forced copulation in captive mallards iii. sperm competition

FORCED
COPULATION
III. SPERM
IN CAPTIVE
COMPETITION
MALLARDS
KIMBERLYM. CHENG,1 JEFFREY
T. BURNS,AND FRANKMcKINNEY
Department
of Ecology
andBehavioral
Biology,BellMuseumof NaturalHisto•,
Universityof Minnesota,Minneapolis,
Minesota 55455USA
ABSTRACT.--Previous
observationsof forced copulation (FC) in captive Mallards (Anas
platyrhynchos)
showedthat most FC attemptswere directedat femalesin prelayingand
layingconditionand thatmostFC'soccurredin the morningwhen the femaleswereleaving
their nestsafter egg laying. In order to determinewhether or not there is a physiological
basis for these observedtemporal patterns, sperm competition in captive Mallards was
examinedusingartificialinseminationand geneticmarkers.Resultsindicatedthatif a female
was inseminatedwith two competingdosesof semenat different time intervals,the proportionof progenyfromthefirstandthe secondinseminations
wasnot significantlydifferent
if these inseminationswere simultaneous,I h, or 3 h apart. There was a preponderanceof
progeny(70%)fromthe secondinsemination,however,if the inseminations
were6 h apart.
Inseminationof femaleslessthan I h after egg laying resultedin 25% of the eggslaid the
followingmorningbeing fertile. Only I of 179eggslaid the followingmorningwas fertile
when the femaleswere inseminatedmore than i h after egg laying.
Our experimentdemonstratedthat there is an insemination"window," a short period
when new spermareleastlikely to meet competitionfrom spermalreadyin the oviductand
from spermintroducedlater, and it provided a possibleexplanationfor the observedtiming
of FC attempts.Received12 May 1982,accepted
3 December1982.
PARKER(1970: 527) defined sperm competi- species(the Mallard, Anas platyrhynchos)in
tion as "the competitionwithin a singlefemale which pairedfemalesare inseminatedby males
betweenthe spermfrom two or more malesfor other than the mate during FC, thus providing
the fertilization of the ova." Although best an opportunityto investigatethe mechanism
known in insects (Parker 1970), this phenom-
enon occursalsoin many other animal groups
(Allison 1977, Bertram 1976, Smith in press).
Amongbirds, spermcompetitionundoubtedly
occursin a number of polyandrousspecies,in
which one female may copulatewith several
males (Jenni 1974), and probably occursin a
number of promiscuousspeciesas well (Wittenberger1979). While most speciesof birds
are consideredto be monogamous(Lack 1968)
and copulationsoccurprimarilybetweenmates,
there is evidence that sperm competition may
occur in some of these speciesalso. For example, paired femaleshave been observedto
acceptor solicitcopulationsor to be subjected
to forcedcopulations(FC)frommalesotherthan
their mates (Gladstone 1979, McKinney et al.
in press). This paper deals with one duck
• Presentaddress:Department of Poultry Science,
The University of British Columbia, Vancouver,Brit-
of sperm competition.
Earlier papersin this serieson captiveMal-
lards documentedthat eggscouldbe fertilized
by FC (Bumset al. 1980)and describedtemporal relationsbetweenFC and femalebreeding conditionandbehavior(Chenget al. 1982).
In the latter analysis,it was shownthat almost
all FC's were directed at females in the laying
phase and that most FC attemptsoccurredin
the morning hours, especiallywhen females
were detectedleaving their nests.Basedon the
ovulation pattern and sperm-storagemechanism
demonstrated
in
domestic
chickens
(Comptonet al. 1978,Sturkie 1976),Cheng et
al. (1982)hypothesizedthat Mallard maleswere
attemptingFC's at that time of the day when
their sperm would competemost effectively
with spermfrom pair copulations.
In studyingsperm-storage
mechanisms
of the
uterovaginal(UV) glandsin chickens,Compton et al. (1978) showed that, if hens were
ish Columbia V6T 2A2, Canada.
3O2
subjectedto artificialinsemination(AI) with
semen from one type of roosterand then reThe Auk 100:302-310.April 1983
April1983]
Sperm
Competition
in Mallards
inseminated4 h later with an equalamountof
semen from roosters with a different genetic
marker, 80% of the progenyresultedfrom the
secondinsemination.If the two types of male
gameteswere mixed in equal proportionsbefore AI, however, the resultingphenotypesof
progenyapproximateda 1:1 distribution. Data
from this experimentsuggestthat spermfrom
differentinseminationssequentiallyfill the UV
glandsand producea stablestratificationof
spermcells,so that the most recentlyinseminated sperm stay on top and mixing of sperm
cells from different
inseminations
does not oc-
cur within a gland. Spermfrom the top layer
are
also
the first
released
for
fertilization
(DeMerritt 1979), and so most of the progeny
are sired by the male that performedthe most
recent
insemination.
If a similar
mechanism
were to exist in the Mallard, then even though
males were attempting FC on females during
their layingperiod,spermfrominfrequentFC's
would be likely to be coveredby sperm from
repeatedpair copulationsand would not compete effectivelyin fertilizing eggs.
303
males.Thus, progenypossessing
wild-type plumage
could be attributed to sperm from wild-type males;
those with white plumage could be attributed to
sperm from recessivemales.
Semenwas collectedfrom the malesby the mas-
sagemethodof Chengand Otis (in prep.), which is
a modificationof the techniquedevelopedfor A! of
chickens and turkeys (Burrows and Quinn 1937).
Fresh undiluted semen was inseminated intravaginally by everting the oviduct (Kinney and Burger
1960). Pooled semen samplesfrom a minimum of
three malesof the samegenotypewere used for all
inseminations
to reduce individual
male effect. Fe-
maleswith a hard-shelleggin the oviductat the time
of inseminationwere not used. For 14 daysafter each
seriesof inseminations,eggswere collecteddaily,
identifiedby femaleand date,and storedat 10øCand
65% humidity (Chenget al. 1980).They were then
incubated
in a Robbins
Model
IHA
electric forced-
air incubator and were candied on the 7th, 14th, and
23rd days to determineembryoviability. Eggscon-
taillingviableembryosweretransferred
on the 23rd
day of incubationto individualhatchingbasketslocated in a hatcher attached to the incubator.
At the
time of the first candling,any apparentlyinfertile
eggswerebrokenout and macroscopically
classified
as an early embryonicdeathor as infertile (Kosin
On the other hand, ovulation in chickens
1944).Dead embryosdetectedduring the secondand
normallyoccurswithin 15--75min after the lay- third candlingwerealsorecorded.Becauseducklings
ing of the previous egg (Sturkie 1976). The homozygousfor the recessivewhite gene(i.e. those
ovulated ovum remains fertilizable for only ducklingssired by white males)have yellow down,
about 15 min before albumen is deposited whilethosesiredby wild-typemaleshavethe typical
around the yolk (Gilbert 1971).If FC attempts yellow and brown down pattern,paternitycouldbe
were so timed that sperm from such insemi- determinedfor all ducklingsat hatchingand for emnations reached the infundibulum (the site of bryosthat died after 19 daysof incubation.The exfertilization)at the time of ovulation,they might perimentswere conductedat the facilitiesof the Departmentof Animal Science,Universityof Minnesota,
have a good chanceof fertilizing the egg that St. Paul, Minnesota.
would be laid the next morning. Sperm from
pair copulationslater in the day would not be
EXPERIMENT I
in time to competefor fertilization of this particular egg.
A pilot study using White Pekings and
The purposeof this study was to determine
whether or not (1) sperm storageand utiliza- Rouens (domestic breeds of Mallard) was conductedto developAI techniquesand to obtain
tion mechanisms in the Mallard are similar to
thosein the chickenand (2) spermintroduced preliminary data. We inseminated16 White
into the femaleshortlyafter ovipositioncan ef- Peking (recessive)femaleswith semen from
fectivelyfertilize the egg to be laid on the fol- White Peking males and Rouen (wild-type)
males (and viceversa)at varioustime intervals,
lowing morning.
and the data collected (Table 1) appeared to
GENERAL MET•OVS
In order to determine how sperm from different
inseminations compete for fertilization of the ova,
agreewith thosereportedfor chickens(Compton et al. 1978). Problems existed with regard
to the adaptiveness
of White Pekingfemalesto
wire-floor
cages,
however.
Becauseof their
the recessivewhite plumage gene of the Mallacd
(Lancaster1963)was used as a geneticmarker. Fe- body weight, some femalesdevelopedfoot
maleshomozygous
for this allelewereartificiallyin- problems,and eggswere brokenas a resultof
seminatedwith semenfrom the homozygousdom- femalessteppingon them.
inant(wild-typeplumage)malesandrecessive
(white)
To obtain more conclusive evidence on the
304
CHENG,BURNS,AND McKINNEY
[Auk, Vol. 100
TABLE1. Comparisonof the phenotypicratio of progenyfollowinginseminationswith two typesof semen
(Peking and Rouen) at different time intervals.
First
Second
Time
Number of progenya
Treatment
insemination
insemination
interval (h)
White
Wild-type
1
2
Peking
Rouen
Rouen
Peking
I
I
6
7
5
4
3
Peking
Rouen
5
11
4
Controls
Rouen
Peking
Mixed semen
6
6
0
11
23
Ratiob
10:12
5
26
10:22c
(23:26)
Females that did not lay any fertile eggsafter the inseminations were not included.
Ratioof thenumberof progenyresultingfromthe firstinsemination
versusthenumberof progenyresultingfromthe secondinsemination.
X• - 4.50, basedon the expected1:1 ratio, X•005with 1 df - 3.84.
relative effectivenessof competing insemina-
1. There were basicallyfour treatmentswith four fe-
tions at different time intervals, a further ex-
males in each treatment: (1) females were insemi-
nated simultaneouslywith both GF and DW semen
(8, 16, 20 and 23), (2) females were subjectedto sequential AI 1 h apart (6, 7, 9, and 10), (3) females
(DW) Mallards. These two "breeds" were used
were subjectedto sequentialA! 3 h apart (4, 5, 11,
rather than White Pekingsand Rouensbecause and 12), and (4) femaleswere subjectedto sequential
periment on a larger scalewas conductedusing
game-farm (GF) Mallards and domesticwhite
of their small body size (adult DW females AI 6 h apart (2, 3, 13, and 14). Within Treatment 1,
weighed about 1,450 g) and reliably high rate femalesreceiveda single insemination consistingof
of egg production.
0.2 ml semen from both GF and DW males mixed
in
equal volume. Within each of Treatments 2, 3, and
METHODS
4, two of the four females were each inseminated with
0.1 ml of pooled semenfrom DW malesand re-insemi-
Two hundred 1-day-old DW Mallard ducklings nated with 0.1 ml of pooled semen from GF males
were purchasedfrom the Pietrus Hatchery (Sleepy (Part A of scheme). The other two females in each
Eye, Minnesota)in June1979.The PietrusHatchery treatment were first inseminated with semen from
obtained
their stock from a farm in Iowa and has
maintained a breeding flock of about 300 DW Mallardsfor four generationswith a maleto femaleratio
of 1:5. Althoughthe DW Mallard is a differentbreed
from the White Peking,they arehomozygous
for the
same white plumage gene. Sixty male GF Mallard
(wild-type) ducklingswere also obtainedfrom the
GF males and re-inseminated
with
semen from
DW
males according to the same schedule (Part B of
scheme).In caseswhere there was only one insemination, 0.2 ml of semen was used in order to be
comparable with caseswhere two separate doses of
0.1 ml were given. The common dosagefor AI in
geeseis 0.05 ml of undiluted semen (Kinney and
Burger 1960, Kurbatov et al. 1976).AI dosagesranging from 0.01 to 0.025 ml of undiluted semenyield
good fertility in Muscovy ducks(Huang and Chow
1974), and a dosage of 0.1 ml diluted semen (1 part
Cheng et al. 1980). The ducklingswere raised to- of semen to 3 parts of extender) per insemination
getherundera daily8-h light schedulein a 12.2-m x resulted in good fertility in Peking ducks (Davtyan
6.1-m floor pen until the end of January1980,when and Starygin1974).The dosagewe were usingwould
insure the introduction of adequate sperm numbers.
they were moved to laying cages.
Femaleswere housedindividuallyin 31-cmwide x It is not known how much semenis normallytrans~
41-cm high x 46.5-cm deep chicken laying cages. ferredthroughnaturalmating.In our study,the mean
Cage-floorswere slantedsothat eggslaid would roll volume of ejaculatethrough the massagemethod was
to a troughoutsidethe cage.We kept 50 malesof each 0.12 ml (Cheng and Otis in prep.).
Treatment 1 and the reciprocal order of insemibreed, four to a cage, in 76-cm x 92-cm x 76-cm
cages. The lighting schedule for all birds was in- nations in the two parts of the scheme insure that
creasedto 15 h (0500-2000)of light per day, which any differencein the effectiveness
of competitionberesulted in most (75/96) of the females commencing tween the two types of semen can be detected and
egg production 10 days after the photoperiodwas controlled.To controlfurther for fertility differences
increased.Femaleswere divided into three groups, and the remote possibility that someGF males used
each consistingof 24 experimentalfemalesand 6 re- may havebeen heterozygousfor the wild-type gene,
serves.The experimentalfemalesin eachgroupwere 0.2 ml of each pooled semen sample was used to
inseminatedaccordingto the schemeshown in Fig. inseminate the eight females (1, 15, 17, 18, 19, 21,
Northern Prairie Wildlife Research Center in Jamestown, North Dakota. These GF Mallards were descendentsof wild birds that were captured in 1949
at the Frost Game Farm in Wisconsin (Frost 1972,
April1983]
Sperm
Competition
inMallards
305
•r22
POOLEO
2,3
•
GF
:EMEN
•19
t0•2m•
POOLED
Dw
•4,
POOLED
S •
GF
SEMEN
EMEN
TO2ml
•'6,7 •
•'8
POOLEO
GFSEMEN
•'16
'23
POOLED
ß
tO'9,
10'•'•"•=DW
SEMEN
70
2ml
•-18
POOLEO
GF
------
1
POOLEO
) •'11, 12 •]
OW
SEMEN
SEMEN
70.2
ml
•-21
•13,
14
<•
•-15
OHR.
POOLEO
DW SEMEN
70.2
ml
•-24
0 't'IHR. 0 +3HE. O+6HR.
Fig. 1. Schedule
of artificialinseminations.
Exceptwherespecified
otherwise,
eachfemalereceived
a
doseof 0.1 ml of semenper insemination.
22, and 24) not used in the four treatments.Eggs was not significantlydifferent from the hatch-
were collectedfor 14 days after the inseminations
ability of those(148)fertilizedby GF semen
and then the femaleswere randomlyreassigned
(85%).Thefertilityof eggsfromfemalesover
within groupsfor the nextreplication.
The experi- the 14-dayperiod after inseminationwith GF
ment was repeatedoncewith all 3 groupsof females
and againwith only2 groups(i.e. altogether
8 replications).In total,2,338eggsweresetfor incubation.
semen,however,wasonly47.4%(148/312)and
wassignificantly
lowerthanthatof eggsfrom
those females inseminated with DW semen
(61.5%;206/335)
(Chengand Otis in prep.).
RESULTS
Apparently,
thiswasmainlycaused
by differFertilityandhatchability.--Data
collected
from encesin durationof fertilityoverthe 14-day
control females1, 15, 17, 18, 19, 21, 22, and 24
of each replicationshowedthat the hatchability of 206 eggsfertilized by DW semen(86%)
period.Only one of the eggscollectedon the
firstdayafterAI wasfertile.Fertilityof eggs
from femalesinseminatedwith DW semenwas
306
CHENG,BURNS,AND McK•NNEY
[Auk, Vol. 100
TABLE
2. Comparison
of phenotypic
ratioof progenyfor days2-7 followingsequential
AI with genetically
marked (GF, DW) semen.
White vs.
Treatment
SequenceTime Number
ofprogenywild-type
ratio
of AI
interval (h)
White
Wild-type
First AI vs.
secondAI
ratio
I
Mixed
0
52
49
52:49
(52:49)
2
GF-DW
DW-GF
1
1
12
21
12
24
33:36
33:36
3
GF-DW
DW-GF
3
3
16
24
21
16
40:37
45:32
4
GF-DW
DW-GF
6
6
20
18
36
38:41
23:56•
5
13.78, basedon the expected1:1 distribution, X%.otwith I df = 6.63.
above80% from day 2 to day 8, while that from 0.005) white progeny than wild-type progeny
females inseminated with GF semen was only in the secondperiod(days8-14, Table3). This
above 80% from day 2 to day 5. Thereafter, is consistent with the observations from confertility of eggs from these femalesdeclined trol females that the duration of fertility was
steadily.
longerfor DW than for GF semen.
To avoid the bias causedby the difference
Phenotypicratio of progeny.--Asthere were
no significant differences within each treat- in the duration of fertility between the two
ment in the phenotypic ratio of progeny be- typesof semen,only datacollectedduring the
tween replications, data from all eight repli- period from 2 to 7 dayspostinseminationwere
cationswere pooledfor analysis.Becauseof the consideredin estimating the proportion of
differencein the duration of fertility between progenyresulting
fromsequential
AI. As shown
the two typesof males,datafor days2 to 7 and
days 8 to 14 are summarized separatelyand
presentedin Tables2 and 3, respectively.
Equal
weight was given to eachindividual femalein
in Table2, the phenotypicratio of progenyresultingfrom AI with mixed semen(Treatment
1) and inseminations1 h apart (Treatment2)
approximatedthe expected1:1 ratio and was
calculatingthe progenyphenotypicratio. The consistent with the observations made in the
phenotypicratio of wild-type GF progeny to pilot study. Furthermore,the 1:1 ratio was
white progeny in all four treatmentsdid not maintained even when the inseminations were
differ from the expected1:1 ratio for the period 3 h apart (Treatment3). Followingsequential
of days2-7 (Table2). On the otherhand, there inseminations6 h apart (Treatment4), howmoreprogeny(70.9%)were
weresignificantly
more(X'•= 20.3,df = 1, P < ever,significantly
TABLE3. Comparisonof phenotypicratio of progenyfor days8-14 followingsequentialAI with genetically marked (GF, DW) semen.
Treatment
I
Sequence
of AI
Mixed
Time
interval (h)
Number
ofprogeny
White
vs.
White
Wild-type
wild-type
ratioa
FirstAI vs.
secondAI ratio
0
18
16
18:16
(18:16)
2
GF-DW
DW-GF
1
I
6
7
8
15
13:23
15:21
3
GF-DW
3
3
24
12
8
36:15
32:19
4
GF-DW
6
33
7
66:15
DW-GF
6
33
8
(32.11)**
DW-GF
Total
7
(8.65)**
133: 69
(20.28)**
(in parentheses)
basedon the expected1:1 distribution;** P < 0.01•
40:41
April 1983]
3O7
SpermCompetitionin Mallards
TABLE4. Fertility of eggsfrom inseminationstimed with egglaying.
Number
Time of insemination
Groups
Number of
eggsfertile
168
1
Expt. 1
More than 1 hour after laying
of
eggslaid
next morning
Expt. 2
Control
Re-insemination
Total
7
4
0
0a
179
1
Expt. 2
Lessthan 1 hour afterlaying
Treatment
1
Re-insemination
Treatment 2
Re-insemination
Total
13
3
7
12
4
5a
1
0a
36
9
Numberof eggsfertilizedby spermfromlastinsemination.
attributed
to the second of the two insemina-
RESULTS
tions.
The resultsof Experiment2 are summarized
EXPERIMENT
2
in Table 4. Fourteen of the 20 females insemihated with DW semen in Treatment 1 laid in
Mostfemaleslaid eggsbetween0500and0900 the morning following the day of AI, and 3 of
each day. In Experiment1, the first insemina- the 14 eggs were fertile. Subsequent data
tion was carried out at 1000, which would be showedthat one of the femalesdid not lay any
an hour or more after egg laying for most fe- fertile eggsduring the 1-week period after the
males. Of the 168 eggscollectedon the morn- inseminations, and her record was excluded
ing following the day of inseminations,only from the summaryof data. Therefore,in effect,
oneeggwasfertile.The purposeof Experiment 3 of 13 eggs(23.1%) were fertile.
2 was to determinewhether or not eggslaid
Of the 20 femalesassignedfor Treatment2,
the morning following the day of insemina- we were able to observelaying time for only
tions could be fertilized if females were insem18 on the morning of AI. Thereforeonly 18 females were inseminated
with semen from GF
inatedwithin an hour after laying.
males. Of the 13 eggs collectedthe morning
METHODS
after, only 1 was fertile. One female in this
group was infertile, and her record was exFifty DW femalesthat had been laying infertile cluded (i.e. in effect, 1 of the 12 eggs (8.3%)
eggs(i.e. femalesthat had not been subjectedto AI
within a 2-week period) were used. Thesebirds were
under the same 15-h light scheduleas those in Ex-
was fertile).
Seven control females were inseminated
more
periment1. Layingtime for eachfemalewas record- than an hour after their laying, and none of the
ed in the morning, and 20 femaleswere individually eggslaid the following morning was fertile.
For re-inseminations 3 days later, we were
inseminatedwithin 1 h after laying with 0.2 ml of
pooledsemenfrom DW males(Treatment1). Another
20 females were inseminated with 0.2 ml of pooled
able to record laying time for 14 of the 20 females from Treatment 1 (females inseminated
semenfrom GF males, alsowithin 1 h after laying
(Treatment2). The remaining 10 femaleswere used
with DW semen). Eight eggs were collected
as controlsand were inseminated during the period
men. One egg was from the infertile female,
between1-2 h after their laying.
Threedaysafterthe first setof inseminations,laying times for thesefemaleswere again recorded.Fe-
from
the 14 females
inseminated
with
GF se-
and we hatchedout 5 wild-type ducklingsand
2 white ducklings. Sperm from this later inseminationfertilized 5 of the 7 eggslaid the
mmes that had been inseminated
with DW semen 3
days previously(Treatment 1) were re-inseminated morning afterAI despitethe presenceof sperm
with GF semen and females in Treatment 2 were refrompreviousinseminationsin theUV glands.
inseminatedwith DW semenwithin 1 h afterlaying.
Time of laying was recordedfor 12 females
308
from
CHENG,
BURNS,
ANDMcKINNEY
Treatment
2. These
females
were
re-in-
[Auk,Vol. 100
seminatedwith DW semen and five eggswere
collectedthe following morning. One embryo
inations I h and 3 h apart probably mixed in
the oviductbeforeenteringthe UV glandsand
resulted in an equal probability of fertilizing
died within the first week of incubation, there-
ova in subsequent ovulations.
by precludingdeterminationof its phenotype.
Observationsof captiveMallardsshowedthat
Of the remainingfour eggs,all producedwild- the frequencyof apparentlysuccessful
FC was
type ducklings.None of the four eggswas fer- low comparedto the frequencyof pair coputilized by spermfrom the last insemination.In lations (Cheng et al. 1982). Therefore, a dose
total, with semen from previous insemina- of semenfromFC probablyhasto competewith
tionsalreadypresentin the oviduct,spermfrom more than one dose of semen from pair coputhe last insemination fertilized 5 of the 11 eggs lationsduring the prelayingand laying period
of the female. Near the time of ovulation, how(45.5%) laid the following morning.
Five of the control females were re-insemiever, recentlyinseminatedsperm can traverse
natedbetweenI and 2 h afterlaying. Of thefour the UV junction and reach the infundibulum
eggscollectedand hatched, none was fertilized
by the last insemination.
Combiningthe datafromboth treatmentsand
all inseminations, it was found that sperm in-
seminated within an hour after laying fertilized 9 of 36 eggs (25%) laid the morning after
the day of AI (Table4).
within a few minutes (Mimura 1939, Bobr et
al. 1964, Howarth 1971). This is the only time
when sperm transport in the oviduct is not obstructed by the presenceof an egg (Morzenti
et al. 1978). Data from Experiment 2 showed
that sperm from 9 of 36 (25%) inseminations
administeredwithin an hour after laying were
successfulin fertilizing the egg to be laid the
DISCUSSION
next morning. Although the percentageof sucSome birds with large clutch sizes (e.g. cesswas not high, this resultshowedthat there
chickens and turkeys) have sperm-storage is an insemination "window," a short period
glands in the oviduct, in which the functional when new sperm are least likely to meet comcapacityof the sperm is prolonged and from petition from spermalreadyin the oviduct and
which spermare released(Lorenz1966,Comp- from sperm introduced later.
ton and Van Krey 1979).Poultrybreedershave
There is someevidenceto suggestthat male
known for a long time that sperm competition Mallards are timing their FC attempts in relacan take placein chickensand turkeys,and a tion to this especiallyfavorable period. In a
good deal of experimentationon this topic has flight-pen study, most FC attempts occurred
been carriedout in conjunctionwith maximiz- during the morninghours(when eggsare laid),
ing fertility by AI (e.g. Allen and Champion and they were directed especiallyat females
1955, Payne and Kahr 1961, Reinhart and Je- leaving their nests(Cheng et al. 1982).The time
rome 1971, Lake 1975, Classen and Smith 1975, intervals between laying, ovulation, and deCompton et al. 1978, DeMerritt 1979). Many parture from the nest have not been studied in
studiesindicatedthat, in general,the most re- ducks, but there is some information on the
centof competinginseminationsis likely to be total time spent at the nest on the dayswhen
the mosteffectivein fertilizingeggs,not only eggs are laid. Northern Shovelers(Anas clybecausefertility declineswith the ageof sperm peata) studied by Afton (1977, 1980)spent 94in the UV glandsbut alsobecauseof the way 107 min on the nest during laying of the first
theseglandsare filled and how spermare sub- three eggs of the clutch, but thereafterthey
sequentlyreleased.DomesticMallard females spent increasinglylonger periods on the eggs
can storeviable spermfor up to 17 days(Elder (3-4 h for the 4th and 5th eggs,up to about 12
and Weller 1954, Ash 1962), and data from four h on the day the last eggswere laid). On three
domestic
and
semi-domesticated
breeds
of
Mallard in our study showedthat sperm competition alsotakesplacein thisspeciesin a way
similar to that in chickensand turkeys.Sperm
from the morerecentof two competinginseminations as closeas 6 h apart fertilized 70% of
the eggslaid subsequently.Spermfrom insem-
occasions,Afton (1977) flushed a female 1-2 h
after her arrival
at the nest and found
that a
new egghad alreadybeen laid, suggestingthat
laying occurssoon after the bird arrives. Data
for Mallardsfor the last half of the laying period (Caldwell and Cornwell 1975) suggesta
similar pattern of increasing duration of nest
April1983]
Sperm
Competition
in Mallards
attendanceduring laying. Thesefindings suggestthat, when malesare successful
in forcing
copulationon femalesleaving the nest, these
inseminationsmight coincidewith the favorable postovulation"window" only on the first
few days of laying. This possibilitydeserves
further study.
If FC is an evolved strategyand is effective
in fertilizing eggs, mate-guardingto forestall
FC attempts would be expectedas a counteradaptation. We do know that males attempt to
defend their mates by fighting or trying to dislodgemalesattemptingFC (McKinneyet al. in
press), but, when many males are involved in
FC attemptson the samefemale simultaneously, this is not always effective in preventing
FC. Whether the mate has additional strategies
to reducethe probability of eggsbeing fertilized by sperm deposited by FC is largely an
unexplored question. Such strategiescould revolve around the frequencyand timing of pair
copulations (PC). There is no good information, however, on the timing and frequencyof
PC in relation to female reproductivephysiology in Mallards.Paired malesfrequentlysolicit
copulationfrom their mates, but few are successful. Observations of Mallards (Barrett 1973,
309
ACKNOWLEDGMENTS
Our research on sperm competition in waterfowl
was supportedby the National ScienceFoundation
(Grant BNS-7924692).Jack Otis, Department of Animal Science, assisted in artificial inseminations and
the developmentof the A! techniques.R. N. Shoffner
and the Department of Animal Scienceprovided the
cageand incubationfacilities. P. B. Siegelprovided
helpful suggestionsfor the inseminationscheme.We
alsoacknowledgethe supportof D. F. Parmelee,Field
Biology Program, and thank P. B. Siegel, M. G. Anderson, G. A. Parker, R. O. Bailey, and an anonymous reviewer
for critical
comments
on the manu-
script.
LITERATURE
CITED
AFTON, A. D. 1977. Aspects of reproductive behavior in the Northern Shoveler. Unpublished
M.S. thesis. Minneapolis, Minnesota, Univ.
Minnesota.
1980. Factorsaffectingincubation rhythms
of Northern
Shovelers.
Condor
82: 132-137.
ALLEN, C. J., & L. R. CHAMPION. 1955. Competi-
tive efficiency of turkey sperm. Poultry Sci. 34:
1598-1604.
ALLISON,A.J. 1977. Flockmating in sheep. II. Effect of number of ewes per ram on mating behavior and fertility of two-tooth and mixed-age
Romneyewesrun together.New ZealandJ. Agr.
Barash1977)and severalother dabbling ducks
(McKinney et al. in press)indicate that paired
Res. 20: 123-128.
malessometimesforce copulationon their own ASH, W.J. 1962. Studies of reproductionin ducks.
matesafter they have been subjectedto FC as1. The duration of fertility and hatchability of
White Pekin duck eggs. Poultry Sci. 41: 1123saults. Our findings on sperm competition
1126.
suggestthat a prompt"forcedpair copulation"
BARASH,
D. P. 1977. Sociobiologyof rape in Mal(FPC)would be especiallyadvantageous
for the
lards
(Anas platyrhynchos):responsesof mated
paired male if FC had occurred during the
male. Science 197: 788-789.
postovulation period when fertilization takes
BARRETT,J. 1973. Breeding behavior of captive
place.At othertimes, however,the mate might
Mallards. Unpublished M.S. thesis. Minneapobenefit more by waiting and performing PC
lis, Minnesota, Univ. Minnesota.
after an interval of several hours. Apparently
BERTRAM,B.C. R. 1976. Kin selection in lions and
FPC's are not invariably performedafter FC's,
and, if the female resists,the attempt to mount
in evolution. Pp. 281-301 in Growing points in
ethology(P. P. G. BatesonandR. A. Hinde, Eds.).
Cambridge, England, Cambridge Univ. Press.
is often aborted.
Our experiments demonstrated that sperm BOBR, L. W., F. X. OGASAWARA,• F. W. LORENZ.
1964. Distribution of spermatozoain the ovicompetition can take place in Mallards and
duct and fertility in domesticbirds. II. Transport
provided a physiologicalbasis for a possible
of spermatozoa in the fowl oviduct. J. Reprod.
explanationof some of the temporal patterns
Fert. 8: 49-58.
of FC behavior observed in our flight pens. BURNS,J. T., K. M. CHENC, & F. MCKINNE¾. 1980.
These experiments raise interesting questions
Forcedcopulation in captive Mallards. I. Fertilabout the degreeto which behavioralstrategies
ization of eggs. Auk 97: 875-879.
of malesare "finely tuned" to the reproductive BURROWS,
W. H., & J. P. QUINN. 1937. The collecphysiologyof females,and, in particular, they
tion of spermatozoafrom the domesticfowl and
turkey. Poultry Sci. 16: 19-24.
lead to predictions that can guide future behavioral studies of PC, FC, and FPC.
CALDWELL,P. J., & G. W. CORNWELL. 1975. In-
310
CHENG,
BURNS,
ANDMcKINNEY
cubation behavior and temperaturesof the Mallard Duck. Auk 92: 706-731.
CHENG, K. M., R. N. SHOFFNER,R. E. PHILLIPS, &
F. B. LEE. 1980. Reproductiveperformancein
wild and game-farmMallards.Poultry Sci. 59:
1970-1976.
ßJ. t. BURNS,& F. McKINNEY. 1982. Forced
copulationin captiveMallards. II. Temporal factors. Anim.
Behav. 30: 695-699.
KOSIN,I.L. 1944. Macro- and microscopic
methods
of detectingfertility in unincubatedeggs.Poultry Sci. 23: 266-269.
KURBATOV,A.D.,
itive fertilizationin the domesticfowl. Poultry
Sci. 54: 1748.
COMPTON,M. M., & H. P. VAN KREY. 1979. Emptying of the uterovaginalsperm storageglands
in the absenceof ovulation and oviposition in
the domestichen. Poultry Sci. 58: 187-190.
, --,
& P. B. SIEGE•,.1978. The filling
and emptying of the uterovaginalsperm-host
glandsin the domestichen. PoultrySci. 57: 16961700.
& M. I. STAXYGIN. 1974. Artificial
inseminationin duck production.Proc.and Abs.
15th World's Poultry Congr. New Orleans:384385.
Artificial Insemination, Cracow 6: 1009-1012.
LACK,D. 1968. Ecologicaladaptationsfor breeding
LIKE, P. E. 1975. Gameteproductionand the fertile
period with particular reference to domestic
birds. Symp.Zool. Soc.,London53: 225-244.
LANCASTER,
F. M. 1963. The inheritanceof plumage color in the common duck. Bibliographica
Genetica
DEMERRITT,
R.J. 1979. The role of the uterovaginal
junction in sperm cell storageand releasein the
domesticfowl. Poultry Sci. 58: 1049.
1954.
Duration
of fertility in the domesticMallardhen afterisolation from the drake. J. Wildl. Mgmt. 18: 495502.
FROST,J. 1972. Observations from commercial wa-
terfowlproducers.Pp. 71-72 in Symposiumfor
the role of hand-reared
ducks in waterfowl
man-
agement.Dundee, Illinois, Max McGraw Wildl.
Found.
29: 317-404.
LORENZ,F.W.
1966. Behavior of spermatozoain
the oviductin relationto fertility. Pp. 39-43 in
Physiologyof the domesticfowl (C. HortonSmith and E. C. Amorosa, Eds.). Edinburgh,
Oliver & Boyd.
McKINNEY, F., S. R. DERRICKSON,& P. MINEAU.
In
press. Force copulation in waterfowl. Behaviour.
MIMURA, H.
ELI•ER, W. H., & M. W. WELLER.
P. G. TSARENKO,•r I. I. POPOV.
1976. Improvingartificialinseminationin geese.
Proc.8th Intern. Congr.Anim. Reproductionand
in birds. London, Methuen & Co.
CLASSEN,
H. L., & J. R. SMITH,JR. 1975. Compet-
DAVTYAN, A.D.,
[Auk,Vol. 100
1939. On the mechanism of travel of
spermatozoathroughthe oviductin the domestic fowl, with special referenceto artificial insemination. Sonderabdruckand Okjimas Folia
Anatomica Japonica. 17(5). [Cited by Sturkie
1976.]
MOaZENTI, A., F. X. OGASAWAKA,•: C. L. FUGUA.
1978. The relationship of infundibula fluid to
the avian sperm transport mechanism. Poultry
Sci. 57: 1173.
PARKER,G. A. 1970. Sperm competition and its
evolutionaryconsequences
in insects.Biol. Rev.
45: 525-567.
GI•,BERT,
A.B. 1971. Transportof eggsthroughthe
oviductand oviposition.Pp. 1345-1352in Physiologyand biochemistryof the domesticfowl (D.
J. Bell and B. M. Freeman, Eds.). New York, Academic Press.
G•,AI•STONE,
D. E. 1979. Promiscuity in monogamous colonial birds. Amer.
Natur.
114: 545-557.
PAYNEßL. F., & A. J. KAHR. 1961. Competitive ef-
ficiencyof turkey sperm.PoultrySci. 40: 15981604.
REINHART,B. C., & F. N. JEROME.1971. Competition between the pooled spermatozoaof two
breeds of sires in fertilization. Poultry Sci. 50:
1623.
HOWARTH,B., JR. 1971. Transport of spermatozoa SMITH, R. L. (Ed.). In press. Sperm competition
in the reproductivetract of turkey hens. Poultry
and the evolutionof animalmatingsystems.New
Sci. 50: 84.
York, Academic Press.
HUANG, H. H., & T. C. CHOW. 1974. Artificial in-
semination in mule duck production. Proc. and
Abs. 15thWorld's Poultry Congr. New Orleans:
261-262.
JENNI,D. A. 1974. Evolutionof polyandryin birds.
Amer.
Zool.
14: 129-144.
KINNEY,T., •r R. E. BURGER.1960. A techniquefor
the inseminationof geese.Poultry Sci. 39: 230232.
STURKIE,P. D. 1976. Avian physiology, 3rd ed.
New York, Springer-Verlag.
WITTENBERGER,
J.F. 1979. The evolutionof mating
systemsin birds and mammals.Pp. 271-349in
Handbook of behavioral neurobiology, Vol. 3.
Social behavior and communication (P. Marler
and J. Vandenbergh,Eds.). New York, Plenum.