et al. 1987, 1979,DowandFredga 1984, Rohwer 1985, Les- Bradley et al. 1989, Aebischerand Coulson 1990). sells 1986, Rockwell et al. 1987). Such results The main group of birds that show declining age- coupled with the independence of parental sur- specific survival are long-lived , where vival from brood size suggestthat post-hatching small declines in age-specific survival could be factors have little influence on optimal repro- an important selective force shaping age-related ductive output in waterfowl. variation in reproductive effort (Charlesworth Late-season nests were less frequently parasit- 1980). Unfortunately, it is questionable whether ized than early nests in our population and in declining age-specificsurvival represents an in- several other studies (Morse and Wight 1969, trinsic decline in survival (Pianka and Parker Clawson et al. 1979, Dow and Fredga 1984, Lank 1975). For example, in California (Lams et al. 1989). We assume that the higher level of calijixnicus) the age-specificdecline in survival parasitism of early nesters reflects higher levels (Pugesek 1983) may be extrinsic; that is, caused 822 FRANK C. ROHWER AND H. W. HEUSMANN

by their high reproductive effort. Old-aged birds COULSON,J. C. 1984. The population dynamics of that failed early in the reproductive stagehad the the Eider Duck Somateria mollissima and evi- dence of extensive non-breeding by adult ducks. same survival rates as middle-aged birds (Pu- Ibis 126:525-543. gesekand Diem 1990). We believe the effectsof COUJSON, J. C., AND J. HOROBIN. 1976. The influence changing age-specific survival on reproductive of age on the breeding biology and survival of the patterns are not clear and should be the subject Arctic Tern Sternaparadisaea.J. Zool. Lond. 178: of further study. 247-260. CURIO,E. 1983. Why do young birds reproduce less well? Ibis 125:400-404. ACKNOWLEDGEMENTS CURRY,R. L., AND P. R. GRANT. 1989. Demography of the cooperatively breeding Galapagos Mock- We especially thank D. Grice and R. Bellville, who ingbird, Nesomimus parvulus, in a climatically were involved in collecting field data. A. D. Afton, J. variable environment. J. Anim. Ecol. 58:441463. E. Gates, J. L. Hoogland, and J. Sedinger provided DE STEVEN,D. 1980. Clutch size, breeding success, helpful comments on the manuscript.The help of Shei- and parental survival in the Tree Swallow (Zri- la Bahruth Rohwer in preparing the manuscript was doprocnebicolor). 34~278-291. much appreciated. DWKSTRA,C., A. BULT,S. BULSMA,S. DAAN,T. MELIER, AND M. ZIJLSTRA. 1990. Brood size manipula- LITERATURE CITED tions in the Kestrel (Falco tinnunculus):effects on offspring and parent survival. J. Anim. Ecol. 59: AEBISCHER,N. J., ANDJ. C. COULSON.1990. Survival 269-285. of the Kittiwake in relation to sex, year, breeding Dow, H., AND S. FREDGA. 1984. Factors affecting experience and position in the colony. J. Anim. reproductive output of the Goldeneye Duck Bu- Ecol. 59:1063-1071. cephalaclan&a. J. Anim. Ecol. 53:679-692. AFTON,A. D. 1983. Male and female strategiesfor DUNNET,G. M., AND J. C. OLLASON. 1978. The es- reproduction in LesserScaup. Unpubl. Ph.D.diss., timation of survival rate in the , Fulmarus Univ. of North Dakota, Grand Forks, ND. glacialis. J. Anim. Ecol. 47:507-520. AFTON, A. D. 1984. Influence of age and time on GIBBS,H. L., AND P. R. GRANT. 1987. Adult survi- reproductive performance of female LesserScaup. vorship in Darwin’s Ground Finch (Geospiza) Auk 101:255-265. populations in a variable environment. J. Anim. AINLEY,D. G., ANDD. P. DE MASTER. 1980. Survival Ecol. 56~797-813. and mortality in a population of Ad&lie Penguins. GRICE,D., ANDJ. P. ROGERS.1965. The Wood Duck Ecology 61:522-530. in Massachusetts.MA Div. Fisheries and Game, ALERSTAM,T., AND G. HGGSTEDT. 1984. How im- P. R. Project W-19-R, Westboro, MA. portant is clutch size dependent adult mortality? GUINN, S. J. R., AND B. D. J. BATT. 1985. Activity Oikos 43:253-254. budgetsofNorthern Pintail hens:influence ofbrood ANDERSSON,M. 1984. Brood parasitism within spe- size, brood age, and date. Can. J. Zool. 63:2114- cies, p. 195-227. In C. J. Barnard [ed.], Producers 2120. and scroungers:strategies of exploitation and par- GUSTAFSSON,L., ANDW. J. SUTHERLAND.1988. The asitism. Croom Helm, London. costs of reproduction in the Collared Flycatcher ASKENMO.C. 1979. Renroductive effort and return Ficedula aibicollis.Nature 335:813-8 15. _ rate df male Pied Flycatchers.Am. Nat. 114:748- HAMILTON.W. D. 1966. The mouldine of senescence 753. by natural selection. J. Theor. Bioc 12:1245. BATT,B.D.J., AND H. H. PRINCE. 1979. Laying dates, HARRIS,M. P. 1970. Breeding ecology of the Swal- clutch size and weight of captive Mallards. low-tailed , Creagrusjiurcatus. Auk 87:215- Condor 81:35-41. 243. BRADLEY,J. S., R. D. WOOLLER,I. J. SKIRA,AND D. HEGNER,R. E., AND J. C. WINGFIELD. 1987. Effects L. SERVENTY.1989. Age-denendent survival of of brood-size manipulations on parental invest- breeding Short-tailed ShGarw&ersPufinus tenui- ment, breeding success,and reoroductive endo- rostris.J. Anim. Ecol. 58:175-188. crinology of H&se Spa&ows. &k 104:470-480. BUCKLAND,S. T. 1982. A mark-recapture survival HENNY, C. J.. AND H. M. WIGHT. 1969. An endan- analysis.J. Anim. Ecol. 51:833-847. gered Osprey population: estimates of mortality BULMER,M. G., AND C. M. PERRINS. 1973. Mortality and production. Auk 86:188-198. in the Great Tit Parus major. Ibis 115:277-28 1. HEPP,G. R., R. T. HOPPE,AND R. A. KENNAMER. 1987. CHARLESWORTH,B. 1980. Evolution in age-struc- Population parametersand philopatry of breeding tured populations. Cambridge Univ. Press, New female Wood Ducks. J. Wildl. Manage. 5 1:401- York, N.Y. 404. CHARNOV,E. L., AND J. R. KREBS. 1974. On clutch- HEPP,G. R., R. A. KENNAMER,AND W. F. HARVEYIV. size and fitness. Ibis 116:217-2 19. 1990. Incubation as a reproductive cost in female CLAWSON.R. L., G. W. HARTMAN.AND L. H. FRED- Wood Ducks. Auk 107:756-764. RI&ON. i979. Dump nesting in a Missouri HEUSMANN,H. W. 1972. Survival of Wood Ducks Wood Duck population.J. Wildl. Manage. 43:347- broods from dump nests. J. Wildl. Manage. 36: 355. 620-624. WOOD DUCK SURVIVAL 823

HEUSMANN,H. W., ANDR. H. BELLVILLE.1982. Wood PIANKA,E. R., ANDW. S. PARKER. 1915. Age-specific Duck researchin Massachusetts,1970-1980. MA reproductive tactics. Am. Nat. 109:453464. Div. Fish. & Wildl. Research Bull. 19, Westboro POTTS,G. R. 1969. The influence of eruptive move- MA. ments, age, population size and other factors on JOHNSON,D. H., AND J. W. GRIER. 1988. Determi- the survival of the Shag- (Phalacrocorax aristotelis nants of breeding distributions of ducks. Wildl. (L.)). J. Anim. Ecol. 38:53-102. Monogr. lOO:l-37. PUGESEK.B. H. 1983. The relationshio between oa- KADLEC,J. A., AND W. H. DRURY. 1968. Structure rental age and reproductive effort inthe California of a New England Herring Gull population. Ecol- Gull (Larus californicus).Behav. Ecol. Sociobiol. ogy 49~644-616. 13:161-171. KORPIM&, E. 1988. Costs of reproduction and suc- PUGFSEK, B. H., AND K. L. DIEM. 1990. The rela- cess of manipulated broods under varying food tionship between reproduction and survival in conditions in Tengmalm’s Owl. J. Anim. Ecol. 57: known-agedCalifornia gulls.Ecoloav 7 1:81 l-8 17. 1027-1039. REID, W. V. -1987. The cost of repro&ction in the LANK, D. B., E. G. COOCH,R. F. ROCKWELL,AND F. Glaucous-wingedGull. Oecologia 74:458-467. COOKE. 1989. Environmental and demoaraphic RICHDALE,L. E. 1957. A population study of pen- correlatesof intraspecificnest parasitism in Lesser guins. Clarendon Press, Oxford. Snow Geese Chen caerulescenscaerulescens. J. RICHDALE,L. E., AND J. WARHAM. 1973. Survival, Anim. Ecol. 58:2945. pair bond retention and nest-site tenacity in Bull- LAZARUS,J., AND I. R. INGLIS. 1978. The breeding er’s Mollymawk. Ibis 115:257-263. behaviour of the Pink-footed Goose: parental care ROCKWELL,R. F., C. S. FINDLAY,AND F. COOKE. 1987. and vigilant behaviour during the fledging period. Is there an optimal clutch size in Snow Geese? Behaviour 65:62-88. Am. Nat. 130:839-863. LESSELLS,C. M. 1986. Brood size in Canada Geese: ROHWER,F. C. 1985. The adaptive significance of a manipulation experiment. J. Anim. Ecol. 55: clutch size in prairie ducks. Auk 102:354-36 1. 669-689. ROHWER,F. C. In press. The evolution of reproduc- LESSELLS,C. M. 1987. Parental investment, brood tive patterns in waterfowl. In B.D.J. Batt et al. size and time budgets:behaviour of Lesser Snow [eds.], The ecology and management of breeding Geese families. Ardea 75: 189-203. waterfowl. Univ. Minnesota Press. Minneapolis. LOERY, G., K. H. POLLOCK,J. D. NICHOLS,AND J. E. ROHWER,F. C., AND M. G. ANDERSON.1988. Female- HINES. 1987. Age-specificity of Black-capped biased philopatry, monogamy, and the timing of Chickadee survival rates: analysis of capture-re- pair formation in migratory waterfowl, p. 187- capture data. Ecology 68: 1038-1044. 22 1. In R. F. Johnston [ed.], Current , LUDWIG,J. P. 1967. Band loss-its effect on banding Vol. 5. Plenum, New York. data and apparent survivorship in the Ring-billed ROHWER,F. C., AND S. FREEMAN. 1989. The distri- Gull population of the Great Lakes. -Banding bution of conspecificnest parasitismin birds. Can. 38:309-323. J. Zool. 611239-253. MORSE,T. E., ANDH. M. WIGHT. 1969. Dump nest- RP)s~, E. 1985. The effect of enlarged brood size ing and its effect on production in Wood Ducks. on the future reproductive potential of the Rook. J. Wildl. Manage. 33:284-293. J. Anim. Ecol. 54:255-260: SAS INSTITUTEINC. 1985. User’s auide: statistics. NICHOLS,J. D., M. J. CONROY,D. R. ANDERSON,AND version 5 edition. Cary, NC. - K. P. BURNHAM. 1984. Compensatory mortality SCHINDLER,M., AND J. LAMPRECHT.1987. Increase in waterfowl populations:a review of the evidence of parental effort with brood size in a nidifugous and implications for research and management. bird. Auk 104:688-693. Trans. N. Amer. Wildl. and Natur. Resour. Conf. SCOTT,D. K. 1980. Functional aspectsof prolonged 491535-554. parental care in Bewick’s Swans. Anim. Behav. NOL, E., AND J.N.M. SMITH. 1987. Effectsof age and 28:938-952. breedingexperience on seasonalreproductive suc- SEBER,G. A. F. 1982. The estimation of abun- cessin the Song Sparrow. J. Anim. Ecol. 56:301- dance and related parameters. Second edition. 313. Macmillan, New York. NUR, N. 1984. The consequencesof brood size for SEDINGER,J. S., AND D. G. RAVELING. 1990. Parental breedingBlue Tits I. Adult survival, weight change behavior of Cackling Canada Geese during brood and the cost of reproduction. J. Anim. Ecol. 53: rearing:division of labor within pairs. Condor 92: 479-496. 174-181. PARKIN,D. T., AND R. WHITE-ROBINSON.1985. The SEMEL,B., AND P. W. SHERMAN. 1986. Dynamics of effect of age on survival in the Canada Goose, nest parasitism in Wood Ducks. Auk 103:813- (Branta canadensis)in Nottinghamshire, p. 323- 816. 340. In B.J.T. Morgan and P. M. North [eds.], SEMEL,B., P. W. SHERMAN,AND S. M. BYERS. 1988. Statistics in ornithology. Springer-Verlag, New Effects of brood parasitism and nest-box place- York. ment on Wood Duck breeding ecology. Condor PETTIFOR.R. A.. C. M. PERRINS.AND R. H. MCCLEERY. 90:920-930. 1988: Individual optimization of clutch size in STEARNS,S. C. 1976. Life-history tactics: a review of Great Tits. Nature 336:160-162. the ideas. Quart. Rev. Biol. 5 1:3119. 824 FRANK C. ROHWER AND H. W. HEUSMANN

WIGGINS,D. A. 1990. Clutch size, offspring quality, determination of clutch size in precocial birds, p. and female survival in Tree Swallows-an exper- 33-68. In R. F. Johnston [ed.], Current omithol- iment. Condor 92534-537. ogy, Vol. 1. Plenum Press, New York. WILLIAMS,G. C. 1966. Natural selection, the costs WOOLFENDON,G. E., AND J. W. FITZPATRICK.1984. of reproduction, and a refinement of Lack’s prin- The Florida Scrub Jay: demography of a cooper- ciple. Am. Nat. 100:687-690. ative-breeding bird. Princeton Univ. Press, WINKLER, D. W., AND J. R. WALTERS. 1983. The Princeton, NJ.