INFLUENCE OF AGE AND TIME ON REPRODUCTIVE PERFORMANCE OF FEMALE LESSER SCAUP
ALAN D. AFTON • Delta WaterfowlResearch Station, Rural Route1, Portagela Prairie,Manitoba RIN 3A1, Canada,and Departmentof Biology,Institute for EcologicalStudies, University of North Dakota, Grand Forks, North Dakota 58202 USA
ABSTRACT.--Iexamined the influence of age and time on female reproductiveperformance in a marked populationof LesserScaup (Aythya affinis) breeding in southwesternManitoba from 1977to 1980.Arrival date and length of prelayingperiod varied amongyears but were not related to female age. Rate of nonbreeding and nest-initiation date varied among years and by female age. Age variation in nonbreedingwas due to the failure of yearlingsand a few 2-yr-olds to breed. Yearlings began laying an average of 5-8 days later than older females.Clutch size increasedwith female age but showed no significantannual variation despiteextreme changesin water conditions.Reproductive performance generally increased with female age and with improving water conditions. These results were consistentwith theoretical models that predict that reproductive effort (RE) increaseswith age or with declining residual reproductivevalue. My data support Williams' (1966)prediction that organismsadapted to living in variableenvironments adjust their RE in relation to probability of success.Female LesserScaup exhibited a temporally dynamic reproductivestrategy, and natural selectionseemingly has favored "yes-if" genes (Williams 1966) that effectbreeding only when conditionsare favorablefor the individual. Received27 April 1983, accepted3 October1983.
PARENTALage and breeding experienceinflu- main obscure due to insufficient information ence the reproductive performance of many on known-age individuals. Trauger (1971) re- avianspecies (Lack 1966, Kfomp 1970, Ryder ported that productivity, measuredby the pro- 1980). Among waterfowl, several breeding pa- portion of returning marked hens that were rameters may vary with age, including spring later observed with broods, increased with fe- arrival date, nestingchronology, and clutch size male age from 1 to 4 yr. The consistentlow (e.g. SowIs 1955, Finney and Cooke 1978, Kra- productivity of yearling femaleswas attributed pu and Doty 1979, Baillie and Milne 1982, to poor nesting success and nonbreeding Rockwell et al. 1983). Reproductive perfor- (Trauger 1971: 78). mance of waterfowl also varies annually in re- I collected reproductive information on a sponse to fluctuating environmental condi- marked population of LesserScaup breeding in tions in breeding, wintering, and migration southwestern Manitoba from 1977 to 1980, a areas (e.g. Crissey 1969, Bengtson 1971, Heit- period during which wetland conditions var- meyer and Fredrickson1981, Davies and Cooke ied dramatically. In this paper, I examine the 1983, Krapu et al. 1983). influence of age and time on the reproductive Female age and prairie drought are impor- performance of females. Breeding parameters tant factorsinfluencing reproductionof Lesser estimated were: (1) arrival date, (2) rate of non- Scaup (Aythya affinis;Rogers 1959, 1964; Traug- breeding, (3) length of prelaying period, (4) er 1971).Nonbreeding and age at first breeding nest-initiation date, (5) clutch size, (6) nest suc- in this specieshave been discussedand spec- cess,(7) rate of renesting,and (8) brood surviv- ulated upon (seeTrauger 1971: 87), but specific al. Results are discussedin relation to predic- factorsaffecting these breeding parameters re- tions of theoretical models concerning age- specificand temporal variation in reproductive effort.
• Presentaddress: Wetland Wildlife Populations and STUDYAREA AND METHODS ResearchGroup, MinnesOta Department of Natural Resources,102 23rd Street,Bemidji, Minnesota 56601 The 777-hastudy area was located 3.3 km southeast USA. of Erickson, Manitoba (50ø30'N, 99ø55'W)and was de- 255 The Auk 101: 255-265. April 1984 256 ALAN D. AFTON [Auk, Vol. 101
TABLE 1. Comparisonsof selectedenvironmental parameters and springarrival datesof LesserScaup, 1977- 1980.
Year
Parameter 1977 1978 1979 1980 Mean maximumApril temperature(øC) 14.6 7.4 1.2 14.5 Total April-May precipitation (mm) 107.3 73.4 138.5 10.0 Maximum snowpackbefore spring thaw (mm H20) 33.4 47.1 111.1 57.7 Mean April-May total soil moisture (mm)• 43.9 66.2 93.0 40.3 Pondice-free date 17 April 30 April 18 May 22 April May ponds/km2b 18.5 32.0 35.1 19.7 Firstarrival of unmarkedLesser Scaup 14 April 19 April 1 May 16 April First arrival of marked LesserScaup _c 2 May 13 May 21 April Water holding capacity- 177.5 ram. 1962-1981average 32.8/km2 (Stoudt 1982,J. Serie pers.comm.). cA single marked individual returned on 4 May 1977.
scribed in detail by Rogers (1964), Hammell (1973), comm.) from transectsnear Minnedosa (23 km south and Sunde and Barica(1975). Reproductive data were of the study area) and weather data for grid-point collectedfrom known-age females (banded as duck- 237 (18 km east, Street and Findlay 1981) to index lings) or those aged by eye color (Trauger 1974, Af- annual changes in wetland conditions. Total soil ton in prep.). Adults were captured in decoy traps moisture,considered a prime parameterfor express- (Anderson et al. 1980), nest traps (Weller 1957), and ing meteorologicaldrought, was estimatedfrom a cli- bail traps (Sugdenand Poston1970) modified with a matic water balance model (Street and Findlay 1981). radio-controlled release mechanism. Ducklings and Reproductiveparameters with categoricalresponses brood hens were capturedby drive-trapping (Cowan (nonbreeding, nest success,renesting) were analyzed and Hatter 1952) and night-lighting (Lindmeier and with linear models (Grizzle et al. 1969) using the Jessen1961). Captured birds were banded and indi- FUNCAT procedureof the StatisticalAnalysis Sys- vidually markedwith nasalsaddles (Sugden and Pos- tem (Helwig and Council 1979). Parameters with ton 1968). continuous responseswere analyzed with regression The breeding statusof marked femaleswas deter- techniques(Kleinbaum and Kupper 1978:362) using mined by: (1) intensive behavioral observation,(2) the GLM procedure (Freund and Littell 1981). Dun- locating nests, and (3) external physical indicators can's multiple range test with Kramer's adjustment (brood patch, wide pubis spread, egg in oviduct). was used to detect differencesamong means of main When estimatingthe rate of nonbreeding, I also in- effects (Duncan 1955, Kramer 1956). cluded data from 18 unmarked females that were col- Four female age classeswere used:1-yr, 2-yr, 3-yr, lected 2-12 km from the study area. The breeding and 4-yr and older. Arrival and nest-initiation dates statusof these females was determined by ovarian were coded as Julian dates in analyses. I used pro- examination (McKnight and Buss 1962). The breeding statusof 5, 1, 4, and 22 marked fe- males could not be accurately determined in years TABLE2. Probability levels from analysesof the in- 1977-1980, respectively;consequently, they were ex- fluence of year, female age, and year-by-ageinter- cluded from analyses.Sixteen (50%) of thesefemales action on various breeding parameters. were probably migrants,because they were captured early in the seasonand were never observed again Explanatory variables in that year or in subsequentyears. The remainder (15 yearlings and 1 4-yr-old) were known residents Parameter Year Age Y by A of the study area. Arrival date a 0.0005 0.6659 0.8666 Nestswere found by watching femalesfly or swim Rate of to their nestsand by flushing hens from coverwith nonbreeding b 0.0204 0.0006 0.2551 a trained dog. One or two assistantsand I observed Prelaying perioda 0.0001 0.6598 0.1404 Nest-initiation all marked hens on the study area during the nest- date a 0.0001 0.0001 0.0674 initiation period each year. Marked pairs were ob- Clutch size • 0.1374 0.0001 0.4490 served for 265 h, 896 h, and 685 h in years 1977- Nest success b 0.7992 0.0554 0.6343 1979, respectively. Observational effort in 1980 was Rate of renestingb 0.2085 0.6915 0.8090 comparableto previous years, although total hours Brood survivaP 0.1643 0.4581 0.5778 observed were not recorded. GLM analysis(see text). I used pond counts (Stoudt 1982, J. Serie pers. FUNCAT analysis. April 1984] LesserScaup Reproductive Performance 257
TABLE3. Summarystatistics by year for eachbreeding parameter.Statistics presented are mean + SE (sample size) and range for parameterswith continuousresponses, and percentages(sample size) for those with categoricalresponses.
Year
Parameter 1977 1978 1979 1980
Arrival date -- 133.5 + 2.3 (8) 136.8 + 0.8 (17) 123.8 + 1.9 (45) 122-143 133-145 111-167 Rate of nonbreeding 28.0% (25) 17.2%(29) 2.3% (43) 15.3%(59) Prelayingperiod -- 25.8 + 2.6 (5) 32.2 + 0.9 (15) 43.0 + 1.9 (13) 21-35 24-38 25-56 Nest-initiation date 168.2 ñ 3.8 (6) 164.7 + 1.9 (17) 169.4 + 0.7 (31) 163.4 + 1.2 (24) 154-179 152-178 162-179 152-176 Clutch size 10.4 + 0.8 (5) 10.4 + 0.4 (17) 10.2 + 0.2 (31) 10.3 + 0.3 (24) 8-12 8-14 8-13 8-13 Nest success 17.7% (17) 29.2% (24) 40.4% (47) 26.8% (41) Rate of renesting 0.0%(14) 6.3%(16) 38.9%(18) 16.0%(25) Brood survival 0.518 + 0.134 (3) 0.446 ñ 0.101 (7) 0.737 + 0.076 (19) 0.766 + 0.067 (10) 0.250-0.667 0.091-0.846 0.000-1.000 0.333-1.000
portionsof young in eachbrood surviving until age idents varied significantly among years 1978- classII (>20 days)for analysisof broodsurvival (Gol- 1980 but were not related to female age (Table lop and Marshall 1954).Identification of older broods 2). The mean arrival date in the early spring of for counting was frequently impossiblebecause of 1980 (4 May) was earlier (P < 0.05) than the henstemporarily or permanentlyleaving their broods. mean arrival dates in the later springs of 1978 Survival to age classII was considereda good index (14 May) and 1979 (17 May, Table 3). Annual of fledgingsuccess, because most duckling mortality means and variances of arrival dates were cor- occursduring the first week of life (Hammell 1973, Afton 1983). related (r = -0.998, df = 1, P < 0.05), suggest- ing that females arrived more synchronously RESULTS in late springs. In 1980, nonbreeding yearlings arrived later Water conditions and spring phenology var- (• = 130.6, SE = 6.2 days, n -- 7), on average, ied dramaticallyamong years 1977-1980 (Table than yearlings that nested (• = 119.7, SE = 1.2, 1). The springsof 1977 and 1980 were early in n = 7), but the difference was not statistically comparisonto the delayed spring of 1979, as significant (Mann-Whitney test, P > 0.05). reflected by mean maximum April tempera- Variance in arrival date was greater for non- tures and dates on which all study area ponds breeders than for breeding yearling females became ice-free. Wetland numbers were high- (Siegel-Tukeytest, P < 0.01). est in 1979 and lowest during the severe 1977 Rate of nonbreeding varied significantly drought, when wide mud flatswere evident on among years 1977-1980 and by female age (Ta- permanent ponds. Water conditions (indexed ble 2). Age variation in nonbreeding was due by pond numbers and soil moisture)improved to the failure of yearlingsand a few 2-yr-olds from 1977 through 1979, due to increasing to breed(Table 4). All 3-yr-old and 4-yr-old and amountsof winter and spring precipitation,but older females nested. Proportions of yearling deteriorated in 1980 because of reduced winter females that bred (r = 0.735) and, to a lesser snowfall and almost nonexistent spring rains extent, proportions of 2-yr-olds that bred (r = (Table 1). 0.201)generally increased as water conditions First-arrival dates of unmarked LesserScaup (pond numbers) improved, but these relation- differed by 17 days among yearsand were cor- ships were not statisticallysignificant (df = 2, related (r = -0.951, df = 2, P < 0.05) with mean P > 0.05). maximumApril temperatures(Table 1). Marked The length of the prelaying period (number residentsbegan arriving 1 to 2 weeks later than of days from arrival until first egg) varied sig- unmarkedbirds eachyear. Arrival datesof res- nificantly among years 1978-1980 but was not 258 ALAND. AFTON [Auk,Vol. 101
TABLE4. Summary statisticsby female age for each breeding parameter. Statisticspresented are mean _+SE (samplesize) and range for parameterswith continuousresponses, and percentages(sample size) for those with categorical responses.
Female age (yr) Parameter 1 2 3 > 4
Arrival date 129.1 + 2.3 (37) 127.0 + 2.9 (12) 132.2 + 3.2 (9) 122.8 + 2.4 (12) 111-167 116-140 112-143 112-134 Rate of nonbreeding 29.3% (58) 9.6% (52) 0.0% (23) 0.0% (23) Prelaying period 36.6 + 3.3 (9) 39.3 + 2.2 (6) 29.9 + 2.7 (9) 37.4 + 2.3 (9) 24-56 32-44 21-47 25-47 Nest-initiation date 171.1 + 0.9 (26) 165.5 + 1.3 (22) 162.8 + 1.7 (16) 163.4 + 1.3 (14) 163-178 157-179 152-171 154-169 Clutch size 9.0 -+ 0.1 (26) 10.0 -+ 0.2 (21) 10.9 + 0.3 (16) 12.1 + 0.2 (14) 8-10 8-12 9-12 11-14 Nest success 26.3% (38) 22.2% (45) 45.5% (22) 41.7% (24) Rate of renesting 8.7%(23) 17.9%(28) 27.3%(11) 18.2%(11) Brood survival 0.693 + 0.101 (10) 0.672 + 0.106 (10) 0.575 + 0.118 (9) 0.751 + 0.070 (10) 0.111-1.000 0.000-1.000 0.000-1.000 0.250-1.000
related to female age (Table 2). Mean lengths female age was 4.3 times more important than of prelaying periods for years 1978-1980 were nest-initiation date in predicting clutch size. all different (P < 0.05, Table 3). Nest success(proportion of nestsin which at The nest-initiation date (first egg date in first leastone egg hatched)tended to vary (P < 0.06) nests only) varied significantly among years with female age but not among years (Table 2). 1977-1980 and by female age (Table 2). Mean Nest successof 1 and 2-yr-old females com- date for yearlings(20 June)was later (P < 0.05) bined (24%, n = 83) was lower (Chi-square test, than dates for hens of 2-yr (15 June), 3-yr (12 P < 0.05) than that of older females (43%, n = June), and 4-yr and older (12 June, Table 4). 46). Nest successgenerally increased as water Linear regressionanalysis for years 1978-1980 conditions (pond numbers) improved, but the combinedindicated that arrival date explained relationship was not statistically significant 26%of the variation in nest-initiation date (Fig. (r = 0.862, df = 2, P > 0.05). 1). Arrival and nest-initiation dates were not Most unsuccessfulnests (90%, n = 89) result- significantlycorrelated (P > 0.05) within any ed from destructionby mink (Mustela vison), year,although the small 1978sample suggested raccoons(Procyon lotor), red fox (Vulpesvulpes), a positive relationship (r = 0.697, df = 3, P > or striped skunks (Mephitis rnephitis);the re- 0.05). mainder were caused by farming operations Clutch size (first nests only) increased sig- (7%) and investigator disturbance(3%). Seven nificantly with female age but showedno sig- marked females (3 yearlings, 3 2-yr-olds, 1 4- nificant annual variation (Table 2). Covariance yr-old) were killed on the nest in 1979 (4 by analysis,in which the effect of nest-initiation mink, 2 by red fox) and 1980 (1 by red fox). date was removed, gave identical results;age Multivariate analysisindicated that the pro- effect was significant(P < 0.001) but year and portion of femalesrenesting, after lossof their year-by-ageinteraction were not (P > 0.05).The first nests,did not vary significantlywith age mean clutch sizesof all age classeswere differ- or among years 1977-1980 (Table 2). Univariate ent (P < 0.05, Table 4). analysis,after all ages were combined, how- Multiple regressionanalysis indicated that ever, indicated that proportions varied among age (P < 0.001) and nest-initiationdate (P < years (Chi-square test, P < 0.02). Significant 0.05) together explained 71% of the variation variation resulted primarily from lower and in clutch size (CSIZE = 14.39 + 0.94 (AGE) - higher than expectedproportions in 1977and 0.04 (NDATE), P < 0.001, n = 77). A compari- 1979, respectively.Renesting rates generally son of standard partial regression coefficients increased as water conditions (pond numbers) (see Steel and Torrie 1960: 299) indicated that improved, but the relationshipwas not statis- April 1984] LesserScaup Reproductive Performance 259
180 Y = 127.050 + 0.293X r 2 = O.262 N = 33 175 P <0,01
i1.1
170
165
160 ß ß
ß ß ß ß : 1978 155 ß : 1979
ß ß : 1980
150 I10 115 120 125 1:50 1:55 140 145 150 ARRIVAL DATE Fig. 1. Relationshipof nest-initiationdate to springarrival date for years1978-1980. Dates are in Julian days. tically significant (r = 0.627, df = 2, P > 0.05). residents in years 1978-1980, these birds were The renesting rate of 2-yr-old and older fe- migrants.Gates (1962) noted a similar delay in malescombined (20%, n = 50) appearedto be returns of resident female Gadwall (Anas stre- greaterthan that of yearlings(Table 4), but the pera) in Utah. These observations contrast with differencewas not statisticallysignificant (Chi- Sowl's(1955) findings that in a smallsample of square test, P > 0.05). color-banded Northern Pintails (Anas acuta), Broodsurvival did not vary significantlywith residents were among the first to arrive in femaleage or amongyears 1977-1980 (Table 2). spring. Additional studies of marked birds are Mean proportion surviving to age classII for needed to determine speciesdifferences and all years and age classescombined was 0.675 whetheror not aspectsof the breedingstrategy (SE = 0.049, n = 39 broods). Brood survival was (e.g. early nestingversus late nesting)are re- not related to water conditions,as indexed by lated to arrival patterns. pond numbers (r = -0.032, df = 2, P > 0.05). I found no age-specificvariation in arrival dates, after adjusting for annual differences. This contradictsprevious conclusions that adult DISCUSSION femaleresidents generally precede yearlings to the breeding grounds, as reported for Lesser ! found a generalrelationship between tem- Scaup(Trauger 1971), Redheads (Aythya amer- perature and spring arrival of LesserScaup. icana, Johnson 1978), and Northern Pintails Similar relationshipshave been noted for dab- (Sowls1955). These earlier findingswere most- bling and diving ducks(Anas and Aythyaspp.) ly anecdotal,however, were unsupportedby on Delta Marsh, Manitoba (Sowls 1955) and in statisticaltests, or includeddirect comparisons North Dakota (Hammond and JohnsonMS), for of wild adultsand hand-reared yearlings. LesserScaup at Yellowknife, Northwest Terri- The tendency of older, more experiencedin- tories(Trauger 1971),and for a variety of avian dividuals to arrive earlier in the season has been speciesat other locations (e.g. Saunders 1959, documentedin other avian groups,e.g. Arctic Mikkonen 1981). Terns (Sternaparadisaea, Coulson and Horobin The firstLesser Scaup to returnin springwere 1976), Prairie Warblers (Dendroicadiscolor, No- unmarked, and, based on the return of marked lan 1978),and SpottedSandpipers (Actitis mac- 260 ALAND. AFTON [Auk,Vol. 101 cularia, Oring and Lank 1982). In the latter estimateof the rate of nonbreeding by year- species, delayed arrival by inexperienced in- lings eachyear could be slightly low. dividuals appeared to maximize their success Residentsspent an average of 4-6 weeks on in entering the breeding population, which is the study area before nesting. Prelaying pe- consistentwith the hypothesis that sexual se- riods were longestin 1980,when birds arrived lection drives arrival times. Sexual selection an averageof 10-13 days earlier than in 1978 probably does not directly influence arrival and 1979. Arrival dates in 1978 and 1979 were times of ducks, becausepairing takes place on similar, but the mean prelaying period in 1979 wintering and migration areas. It may, how- was longer due to delayed nesting, which cor- ever, influence age variation in winter distri- related with the late date that ponds became butions and initiation of spring migration (see ice-free in 1979 (Table 1). Late-nestingdab- Saylerand Afton 1981). bling ducks (Anasspp.) also spend a lengthy The large variation in arrival dates by non- period on breeding areasbefore laying (Gates breeding yearlings in 1980 suggeststhat the 1962, Afton 1979). "decision" not to breed occurs after arrival on Yearlingsbegan laying an averageof 5-8 days breeding areas for some females (some non- later than older females.Delayed nesting by breeders arrive early) but before arrival for first-timebreeders has been reported frequent- other individuals (some arrive late). Three late- ly for waterfowl (Mendall 1958, Gates 1962, arriving yearlings were captured and did not Bellroseet al. 1964, Brakhage1965, Dane 1965, have brood patches, so their late arrival was Grice and Rogers1965, Coulter and Miller 1968, unlikely to be due to breeding attempts else- Mihelsons et al. 1970, Kear 1973, Finney and where. Thus, factors during winter or spring Cooke 1978, Johnson 1978, Alliston 1979a, Kra- migration may be important in determining an pu and Doty 1979, Baillie and Milne 1982) and individual's future reproductive performance. other birds (seeKlomp 1970, Perrins 1970, Ry- Proportions of I and 2-yr-old females that der 1980). nested tended to increasewith improving water Annual variationin nestingchronology was conditions. Water levels on the study area in partially related to variation in arrival dates 1980 were much higher than in 1977, contrary (Fig. 1). Mean date of nest initiation was ear- to that indexed by soil moisture and pond liest in 1980 (12 June), when residents arrived numbers from transects near Minnedosa (Table early, and was delayed in 1979 (18 June),when 1). Pondswere initially full in 1980 due to car- residentsarrived, on average,13 dayslater than ry-over from 1979, and water levels in 1980 in 1980. Individuals did not show a distinct re- were similar to levels in 1978 (based on extent lationship within years, i.e. the early arrivals of mud flats). Although Minnedosa is in close in a given year did not initiate nests signifi- proximity to the study area (23 km), there are cantly earlier than later arrivals in the same slight differencesin pond size and permanency year. between the areas (Kiel et al. 1972: 27). Wet- Although sample size was small, the severe lands in the Ericksonarea are, on average,more droughtin 1977appeared to delaynesting (ever• permanent, larger in size, and of lower density though residentsprobably arrived as early or per unit area. Thus, a statisticallysignificant earlier than in 1980), basedon April tempera- relationshipmay have been recordedbetween tures,arrival of unmarked birds, and pond ice- rate of nonbreeding and annual water condi- free dates (Table 1). Nesting by Canvasback tions, if water levels had been measured on the (Aythya valisineria,Stoudt 1982) and Mallards studyarea. Perhaps more importantis the fact (Anas platyrhynchos,Krapu et al. 1983) also is that the power of this correlationwas very low delayed during drought. due to my small sample of years (n = 4). Amphipods (Crustacea:Amphipoda) are the Most (94%, n = 16) of the known residents most important foods of LesserScaup during whosebreeding status was not determinedwere the breeding season (Rogers and Korschgen yearling females.I was unable to determine 1966, Bartonek and Hickey 1969, Afton in their statusprimarily becauseof their frequent prep.). Egg-laying by Lesser Scaup generally widespreadmovements and failure to develop was correlated with June peaks in Garnrnarus small discretehome ranges.Many of these fe- lacustrisdensities in area lakes,although max- malesmay have been nonbreedersand thus my imum amphipod biomass occurs later in April 1984] LesserScaup Reproductive Performance 261
August (Mathias and Papst 1981, Salki 1981, terfowl species(e.g. Bengtson1971, Davies and Mathias et al. 1982, Mathias pers. comm.). Egg- Cooke 1983, Krapu et al. 1983). Interestingly, laying by Tufted Ducks(Aythya fuligula, Laugh- annual mean clutch size increased during lin 1975) and Ruddy Ducks (Oxyurajamaicensis, drought for someprairie-nesting ducks in two Gray 1980) also is synchronized with peak den- long-termstudies (Smith 1971,Stoudt 1971). My sities or maximum biomass of their principal results support Bailey's (1981) prediction that foods (chironomid larvae). this was due to only the oldest femalesbreed- Clutch size increasedan average of one egg ing in these years, but, in addition, renesting per year of age, from ages1-yr to 4-yr and old- was probably uncommon, and thus annual er. Small samples prevented me from deter- means were not reduced by smaller renest mining whether or not clutch size continues to clutches (see Coulter and Miller 1968: 27). increase in the 4-yr and older group, but, be- Nest successof 1 and 2-yr-olds was lower cause a maximum clutch size of 14 was record- than that of older female LesserScaup. In con- ed only once, any average increase must be trast,female age doesnot influence nestsuccess smaller than that of younger birds. Yearling of several other waterfowl species(Brakhage ducks frequently have been reported to lay 1965, Johnson 1978, Baillie and Milne 1982). smaller clutches than older females (Stotts and Nest successof Lesser Scaup generally in- Davis 1960, Dane 1965, Coulter and Miller 1968, creasedwith improving water conditions(Rog- Morse et al. 1969, Alliston 1979a, Krapu and ers 1964,this study).Factors contributing to age- Doty 1979), but variations in clutch size of old- specificand temporal variation in nest success er females rarely have been documented(Bail- requirefurther study. lie and Milne 1982), except for geese(Kossack The proportion of hens that renested, after 1950, Brakhage 1965, Kear 1973, Cooper 1978, loss of their first clutch, tended to increase with Finney and Cooke 1978, Rockwell et al. 1983). improving water conditions.Renesting rates of Clutch size increaseswith age in many other yearlingsappeared to be lower than that of old- avian species(see Lack 1966, Klomp 1970, Ry- er birds,but the differencewas not statistically der 1980). Some species,however, show no ef- significant.A higherrate of renestingby older fect of female age (e.g. Koskimies 1957, Lei- femaleswas suggestedfor Gadwall (Gates1962), nonen 1973), and clutch size even declines with Wood Ducks(Aix sponsa,Grice and Rogers1965), age in a few instances(Klomp 1970). LesserScaup (Trauger 1971), Redheads(Allis- Clutch size did not vary among years 1977- ton 1979b), and Mallards (Krapu and Doty 1980 after adjusting for female age, despite 1979), but was demonstratedonly for Blue- marked changesin wetland conditions. Rogers winged Teal (Anasdiscors, Strohmeyer 1967) and (1962: 56) claimed but did not demonstrate that indirectly for Mallards and Black Ducks (Anas densities of Gammarusspp. were much reduced rubripes,Coulter and Miller 1968). during low water years. In contrast,J. Mathias Broodsurvival did not vary significantlywith (pers. comm.) did not find large variations in female age or among years.These resultsare G. lacustrisand Hyalella aztecadensities among tentative becausesamples were sosmall in some three lakes near the study area or among years years, due to nonbreeding and poor nest suc- 1976-1979, despite marked changes in water cess,that demonstrationof significant differ- levels. Basedon stomachanalyses of tiger sal- ences would be difficult. Lack of annual varia- amanders (Ambystomatigrinum) and stocked tion in brood survival was not surprising, rainbow trout (Salmogairdneri), Olenick and Gee becausefemales moved broods to the largest (1981) suggested that Gammarus spp. were and most permanent ponds (Hammell 1973, abundant during the severe 1977 drought. Afton in prep.), and amphipodpopulations, an These observationssuggest that food abun- importantduckling food (Bartonekand Hickey dance was not limiting clutch size during the 1969, Sugden 1973), did not fluctuategreatly drought for those females that nested. among years. Variations in abundance, distribution, and Severaltheoretical papers have predicted that quality of food resources, associated with reproductive effort (RE) should increase with drought and other environmental fluctuations, age, or that RE should vary inversely with re- have been identified as proximate causesof an- sidual reproductivevalue (RRV) (Williams 1966, nual variation in clutch size of a variety of wa- Gadgil and Bossert 1970, Pianka and Parker 262 ALAND. AFTON [Auk, Vol. 101
1975, Charlesworth and Leon 1976, Pianka 1976; Poor reproductiveperformance of young fe- but see Fagen 1972, Schaffer 1974). Williams male waterfowl has been attributed to less ex- (1966)predicted that organismsadapted to liv- perience in locating, competingfor, and ex- ing in variableenvironments should adjust their ploiting food resources,to less ability to RE relative to their probability of success. accumulatelarge endogenousreserves, and to In general,my datawere consistentwith pre- lower efficiency at foraging for specific nu- dictions that RE increaseswith age or with de- trients needed for egg-formation(e.g. Batt and clining RRV. Although RE was not measured Prince 1978, Batt 1979, Krapu and Doty 1979, directly, proportions of females breeding and Baillie and Milne 1982). Young birds are less •enesting, and perhapsclutch size, should be efficient feedersthan adults in somespecies (see positively correlated with RE. My data also citations in Baillie and Milne 1982), but age supportWilliams' (1966) prediction that indi- variation in feeding efficiency has not been vidualsadjust their RE in relation to probabil- documented for waterfowl. ity of success.Females exhibited a temporally The best evidence that age variation in re- dynamic reproductivestrategy (cf. Nichols et productiveperformance of ducksis related to al. 1976), and natural selection seemingly has feeding comesfrom the captiveexperiments of favored"yes-if" genes (Williams 1966:173) that Batt and Prince (1978). They found no age-spe- effect breeding only when conditions are fa- cific variation in the breeding parameters of vorable for the individual. Females 3-yr and Mallards under controlled conditions, in the older made one nesting attempt regardlessof absenceof competition, and when food was environmental conditions,which probably re- provided ad libitum,a finding that was incon- flectstheir higher probability of offspringsuc- sistentwith the field data of Krapu and Doty cess(higher nest success)and their relatively (1979). Studies of age variation in feeding ef- small RRV. Changes in RRV with age were ficiency and how such variation relates to re- probablya function of survivorship,because I productiveperformance are urgently needed. found no evidence that fecundity decreased Batt (1979) suggestedthat delayed breeding with age. and smaller clutchesof yearling Mallards and Nichols et al. (1976) noted that high annual other waterfowl were consequencesof inex- variation in reproductive performance could perience and competition with older birds indicate either variation in RE or variation in rather than part of an evolved reproductive food levels with constant RE. Becausepopula- strategyper se.I believe that he was implying tionsof the predominantfood sourcesof Lesser that the REs of yearlings and older females Scaupdid not fluctuategreatly among years (see were similar but that the total amount of re- p. 261), RE apparently did vary among years. sourcesavailable to yearlings was less (due to Dynamic reproductivestrategists apparently lower feeding efficiencyand/or displacement "predict" offspring survival from environmen- to inferior territories). This seems likely for tal cues (Cohen 1967, Nichols et al. 1976). Nest Mallards, when one considers Batt's experi- success(influenced primarily by predation) de- mental results, observations of territorial be- creased with deteriorating water conditions, havior (e.g. Titman 1983), and Krapu and Do- while brood survival remained relatively con- ty's (1979) findings that, although the clutch stant (Rogers1964, this study); consequently, size of wild yearlings averaged one egg less nest successwas the critical factor affecting than that of adults, the clutch size of adults (2- variation in offspring survival. Thus, females yr and older) did not increasefurther with probablyuse proximateenvironmental factors age. associated with water-level fluctuations (see I believe, however, that clutch-size variation Rogers1962, 1964) to predicttheir probability of Lesser Scaup was not due solely to experi- of success,whereas the primary ultimate factor ence or competition but was at least partially affecting nonbreeding (and RE) probably is related to age variation in RE. Breeding pairs predation upon nests. Becausepredominant did not defend territories but, instead, had rel- foods did not fluctuate greatly among years, atively small, highly overlappinghome ranges food abundanceon breeding areas apparently (Hammell 1973, Afton in prep.). Although in- was not an important proximate factor affect- experiencein locatingand efficientlyexploit- ing nonbreeding of LesserScaup. ing food resourceson breedingareas could ex- April 1984] LesserScaup Reproductive Performance 263 plain lower reproductive performances of 1979b. Renesting by the Redhead Duck. yearlings, such explanations alone seem un- Wildfowl 30: 40-44. likely to accountfor subsequentlarge increases ANDERSON,M. G., g. D. SAYLER,& A.D. AFTON. 1980. in clutch size of older females (see also Curio A decoy trap for diving ducks. J. Wildl. Mgmt. 1983). Captive breeding experiments with 44: 217-219. Lesser Scaup are needed to examine the rela- BAILEY,R.O. 1981. A theoretical approach to prob- lems in waterfowl management. Trans. North tive importance of experience and age varia- Amer. Wildl. Nat. Resources Conf. 46: 58-71. tion in RE more critically. Comparisons of BAILLIE,S. R., & H. MILNE. 1982. The influence of clutch size of wild first-time breeding females female age on breeding in the Eider Somateria of different ages might provide a strong field mollissima.Bird Study 29: 55-66. test of the hypothesis. BARTONEK,J. C., & J. J. HICKEY. 1969. Food habits of Canvasbacks,Redheads, and Lesser Scaup in ACKNOWLEDGMENTS Manitoba. Condor 71: 280-290. BATT, B. D. J. 1979. Individual variation and the I wish to acknowledgethe considerableassistance analysis of Mallard populations. Pp. 95-102 in of J. Afton, M. Afton, G. Hammell, K. Warrington, Waterfowl and wetlands--an integrated review local landowners, and numerous Webster Fellows and (T. A. Bookout, Ed.). Madison, Wisconsin, Proc. Delta associates.! have profited greatly from discus- 1977Symp., North Central Sect.,The Wildl. Soc. sionswith M. Anderson and R. Saylet concerning the , & H. H. PRINCE. 1978. Some reproductive behavior of Aythyaspecies. I thank J. Serie, Northern parametersof Mallards in relation to age, captiv- Prairie Wildlife Research Center, Jamestown, North ity, and geographicorigin. J. Wildl. Mgmt. 42: Dakota for providing assistance,equipment, and 834-842. pond-survey data. I am grateful to B. Findlay and G. BELLROSE,F. C., K. L. JOHNSON,& T. U. MEYERS. 1964. A. McKay, Canadian Climate Centre, Downsview, Relative value of natural cavities and nesting Ontario for providing weather data and discussion housesfor Wood Ducks. J. Wildl. Mgmt. 28: 661- of meteorologicaldrought indices.I thank J. Mathias, 676. Freshwater Institute, Winnipeg, Manitoba for allow- BENGTSON,S.-A. 1971. Variations in dutch-size in ing me to cite unpublished information on amphi- ducks in relation to the food supply. Ibis 113: pod populations. The Canadian Wildlife Service 523-526. granted banding and collecting permits. M. Ander- BRAKHACE,G.K. 1965. Biology and behavior of tub- son, R. Bailey, B. Batt, R. Crawford, R. Drobney, S. nesting Canada Geese.J. Wildl. Mgmt. 29: 751- Findlay,D. Johnson,J. Nichols,L. Oring, and R. Say- 771. let provided helpful commentson the manuscript. CHARLESWORTH,B., & J. A. LEON. 1976. The relation Financial support was provided by the Alumni As- of reproductive effort to age. Amer. Natur. 110: sociation, Department of Biology, Institute for Eco- 449-459. logical Studies,and Sigma Xi chapter of the Univer- COHEN,D. 1967. Optimizing reproductionin a ran- sity of North Dakota; James Ford Bell Sr. Memorial domly varying environment when a correlation Fellowship Fund; North American Wildlife Foun- may exist between the conditions at the time a dation through the Delta Waterfowl Research Sta- choice has to be made and the subsequentout- tion; and Minnesota Department of Natural Re- come. J. Theor. Biol. 16: 1-14. sources.The University of North Dakota Computer COOPER,J. A. 1978. The history and breeding biol- Center provided computer time, keypunching, and ogy of the Canada Geeseof Marshy Point, Man- programming assistance.This is ResearchReport No. itoba. Wildl. Monogr. No. 61. 35 of the Institute for EcologicalStudies. COULSON,J. C., & J. HOROBIN. 1976. The influence of age on the breeding biology and survival of LITERATURE CITED the Arctic Tern Sternaparadisaea. J. Zool., London 178: 247-260. AFTON,A.D. 1979. Time budget of breeding North- COULTER,M. W., & W. g. MILLER. 1968. Nesting bi- ern Shovelers. Wilson Bull. 91: 42-49. ology of Black Ducks and Mallards in northern 1983. Male and female strategiesfor repro- New England. Vermont Fish and Game Dept. duction in LesserScaup. Unpublished Ph.D. dis- Bull. 68: 1-74. sertation. Grand Forks, North Dakota, Univ. COWAN,I. Met., & J. HATTER.1952. A trap and tech- North Dakota. nique for the capture of diving waterfowl. J. ALLISTON,W. G. 1979a. The population ecologyof Wildl. Mgmt. 16: 438-441. an isolatednesting population of Redheads(Ay- CRISSEY,W. F. 1969. Prairie potholes from a conti- thya americana).Unpublished Ph.D. dissertation. nental viewpoint. Pp. 161-171 in Saskatoonwet- Ithaca, New York, Cornell Univ. lands seminar. Can. Wildl. Serv. Rept. No. 6. 264 ALAND. AFTON [Auk, Vol. 101
CURIO,E. 1983. Why do young birds reproduceless Waterfowl habitat trends in the aspenparklands well? Ibis 125: 400-404. of Manitoba. Can. Wildl. Serv. Rept. Ser. No. 18. DANEßC.W. 1965. The influence of age in the de- KLEINBAUM,D. G., & L. L. KUPPER. 1978. Applied velopment and reproductive capability of the regressionanalysis and other multivariatemeth- Blue-winged Teal (Anas discorsLinnaeus). Un- ods. North Scituate, MassachusettsßDuxbury published Ph.D. dissertation.Lafayette, Indiana, Press. Purdue Univ. KLOMP, H. 1970. The determination of clutch-size DAVIESßJ. C., & F. COOKE. 1983. Annual nesting pro- in birds. A review. Ardea 58: 1-124. ductivity in Snow Geese:prairie droughts and KOSI Some aspectsof the breeding biology of the SCHAFFER,W. M. 1974. Selection for optimal life Hooded Merganser.J. Wildl. Mgmt. 33: 596-604. histories:the effectsof age structure.Ecology 55: NICHOLS,J. D., W. CONLEY,B. BATT, &: g. R. TIPTON. 291-303. 1976. Temporally dynamic reproductive strate- SMITH,A. G. 1971. Ecologicalfactors affecting wa- giesand the conceptof r- and K-selection.Amer. terfowl productionin the Alberta parklands.U.S. Natur. 110: 995-1005. Fish Wildl. Serv. Resources Publ. No. 98. NOLAN,V., JR. 1978. The ecology and behavior of SOWLS,L. K. 1955. Prairie ducks. Washington, D.C., the Prairie Warbler Dendroica discolor. Ornithol. Wildl. Mgmt. Inst. Monogr. 26: 1-595. STEEL,R. G. D., &:J. H. TORRIE.1960. Principlesand OLENICK,R. J., &:J. I-I. GEE. 1981. Tiger salamanders proceduresof statistics.New York, McGraw-Hill (Arabystomatigrinum) and stockedrainbow trout Book Co. (Salmogairdneri): potential competitorsfor food STOWIS,V. D., &: D. E. DAVISß 1960. The Black Duck in Manitoba prairie pothole lakes. Can. Field- in the ChesapeakeBay of Maryland: breeding Nat. 95: 129-132. behavior and biology. ChesapeakeSci. 1: 127- ORING, L. W., &: D. B. LANK. 1982. Sexual selection, 154. arrival times,philopatry and site fidelity in the $TOUDT,J. H. 1971. Ecologicalfactors affecting wa- polyandrousSpotted Sandpiper. Behav. Ecol. So- terfowl production in the Saskatchewanpark- ciobiol. 10: 185-191. lands. U.S. Fish Wildl. Serv. Resources Publ. No. PERRINS,C.M. 1970. The timing of birds' breeding 99. seasons. Ibis 112: 242-255. ß 1982. Habitat use and productivity of Can- PIANKA,E. R. 1976. Natural selection of optimal re- vasbacks in southwestern Manitoba, 1961-72. U.S. productive tactics.Amer. Zool. 16: 755-784. Fish Wildl. Serv.Spec. Sci. Rept.,Wildl. No. 248. --, & W. S. PARKER.1975. Age-specificrepro- STREET,g. B., &: B. F. FINDLAY. 1981. An objective ductive tactics. Amer. Natur. 109: 453-464. climatological study of prolonged dry spells ROCKWELL,R. F., C. S. FINDLAY, & F. COOKE. 1983. (meteorological drought) in the prairie prov- Life historystudies of the LesserSnow Goose(An- inces.Can. Climate Centre Rept. No. 81-10. ser caerulescenscaerulescens). I. The influence of STROHMEYER,D. L. 1967. The biology of renesting age and time on fecundity. Oecologia 56: 318- by the Blue-winged Teal (Anas discors)in north- 322. west Iowa. Unpublished Ph.D. dissertation.St. ROGERS,J.P. 1959. Low water and LesserScaup re- Paul, Minnesota, Univ. Minnesota. productionnear Erickson, Manitoba. Trans. North SUGDEN,L. G. 1973. Feeding ecology of Pintail, Amer. Wildl. Nat. Resources Conf. 24: 216-224. Gadwall, American Widgeon and LesserScaup 1962. The ecologicaleffects of drought on ducklingsin southernAlberta. Can. Wildl. Serv. reproductionof the LesserScaup, Aythya affinis Rept. Ser. No. 24. (Eyton). Unpublished Ph.D. dissertation.Colum- , & H. J. POSTON. 1968. A nasal marker for bia, Missouri, Univ. Missouri. ducks.J. Wildl. Mgmt. 32: 984-986. 1964. Effectof drought on reproductionof --., & --. 1970. A raft trap for ducks.Bird- the LesserScaup. J. Wildl. Mgmt. 28: 213-222. Banding 41: 128-129. --, & L. J. KORSCHGEN. 1966. Foods of Lesser SUNDE,L. g., &:J. BARICA.1975. Geographyand lake Scaups on breeding, migration and wintering morphometry of the aquaculture study area in areas.J. Wildl. Mgmt. 30: 258-264. the Erickson-Elphinstonedistrict of southwest- RYDER,J.P. 1980. The influence of age on the breed- ern Manitoba. Environ. Can., Fish. Marine Serv. ing biology of colonialnesting seabirds.Pp. 153- Tech. Rept. No. 510. 168 in Behavior of marine animals, Vol. 4 (J. TITMAN,g. D. 1983. Spacingand three-bird flights Burger,B. L. Olla, & H. E. Winn, Eds.).New York, of Mallards breeding in pothole habitat. Can. J. Plenum Publ. Corp. Zool. 61: 839-847. SALKI,A.G. 1981. The crustaceanzooplankton com- TRAUGER,D. L. 1971. Population ecologyof Lesser munities of three prairie lakessubject to varying Scaup(Aythya affinis) in subarctictaiga. Unpub- degrees of anoxia. Unpublished M.S. thesis. lished Ph.D. dissertation. Ames, Iowa, Iowa State Winnipeg, Manitoba, Univ. Manitoba. Univ. SAUNDERS,A.A. 1959. Forty years of spring migra- 1974. Eye color of female LesserScaup in tion in southern Connecticut. Wilson Bull. 71: relation to age. Auk 91: 243-254. 208-219. WELLER,M.W. 1957. An automatic nest-trap for wa- SAYLER,R. D., &: A.D. AFTON. 1981. Ecologicalas- terfowl. J. Wildl. Mgmt. 21: 456-458. pectsof CommonGoldeneyes Bucephala clangula WILLIAMS,G.C. 1966. Adaptation and natural selec- wintering on the upper Mississippi River. Ornis tion. Princeton, New Jersey, Princeton Univ. Scandinavica 12: 99-108. Press.