sign in sign up
<<
Home , Alpine accentor, American tree sparrow, Copulation (zoology), Dunnock, Longspur, Reproductive system

The Auk 110(4):875-888, 1993

ANATOMICAL TO COMPETITION IN SMITH'S LONGSPURS AND OTHER POLYGYNANDROUS

JAMESV. BRISKIE Departmentof Biology,'s University, Kingston, Ontario K7L 3N6, Canada';and Departmentof Animaland Plant Sciences,University of Sheffield, P.O. Box601, Sheffield,S10 2UQ, UnitedKingdom

AI•STRACT.--Icompared the reproductiveanatomy of the polygynandrousSmith's Longspur (Calcariuspictus) with two other polygynandrouspasserines ( [Prunellamodularis] and Alpine [P. collaris])and with a wide rangeof sociallymonogamous . All three polygynandrousspecies were found to have enlarged testesand cloacalprotuberances (i.e. the site of sperm storage)compared to specieswith other systems.Testes lengths in polygynandrousspecies averaged 44% longer and cloacalprotuberances 213% greater in volume than expectedfor the body sizesof these species.Testes mass in Smith's Longspurs comprised4.2% of adult body massor more than twice (2.0%) that found in the congeneric and monogamousLapland Longspur(C. lapponicus).Smith's Longspurs also had larger cloacal protuberances,larger seminal glomeraand higher spermstores, suggesting a greateroverall rate of spermproduction than in LaplandLongspurs. In contrast, of polygynandrous speciesdid not show any consistentdifferences in the pattern of sperm storage due to increasedsperm production by males,although this needsto be evaluatedmore thoroughly. The enlarged male reproductive organs of Smith's Longspursand other polygynandrous speciesappear to have evolvedas a consequenceof spermcompetition, whereby large sperm reservesfunction to insure paternity through diluting or displacingthe ejaculatesof rival males.Received 26 December1991, accepted 25 November1992.

SPERM COMPETITION results whenever females specific comparisonsof primates (Harcourt et mate with more than one male during the span al. 1981,Harvey and Harcourt 1984), (Car- of a single breeding attempt (Parker 1970). Al- tar 1985,Moller 1991)and (Kenagy thoughmultimale mating may be advantageous and Trombulak 1986, Brownell and Rails 1986) to females in somesituations (e.g. Smith 1988, have shown that nonmonogamousanimals have Moller 1988a),it can alsoreduce a partner'spa- relatively larger testesthan monogamousspe- ternity (e.g. Gibbset al. 1990). To counter cies.Since larger testes produce more and larger infidelity and increasethe probabilityof pater- ejaculates(Moller 1988b), an increasein testes nity, males have evolved several elaborate size has been interpreted as an to counter measures.For example, males in some (Harcourt et al. 1981, Moller speciescontinually guard and defend their mates 1988b, 1991). from the advancesof other malesduring the Smith's Longspurs (Calcariuspictus) are un- period when eggsare fertilizable (Beecherand usualamong birds in that femalesreg- Beecher1979, Birkhead 1979). If extrapaircop- ularly pair and copulatewith more than one ulationsoccur, then pair malesmay reducetheir male for a singleclutch of eggsat the sametime subsequentparental investment to avoid wast- that males pair and copulate with several fe- ing energyor resourceson raisinganother male's males (Briskie 1992). In a color-bandedpopu- offspring (Trivers 1972, Burke et al. 1989). lation near Churchill, Manitoba, 76.2% of 21 Sometimesfemale infidelity cannot be pre- females observed mated with two males, 9.5% vented becauseof ecological constraints,but with three males,and only 14.3%with a single malesmay nonethelessincrease paternity con- male (Briskie 1992, 1993). Of 27 males observed fidencethrough frequent copulationsthat di- in the same population, 63.0% mated with two lute or displacerival ejaculates(McKinney et females, 7.4% with three females, and 29.6% with al. 1984, Birkhead et al. 1987). Indeed, inter- a singlefemale. This combinationof and polygynyis termedpolygynandry and has been reported in only two other passerinesto Present address. date (Dunnock, Prunellamodularis [Davies 1985] 875 876 JAM•SV. BRISKIE [Auk,Vol. 110 and , P. collaris [Nakamura 1954,Salt 1954).Sperm producedin the testesmature 1990]). Males in all three speciesassist females in the glomera and are stored there until ejaculated. in feeding offspring; in Smith's Longspurs as Both sperm production and sperm storage capacity many as three males may provide food at a sin- are expected to increase as an adaptation to sperm competition (e.g. Smith 1984); thus, males in species gle nest (Briskie 1993). Becausefemales copulate exposedto high levels of sperm competition should with severalmales during the period when have suchorgans enlarged relative to thosefound in could be fertilized, males risk investing energy monogamous species. in offspring fathered by another male. Possibly Cloacal protuberanceswere measured on 29 pas- to minimize this cost,copulations in these spe- sefine species breeding in Ontario and Manitoba, cies are extraordinarily frequent and, in the Canada from 1988 to 1991 (see Table 1 for a list of Smith's Longspur, average more than 350 per species).Apart from the Smith's Longspur,this sam- clutch (Briskie 1992). Observations of such ex- ple included 2 socially polygynousspecies (Yellow- treme behavior, along with recent findings of headed Blackbird [Xanthocephalusxanthocephalus], high levels of mixed paternity in the Dunnock Brown-headed Cowbird [Molothrus ater]), 2 sequen- tially polyandrousspecies (American Goldfinch [Car- (Burke et al. 1989) and Smith's Longspur (75% duelistristis], Common Redpoll [C. fiammea])and 24 of 12 broods with mixed paternity; R. Mont- sociallymonogamous species (all remaining species goreerie, J. V. Briskie and T. Poldmaa unpubl. listed in Table 1; mating-systemdata from Ehrlich et data), strongly suggestthat intense sperm com- al. 1988, Briskie and Montgomerie 1992, and refer- petition is commonin all polygynandrousspe- encestherein). Using calipers,I measured(to 0.1 mm) cies. the diameter (along left-right axis) and height (on In an earlier paper, I described how sperm one side) of the cloacalprotuberance on each male competition has influenced the mating behav- captured or collected from one week before to one ior of Smith's Longspurs(Briskie 1992). Here, ! week after the peak clutch-initiation period of that examine the reproductiveanatomy of this spe- species. Clutch-initiation dates were determined by direct observationsand most males were captured or cies to determine what adaptations have arisen collected at known nest sites or on territories with to cope with the high levels of sperm compe- nesting females. Becausecloacal-protuberance size tition experienced by males. Recently, Naka- varies over the breeding cycle, for all interspecific mura (1990) and Birkhead et al. (1991) examined comparisonsI used only data from those birds mea- Alpine Accentorsand , respectively, sured during the peak period of clutch initiation; this for the purposeof determining how spermcom- is the period when cloacal protuberance size is at its petition has affected male and female repro- maximum (Quay 1986, Weatherhead and Robertson ductive anatomyin thesespecies. Males in both 1980, unpubl. data). Using these values, I estimated specieswere found to have very large testesand cloacal-protuberancevolume as h•rr•, where h is the cloacalprotuberances (i.e. the site of spermstor- cloacal-protuberanceheight and r is 0.5 timesthe pro- tuberance diameter. Sample sizes per speciesranged age) but in neither study was enough infor- from 1 to 37 (Table 1). mation available to make detailed comparisons To examine the structure of the seminal glomera with socially monogamousspecies. I collected and in greater detail, I collected five male data on cloacal-protuberancesize, testissize, and Smith's Longspurs:three during their copulationpe- sperm-storagepatterns for a wide variety of riod, one after the period, and one post- North American passerines.Using the data from breeding . Within an hour of death, the cloacal Alpine and Dunnocks, ! then exam- protuberancewas removed and preservedin 10%buf- ine how sperm competition has influenced the fered formalin. Each seminal glomerus was later dis- reproductive anatomiesin all three polygynan- sected,measured and weighed. To estimatethe num- drous species. ber of sperm stored by each male, I finely minced each glomerus and washed the sperm into a known METHODS volume of phosphate buffered saline. Five samples were then withdrawn from this solution and the The of male birds comprises number of sperm counted under a microscopeusing the two testes and their connections to the cloaca via an Improved Neubauer counting chamber.Estimates the ductus deferens (see review in Lake 1981). The for both glomerawere summedto obtain an estimate distal end of each ductus deferens is known as the of total number of sperm stored.For comparativepur- seminal glomerusand, during the breeding season, poses,I also estimated the number of sperm in the it becomesexpanded and highly coiled,pushing out seminal glomera of one (C. lap- of the , and forming an externally visible ponicus),one American Tree Sparrow (Spizellaarborea), cloacal protuberancearound the vent (Wolfson 1952, and three Yellow-headed Blackbirds. All birds were October1993] ReproductiveAnatomy of Smith's Longspurs 877

TABLEI. Summaryof data on cloacal-protuberancedimensions (_+SE), testis lengths and samplesizes for speciesused in study.All datafrom my study except for theAlpine Accentor (Nakamura 1990) and Dunnock (Birkhead et al. 1991,A. P. Moller pers. comm.). Cloacalprotuberance Testis Diameter Height length Species (mm) (mm) n (mm) 1 Least Flycatcher(Empidonax minimus) 4.25 -+ 0.19 3.34 -+ 0.17 8 5.70 2 Red-eyedVireo (Vireoolivaceus) 4.93 -+ 0.33 4.15 _+ 0.29 4 8.50 3 Warbling Vireo (V. gilvus) 5.25 + 0.25 3.25 + 0.75 2 6.79 4 American Robin (Turdusmigratorius) 12.21 _+ 0.41 10.51 -+ 0.40 7 14.50 5 Veery (Catharusfuscescens) 7.00 + 1.00 5.75 -+ 0.75 2 9.50 6 Gray Catbird (Dumetellacarolinensis) 7.47 _+ 0.28 4.09 -+ 0.26 11 10.80 7 Black-cappedChickadee (Parus atricapillus) 4.40 4.30 1 6.88 8 Barn Swallow (Hirundo rustica) 5.57 + 0.30 3.79 + 0.24 7 6.90 9 Tree Swallow (Tachycinetabicolor) 7.50 -+ 0.46 6.28 -+ 0.42 4 9.30 10 Horned Lark (Eremophilaalpestris) 6.80 6.00 1 8.80 11 American Pipit (Anthusrubescens) 6.50 6.00 I 9.00 12 Dunnock (Prunella modularis) 8.50 _+ 0.39 6.50 -+ 0.30 6 10.90 13 Alpine Accentor(P. collaris) 14.18 -+ 0.31 12.78 -+ 0.49 13 19.90 14 American Goldfinch (Carduelistristis) 6.00 -+ 0.50 5.35 + 0.35 2 7.00 15 Common Redpoll (C. fiammea) 7.00 _+ 0.58 5.40 -+ 0.31 3 6.80 16 Savannah Sparrow (Passerculussandwichensis) 7.65 -+ 0.44 6.58 -+ 0.35 6 9.61 17 Song Sparrow (Melospizamelodia) 7.00 6.30 1 8.62 18 AmericanTree Sparrow( arborea) 8.80 -+ 0.74 7.94 -+ 0.60 5 11.06 19 (S. passerina) 6.48 _+ 0.43 6.00 -+ 0.32 4 8.33 20 Clay-coloredSparrow (S. pallida) 6.17 -+ 0.22 5.78 -+ 0.11 6 7.23 21 White-crownedSparrow (Zonotrichia leucophrys) 8.41 _+ 0.23 7.14 -+ 0.20 10 10.35 22 Smith's Longspur(Calcarius pictus) 11.49 -+ 0.18 9.89 -+ 0.22 15 14.50 23 Lapland Longspur(C. lapponicus) 6.70 5.70 1 10.10 24 TennesseeWarbler (Vermivoraperegrina) 5.40 5.20 1 5.92 25 Blackpoll Warbler (Dendroicastriata) 6.85 + 0.05 6.60 + 0.40 2 7.03 26 Yellow Warbler (D. petechia) 5.72 -+ 0.14 4.64 -+ 0.16 37 5.30 27 Common Yellowthroat (Geothlypistrichas) 4.00 -+ 0.35 2.90 -+ 0.19 5 7.58 28 American Redstart(Setophaga ruticilla) 5.20 _+ 0.31 5.77 -+ 0.27 3 6.70 29 Yellow-headed Blackbird (Xanthocephalusxanthocephalus) 11.27 -+ 0.43 8.57 -+ 0.20 3 13.30 30 Northern Oriole (Icterusgalbula) 8.62 -+ 0.66 6.72 _+ 0.52 6 12.22 31 Brown-headed Cowbird (Molothrus ater) 6.08 _+ 0.37 3.94 -+ 0.29 8 7.00

collected(under permit) during the copulationperiod 1986)is enlargedand elaboratedin speciessubject to and had well-developed cloacalprotuberances and sperm competition.Thus, I looked for any similar largetestes. These values were then comparedto sem- modificationsthat may have occurredin the papilla inal glomera sizesand/or sperm number estimates of Smith's Longspursrelative to other passerines. obtainedfrom: Birkheadet al. (1991)for the Dunnock, Data on testis size were obtained from museum ZebraFinch (Taeniopygiaguttata) and BengaleseFinch skinshoused in the Royal Ontario Museum (Toronto), (Lonchurastriata); Wolfson (1954) for the HouseSpar- the National Museum of Canada (Ottawa), the - row (Passerdomesticus ), American Robin (Turdus migra- itoba Museum of Man and Nature (Winnipeg), and torius)and Dark-eyedJunco (Junco hymelis); and T. R. from personal observationsof fresh material. To es- Birkhead (pers. comm.) for the Canary (Serinuscan- timate testislength at peak breeding for eachspecies, aria). I fitted a second-orderpolynomial regressionbe- In passerines,the extreme distal end of each ductus tween testis length and date, and used the maxima deferensforms a conical-shapedpapilla that pro- of these regressionsas the average maximum testis trudes in the lumen of the cloaca. These structures length during breeding.At least25 measuresof testis are thought to function, in conjunctionwith the proc- length were available for each speciesincluded in todealchamber of thecloaca, as a phallusduring sperm this study. Testis size in the Alpine Accentor was transfer (Salt 1954). After the seminal glomera were obtained from Nakamura (1990). Birkhead et al. (1991) removed, I cut open the cloacaof the four speciesI did not provide measurementsof testislength in the collectedand measuredthe width and height of the Dunnock, so I obtained these data from Moller (1991, papilla of the ductusdeferens. The phallusin pri- pers. comm.). I also weighed the testesfrom those mates (Dixson 1987) and whales (Brownell and Rails birdscollected for estimatingsperm number and sem- 878 l•a•s v. BinsKin [Auk, Vol. 110

Fig. 1. Lateralview of cloacalprotuberance of a male Smith'sLongspur. Excepl for ring of surroundingvent (arrow),protuberance is mostlybare. Abdominal feathers in this individualhave been removedto exposebase of protuberance;anterior of bird is to left in this view. Scalebar = 10 ram. inal glomerasize. Bodymasses for North American phylogenetic effects.Such methods are required be- specieswere obtainedfrom Dunning (1984). causesome traits may be similar in closely related Femalesof polygynandrousspecies also might ex- speciessimply due to common ancestryrather than hibit certain anatomicalfeatures as a consequenceof independent . Failure to control for this ar- increasedsperm competition in males.For example, tificially weights sample sizes and the influence of when males produce more sperm, females may in- some specieson the analysis.First, I used Sibley and creasethe capacilyfor sperm storage.Sperm storage Ahlquist (1990) to constructa phylogeny for the spe- in female birds occurs in specialized sperm-storage cies studied. Then, using the independent contrast tubules (SSTs)located at the junction of the method (Harvey and Pagel 1991),I calculatedunique and (e.g. Howarth 1974, Shugart 1988). To linear contrasts within each taxon with two or more determine if the patternsof spermstorage in female data points.This comparativemethod calculatescom- Smith'sLongspurs differ from thoseof otherbirds, I parisonsor "contrasts"among daughter taxafor each comparedthe numberand size of SSTsacross 20 spe- node in a phylogeny for which there is variation in ciesof North Americanpasserines. The methodsused the test variable. In essence,a phylogeny is first for estimatingSST numberand size have been de- searchedfor independent evolutionary eventsof the scribedin detail elsewhere (Birkhead and Hunter 1990, character in question (e.g. large testes size). These Briskie 1990,Briskie and Montgomerie 1992,Briskie events are then comparedwith independent evolu- and Birkhead 1993). Some of these data were pre- tionary eventsin a secondor third character(e.g. body sentedpreviously with respectto the evolution of mass,).To test for overall comparative spermsize (Briskieand Montgomerie 1992),but are relations across taxa, I determined whether linear repeatedhere for comparativepurposes along with contrasts for one variable were correlated with those dataon sperm-storagepatterns in the Dunnockfrom of another. I used Pearson'sproduct-moment coeffi- the literature (Birkhead et aL 1991). cient to test for linear relations between two variables In all comparativeanalyses, I used the methods and forced the regressionthrough the origin as rec- describedby Harvey and Pagel(1991) to controlfor ommendedby Harvey and Pagel (1991).A significant October1993] ReproductiveAnatomy of Smith's Longspurs 879 correlation between two variables suggeststhat the evolution of one character is associated with that of • 1.41 Y= 1.2-0.015X the second.Sample sizesin thesesorts of comparative '• 1'2'{'• • S r2=0'53 analyses,thus, are the number of independent con- trastsand not the number of speciesoriginally used - ß in the analysis.For clarity, I presentcorrelations using both raw data and the contrasts. All variables were log transformedbefore analysesto help normalize distributions. Standard errors of the mean (SE) were calculated for all results. 0.4- .; RESULTS 0.2- ß Cloacal-protuberancesize.--During the breed- U o.o ' io' ing season the cloacal protuberance of male June July Smith's Longspursappears as a large, barrel- shapedstructure surrounding the vent (Fig. 1). Calendar Date It averages 11.49 + 0.18 mm (n = 15) in diameter Fig. 2. Declinein cloacal-protuberancesize of male and 9.89 + 0.22 mm in height. The protuber- Smith's Longspursover the breeding season.Males ance is largely bare except for a ring of 24 to arrive on breeding groundsin early June.Copulation 27 (t = 25.7 _+0.88, n = 3) feathers encircling activitybegins about I0 dayslater and ceasesby third the vent. Cloacal feathers are white with a black week in June once all clutches have been laid. Open baseand stand approximately 8.0 mm high pos- circle representscloacal-protuberance size of a single teriorly, but taper to 2.6 mm at the anterior end male known to have renested.Sample size is 30 males. of the protuberance.Cloacal feathers in Smith's Decline in protuberance size in one male measured twice is shown by dotted line. Longspur males did not differ appreciably in number or size from the cloacal feathers ex- amined in one male of the congenericLapland body size; Smith's Longspur cloacalprotuber- Longspur (n = 30 feathers,8.0 mm posteriorly, ancesalone averagedthree times the volume of 2.5 mm anteriorly). Unlike in Alpine Accentors protuberances in the monogamous Lapland (Nakamura 1990), the ducts of the seminal Longspur.Cloacal-protuberance volumes in the glomera of Smith's Longspurswere not visible two polygynousspecies (Yellow-headed Black- through the skin of the protuberance. bird, Brown-headed Cowbird) were 44% less on Cloacal-protuberancesize in Smith's Long- averagethan expectedfor their body size while spurswas already large when birds arrived on the two sequentially polyandrousspecies (Com- the breeding grounds in early June (Fig. 2). Af- mon Redpoll, American Goldfinch) had pro- ter copulation behavior ceasedin mid-June, clo- tuberancesabout 36%larger than expected(Fig. acal-protuberancevolume decreasedgradually 3A). Thus, large cloacal-protuberancesize was such that protuberance volume was less than associatedwith (and perhaps with one-half that of copulating birds by the time sequential polyandry), but not with any other maleswere feedingoffspring in early July (Fig. mating system. Protuberancevolume contrasts 2). Renestingin Smith's Longspursoccurs only between polygynandrous species and their rarely (Briskie 1993),but in one renestingmale, closestmonogamous relatives or groupsof rel- cloacal-protuberance size during his second atives were consistently large and positive but copulation period was similar to that found in low for body size (Fig. 3B),confirming that large copulating males during first nesting attempts cloacal-protuberancesize is associatedwith po- (Fig. 2). lygynandry once phylogenetic effects are re- Protuberance size varied 10-fold across the moved. passerinespecies that I sampled(Fig. 3A); it was Seminalglomera size and number of sperm.--The greatestin the Alpine Accentor and smallestin left and right seminal glomera from three male the Common Yellowthroat (Geothlypistrichas). Smith's Longspurs collected during the copu- Cloacal-protuberance size increased with in- lation period averaged0.158 _+0.017 g and 0.164 creasingbody mass(Fig. 3A). The three polygy- + 0.012g, respectively(Fig. 4). Combinedglom- nandrous species had protuberances approxi- era mass(0.322 g) of copulating-period males mately 213% larger than that expectedfor their was 1.13% of body mass(28.5 g). Glomera mass 880 JAMIiSV. Bmsrde [Auk,Vol. 110

times that of either the Lapland Longspur (0.115 A o13 g, n = 1) or the American Tree Sparrow (0.102 4 g, n = 1). The averageseminal glomera massof 1000: 022//•9 Yellow-headed Blackbirds(0.294 + 0.014 g, n = 18eO12 21 • 3) was similar to that found in Smith's Long- spurseven though blackbirdsare almost three times the massof longspurs.When phylogeny • 100 is controlled, cloacal-protuberancevolume was • 7ee• ] strongly correlatedwith seminal glomera mass acrossspecies (Y = 0.93 X; r 2 = 0.82, P = 0.0003, 27ø n = 10 species);thus, specieswith large cloacal 50 100 protuberanceshave proportionately large sem- •e b•y m.s •) inal glomera,as would be expectedif the cloacal protuberance forms as an expansion of the o.s glomera. The combined number of sperm in both 0.6 glomera of Smith's Longspursduring the cop- 0.4 ulation period was estimated to be 2.17 + 0.22 0.2 x 108 (range 1.76 x 108 to 2.50 x 108;n = 3). This was approximately four times that found o in the glomera of the Lapland Longspur (0.53 -0.2 x 108 sperm) or the American Tree Sparrow (0.58 x 108sperm), but only one-half that found -0.4 in the Yellow-headed Blackbird (4.11 + 0.24 x -o.1 1 02 0.3 04 0.5 0.6 07 0.8 108 sperm) and one-fifth that found in Dun- Con•ts • •e •s nocks (10.6 x 108sperm; Birkhead et al. 1991). Fig. 3. (A) Relation between male body massand However, when controlled for phylogeny, the cloacal-protuberancesize for 31 speciesof passerines. number of spermstored in the seminalglomera Polygynandrousspecies represented by open circles, was strongly correlated with cloacalprotuber- sequentiallypolyandrous species by closedtriangles, ance height acrossspecies (Y = 4.82 X; r 2 = 0.61, polygynous speciesby closedsquares, and monoga- P = 0.039, n = 7 species),indicating that species mousspecies by closedcircles. Numbers refer to spe- with large cloacalprotuberances generally have cies listed in Table 1. (B) Independent contrasts(see large reservesof sperm. Methods) between massand protuberancesize for Cloacalanatorny.--In all four speciesexam- samegroup of species.Open circlesare contrastsbe- ined, the papillae of the ductusdeferens formed tween (1) Smith's and Lapland longspurs,and (2) be- a pair of small conicalstructures projecting into tween the genera Prunella(Dunnock and Alpine Ac- centor) and Anthus. Large positive contrasts in the lumen of the urodeum of the cloaca(Fig. protuberancesize in these two comparisonssuggest 5). In Smith's Longspurscollected during the that an increase in protuberancesize is associated copulation period, each papilla averaged 1.13 with polygynandry. + 0.01 mm at the base and 1.53 + 0.04 mm in height (n = 3). Papillaein the Lapland Longspur (base 1.18 mm, height 1.58 mm, n = 1), Amer- from one postcopulation-periodmale was less ican Tree Sparrow (base 1.08 mm, height 1.18 than one-half (0.115 g combined) that of cop- mm, n = 1), and Yellow-headed Blackbird (base ulating males,while the seminalglomera from 1.41 + 0.17 mm, height 1.65 + 0.15 mm, n = 3) a singlepostbreeding male wasonly 0.030g or were similar in both size and appearance.There lessthan 0.10%of male body mass.The dimen- was no obvious morphological difference be- sionsof left and right glomerain copulating tween the papillae of longspursand the other males were very similar and averaged9.59 + speciesexamined that would be related to the 0.53 mm along the dorsal-ventralaxis, 6.93 + frequent copulations characteristicof Smith's 0.18 mm along the anterior-posterioraxis, and Longspurs.No information is available on the 4.34 + 0.14 mm along the left-right axis. size of the cloacal papillae in either the Dun- The averagecombined seminal glomera mass nock or Alpine Accentor. of Smith'sLongspurs (0.322 g) wasnearly three Testissize.--Testes length in breeding Smith's October1993] ReproductiveAnatomy of Smith's Longspurs 881

c

Fig. 4. Ventral view of seminal glomera (S) of a male Smith's Longspur.The tortuous coiling of ductus deferens is readily apparent. Cloaca (C) appearsbetween the two glomera. Baseof vent feathersare to far left in this view. Scale bar = 3 min.

Longspurspeaked at 14.5 mm and on average length contrastsof the polygynandrousspecies had a combined massof 1.203 g, or about 4.2% were large and positive,but small for body mass, of adult bodymass (Fig. 6). This is slightly great- suggesting that a switch to polygynandry at er than that of the Dunnock (3.36%; Birkhead these nodes resulted in increases in testis size et al. 1991), but less than that of the Alpine but not body mass(Fig. 7B) Accentor(7.7%; Nakamura 1990).Testis length Spermstorage in females.--The mean number was significantly correlated with body mass of sperm-storage tubules in female $mith's acrossspecies (Fig. 7). The three polygynan- Longspurscollected during egglaying was 1,537 drous species(Smith's Longspurs,Alpine Ac- _+ 146 (n = 3 females), somewhat less than that centors and Dunnocks) had testis lengths an found in Lapland Longspurs(2,147 _+204, n = average of 44% larger than expected for birds 6 females), but similar to the Dunnock (1,398; with similar body masses.Testis length in Birkhead et al. 1991). No information is avail- Smith's Longspurswas 4.4 mm greater than able on sperm storage in the Alpine Accentor, Lapland Longspur testislength (10.1 mm) and but the number of SSTs in Smith's Longspurs more than twice the mass(Lapland Longspur and Dunnocks is closeto that expectedfor birds 0.544g or 2.0%of malebody mass). Testis lengths of their body mass(Fig. 8A). Thus, in thesetwo in the two polygynous species(Yellow-headed speciesfemales do not appear to provide more Blackbird, Brown-headed Cowbird) were 28% sperm-storagesites in response to increased less than expectedfor their body sizes,while sperm production by males. the two sequentiallypolyandrous species (Com- The mean length of SSTsin Smith's Long- mon Redpoll, American Goldfinch) had testes spurswas 237.6 + 44.9 #m (n = 3 females),only lengths about 7% lessthan expected(Fig. 7A). slightly longer than thoseof the Lapland Long- Thus, large testessize was associatedwith po- spur (216.2 _+10.1 #m, n = 6), but lessthan that lygynandry, but not with any other mating sys- found in the Dunnock (370.3 _+ 14.4 #m; Birk- tem. When phylogeny was controlled, testis- head et al. 1991). Since SST length appears to 882 JAMESV. BmsrdE [Auk,Vol. 110

Fig. 5. View of papilla of ductusdeferens in urodeum of a male Smith'sLongspur. Papillae appearas two conicalstructures on ventral wall of cloaca(positions indicated by two pins). Vent is at top of photograph. Scale bar = 2 mm. have coevolved with sperm length (Briskie and how many spermare in the ejaculatesof Smith's Montgomerie 1992,Birkhead and Moller 1992), Longspurs,the number of spermstored by fe- the appropriate measurefor comparing sperm- males is less than 0.02% of the sperm present storage capacity of SSTs among speciesis the in the seminal glomera of male Smith's Long- number of layers of sperm that can be accom- spurs.This percentagewas similar for Lapland modated within each SST. When sperm length Longspur(0.04%) and the Dunnock (<0.01%, is compared to SST length (Fig. 8B), it becomes calculated from the figures in Birkhead et al. apparent that Smith's Longspurs can accom- 1991), but information on the amount of sperm modate about two layers of sperm per SST, a storedby femalesin other speciesis not avail- pattern common to most of the speciesof pas- able to make a detailed comparison. setinesexamined to date, including the Lapland Longspur. The Dunnock, however, can accom- DISCUSSION modate up to five layers of sperm, a pattern approached only by the American Robin (Fig. In common with the other two polygynan- 8B). Thus, Dunnocks appear to have an in- drousspecies studied to date, Smith's Longspur creasedcapacity to storesperm within eachSST males have very large testes and doacal pro- relative to other species,but such an increase tuberancesfor their body size. Comparedwith does not appear to be related to polygynandry a wide variety of mostlymonogamous birds, all since it was not found in Smith's Longspurs. three polygynandrousspecies examined here The mean number of sperm within each SST showed the same pattern. For Smith's Long- was estimatedin one Smith's Longspurfemale spurs,testes and cloacal-protubemncesize were to be 25.8 (range 0 to 200, n = I00 SSTscen- particularlylarge and striking comparedto those sused). Since there were approximately 1,500 traits in the socially monogamousand conge- SSTsper female, about 39,000 sperm are stored neric Lapland Longspur. at any given time. Although it is not known Large testesare thought to be an adaptafion October1993] ReproductiveAnatomy of Smith'sLongspurs 883

I I

Fig. 6. (A) View of testesin a breedingmale Smith's Longspur. Scale bar = 20 mm. (B) Close-upview of testesin a secondindividual. Each testis extends from caudaledge of the cage(R) to nearanterior edge of cloacalprotuberance (C). Scalebar = 10 mm. 884 JAMESV. B•JSKIE [Auk, Vol. 110

20- A ol3 5000

28•^,_.•15 14 • _ 1000

26 500

10 50 100 10 100 Male bodymass (g) Femalebody mass (g)

0.3 B • 500' • 400-600'12o 04 3515•3•

ß ß [' 200' •,•23 33-

-0.1 100'

-0.1 0'.1 d.2 d.3 d.4 0'.5 0'.6 0%70'.8 50 100 150 200 250 300 Contrasts in male mass Spermlength (gm) Fig. 7. (A) Relationbetween testes length and male body massfor 31 speciesof passerines.Polygynan- Fig. 8. (A) Relation between female body mass drous speciesrepresented by open circles, sequen- and number of sperm-storagetubules (SST) per fe- tially polyandrousspecies by closedtriangles, polyg- male. Smith's Longspurs(22) and Dunnocks (12) are ynous speciesby closed squares,and monogamous indicatedby open circles.Numbers refer to species speciesby closed circles. Numbers refer to species listedin Table 1 exceptfor the following: (32) Carduelis listed in Table 1. (B) Independent contrastsbetween hornemanni,(33) Plectrophenaxnivalis, (34) Euphagus testislength and body massfor samegroup of species. cyanocephalus,(35) Agelaiusphoeniceus, and (36) Stur- Open circles are contrastsbetween (1) Smith's and nella neglecta.Data from Briskie and Montgomerie Lapland longspurs,and (2) between the genera Pru- (1992) and Birkhead et al. (1991). Contrast analyses nella(Dunnock and Alpine Accentor)and Anthus. Large (seeMethods) for mostof thesedata were presented positive contrastsin testessize in these two compar- in Briskieand Montgomerie(1992) and show a similar isons suggest that an increase in testes size is asso- relationship between female body massand the num- ciated with polygynandry. ber of SSTswhen controlledfor phylogeny. (B) Re- lation between sperm length and length of SSTsin females. Line indicatesSST lengths that can accom- to sperm competition becauselarge testespro- modate two layers of sperm. Numbers and symbols duce larger volumesof more concentratedejac- as above.Although Smith's Longspurappear to store ulates (Metter 1988b). When females coputate spermin a pattern commonto most other passetines, with more than one male, such large numbers Dunnock SSTs are long relative to length of their of spermmay function to dilute or displaceejac- sperm. Contrast analysesusing most of these data ulates from rival males. Thus, males that pro- were presented in Briskie and Montgomerie (1992) duce and inseminate the largest number of and showthe samerelationship between sperm length sperminto a given female are likely to father and SSTlength when phylogenyis controlled. a higher proportion of her offspringand will be favored over males with lower sperm pro- , when sperm from two males are in- duction. This strategyis likely to be most suc- seminated within 4 h of each other, paternity cessful when fertilization successis directly is directly proportional to the number of sperm proportionalto number of spermper ejaculate inseminatedby each male (Martin et at. 1974). and femalesregularly switch from one male to However, after 4 h the last male to coputate another over a short time interval. In domestic generally experiencesgreater success,a pattern October1993] ReproductiveAnatomy ofSmith's Longspurs 885 called last-male sperm precedence (Sims et al. required becausemales must fertilize two or 1987, Birkhead et al. 1988). Last-male prece- more females, while males in socially monog- dence is thought to arise from either the cov- amousspecies need only produceenough sperm ering up of first-male sperm by second-male to fertilize a single female (Cartar 1985, Mailer sperm within the SSTsor by the displacement 1991). However, this sperm-depletion hypoth- of first-male sperm from the SSTs by second- esis seemsunlikely for the speciesconsidered male sperm (Lessells and Birkhead 1990). here. Both the Yellow-headed Blackbird and The pattern of spermprecedence is not known Brown-headedCowbird are consideredpolyg- for any of the polygynandrousspecies studied ynous, yet in neither specieswas testeslength here, but becauseeach copulates very frequent- or cloacal-protuberancevolume larger than ex- ly and at intervals usually lessthan 4 h (Davies pected for their body mass.In fact, the Brown- 1985, Nakamura 1990, Briskie 1992), it is pos- headed Cowbird had the smallest cloacal-pro- sible that paternity is determined by propor- tuberance volume and testeslength relative to tional representation. This seems especially its body size of any bird in my sample. Thus, likely in Dunnocks and Alpine Accentors,in sperm production by males with relatively which females frequently switch or alternate "normal-sized" testesapparently is sufficient to from one mate to another over the copulation fertilize several females.This suggeststhat in- period (Davies 1985, Nakamura 1990). How- creased testes size with polygynandry more ever, repeatedmate switching is infrequent in likely evolved to insure paternity under con- Smith's Longspurs(Briskie 1992, unpubl. data). ditions of intensesperm competition rather than Instead, each female longspur copulatesexclu- through sperm depletion. sively with one male for a period lasting one Neither female Smith'sLongspurs nor female to six days, at which point she switches to a Dunnocks appeared to have any obvious ad- secondmale and copulatesexclusively with him aptations to increased sperm competition in for a period lasting one or more days (Briskie males. Both the number and size of the SSTs 1992). Thus, for male Smith's Longspurs,large were similar to that expected for their body testes and the concomitant increase in sperm mass and found in closely related species. production may be an adaptationto rapidly di- Nonetheless,it is difficult to predict exactlyhow lute and overwhelm sperm at mate switches femalegenital should respond to sperm rather than simply increasingthe proportion of competition. One possibility is that females sperm inseminated. could provide more or larger areas for sperm The large cloacal protuberances of Smith's storage;however, there seemslittle theoretical Longspurs and other polygynandrous birds reason why this should be favored over any seem to function as large sperm reservesfor other pattern. Females might even be able to frequent copulations.Cloacal-protuberance size induce greater sperm competition by reducing was strongly correlated with both seminal the number of sperm-storagesites and, thereby, glomera mass and the number of sperm esti- increasethe scrambleamong ejaculatesfor the mated within the glomera. Thus, large protu- few remaining sites. Briskie and Montgomerie berancesare large simply becausethey contain (1992) have argued that such a mechanismmay huge numbersof sperm.All three polygynan- accountfor the wide variation in spermatozoa drous specieswere found to have large cloacal size found acrossspecies. The number of SSTs protuberancesfor their body sizes, a pattern was slightly less in Smith's Longspursthan in which further suggeststhat largeprotuberances the Lapland Longspur,but the numberof SSTs are an adaptation to increasedsperm competi- did not differ appreciably between Dunnocks tion. Acrossa wide sample of species,Birkhead and their nearest monogamous relatives. In et al. (1993) found that protuberance size was contrast, Dunnocks had relatively longer SSTs correlatedpositively with the frequencyof cop- than most other passerines studied; however, ulation, a pattern that also supports the hy- sperm storage needs to be examined in more pothesis that variation in cloacal-protuberance polygynandrous birds before it can be deter- size is related to the degree of sperm competi- mined whether long SSTsare an adaptation to tion. high sperm production by males. An alternative explanationfor both large tes- Dunnocksappear unusual among birds in that tes and large cloacalprotuberances in polygy- femalesregularly ejecta small volume of ejac- nandrousspecies is that large sperm storesare ulate in responseto cloacal pecking by males 886 JAMESV. BRISKIE [Auk,Vol. 110

(Davies 1983). Cloacal pecking is an obvious in longspursis greatestduring the morning and advantage to a male if it removes sperm from virtually ceasesby the afternoon (Briskie 1992). previous males, but may be advantageousto Thus,it is possiblethat spermreserves may have females as well if it assuresmales of paternity been depleted in the birds I collectedbut this and increasesthe likelihood that they will in- needs to be tested further. vest in subsequentoffspring. Cloacalpecking Although my observationsprovide clear ev- does not occur in Smith's Longspurs,but it is idence that testis size and cloacal-protuberance clear that neither is all sperm later stored by size are larger in speciesin which sperm com- females.Less than 1% of sperm present in the petition is thought to be most intense, it is less seminal glomera of male Smith's Longspursis certain how such adaptationsfunction in as- storedin the SSTsat any given time. This value suring male paternity. For example, it is not is similar to that of Dunnocks and Lapland known how the large reservesof sperm in po- Longspurs. Since most sperm is not stored nor lygynandrousspecies are allocatedto ejaculates ejected, it is likely that most is lost through or whether there is a trade-off between ejacu- defecation, although I have not examined fe- lation frequencyand the number of spermper male fecal sacsto determine if spermare pres- ejaculate. Males in polygynandrous species ent. Alpine Accentorsalso do not perform clo- might be expectedto have larger ejaculatesthan acalpecking but, prior to mounting by the male, monogamousspecies in order to dilute rival the female defecates while the male watches male'ssperm, but they alsomight be forcedto (Nakamura 1990). Although it was not deter- "ration" sperm over the copulation period to mined if spermare voided during this display, insure that some sperm is available for future it may function in an analogousway to sperm use.Which pattern bestinsures paternity in each ejection in Dunnocks. type of mating systemmay well depend upon One possibleexplanation for the large dif- the frequency of encountering rival ejaculates ference in number of sperm stored between and the physiologicalconstraints of sperm pro- malesand femalesis that femalesmay storeonly duction. The advantagesand disadvantagesof a small quantity of sperm simply becausethat either strategy will further depend upon how is all they need to maximize their . If females in each speciesstore and use sperm storing large quantities of sperm is costly to from different ejaculates.Sorting out the inter- females (either energetically or through in- actions between ejaculate size, creasedprobability of contacting parasitesor frequencyand spermstorage patterns may prove diseases),then females should store only that complex,but the resolution of this problem will amount of sperm that insuresa maximum rate provide a basisfor understandingthe evolution of fertility. The rest is superfluousand voided. and adaptive modification of the avian repro- For males, this loss is costly but it should not ductive system. change the benefits of producing large ejacu- lates if doing so increasesthe probability of increasingthe number of sperm(however, small ACKNOWLEDGMENTS in absolutenumbers) that eventually find their The Churchill Northern Studies Centre, the Uni- way into the SSTs. versity of Manitoba Field Station (Delta Marsh) and The number of sperm stored in the seminal the Queen'sUniversity BiologyStation provided fa- glomera of Smith's Longspur males was ap- cilities during the courseof this study. I thank Ron proximately four times that found in Lapland Bazin, Paul Golding, Sarah Kalff, Jan Lifjeld, Chris Longspursand American Tree Sparrows,but far Naugler,and Gilles Seutin for theirhelp in measuring less than that found in Yellow-headed Black- cloacalprotuberances. This project was funded by In- birds and Dunnocks. This result was somewhat dian and Northern Affairs Canada (Northern Scien- unexpectedbecause Smith's Longspurshad the tific TrainingGrants Program), a Natural Sciencesand largest combinedseminal glomera massof any Engineering ResearchCouncil of Canada (NSERC) PostgraduateScholarship and an NSERCoperating of the species for which data were available. grant to Robert Montgomerie. The manuscriptwas One possibility is that the Smith's Longspur written while I held an NSERC Postdoctoral Fellow- males I examined were collected in the after- shipat the Universityof Sheffield.Tim Birkhead,Rob- noon (n = 1) and evening (n = 2), and they may ert Montgomerie,Geoff Hill, Gary Schnell,and two have already usedup a large proportion of their anonymous reviewers provided many useful com- spermreserves by this time. Copulationactivity mentson an earlierdraft of this manuscript,and Mike October1993] ReproductiveAnatomy of Smith's Longspurs 887

Siva-Jothy let me use his microscopefor taking pho- sperm competitionin Dunnocks.Nature 302:334- tographs of the seminal glomera. 336. D^wES,N.B. 1985. Cooperation and conflict among LrrER•tUR• CITED Dunnocks,Prunella modularis, in a variablemating system. Anim. Behav. 33:628-648. BEECHER,g. D., AND I. g. BEECHER. 1979. Sociobiol- DIXSON, A. F. 1987. Observations on the evolution ogy of Bank Swallows:Reproductive strategy of of genitaliaand copulatorybehaviour in male pri- the male. Science 205:1282-1285. mates. J. Zool. (Lond.) 213:423-443. BIRKHEAD,t.R. 1979. Mate guarding in the Magpie DUNNING,J. B., JR. 1984. Bodyweights of 686 species Picapica. Anim. Behav. 27:866-874. of North American birds. West. Bird Banding As- BIRKHEAD, T. R., L. ATKIN, AND A. P. MOLLER. 1987. soc. Monogr. 1:1-38. Copulation behaviour of birds. Behaviour 101: EHRLICH, P. R., D. S. DOBKIN, AND D. WHEYE. 1988. 101-138. The birder's handbook:A field guide to the nat- BIRKHEAD,T. R., J. V. BRISKre,AND A. P. MOLCER. 1993. ural history of North American birds. Simon and Male sperm reservesand copulation frequency Schuster, New York. in birds. Behav. Ecol. Sociobiol. 32:85-93. GIBBS,H. L., P. J. WEATHERHEAD,P. T. BO^G, B. N. BIRKHEAD,T. R., B. J. HATCHWELL,AND N. B. DAVIES. WHrr•, L. M. T^BAK, AND D. J. HO¾SAK. 1990. 1991. Sperm competitionand the reproductive Realizedreproductive success of polygynousRed- organsof the male and female Dunnock Prunella winged Blackbirdsrevealed by hypervariable modularis. Ibis 133:306-311. DNA markers. Science 250:1394-1397. BIRKHEAD,T. R., AND F. M. HUNTER. 1990. Numbers HARCOURT, A. H., P. H. HARVEY, S. G. LARSON, AND of sperm-storagetubules in the Zebra Finch(Poe- R. V. SHORT. 1981. Testis weight, body weight philaguttata) and BengaleseFinch (Lonchurastria- and breeding systemin primates.Nature 293:55- ta). Auk 107:193-197. 57. BIRKHEAD,T. R., AND A. P. MOLLER. 1992. Sperm HARVEY,P. H., AND A. H. HARCOURT.1984. Sperm competition in birds: Evolutionary causesand competition, testessize, and breeding systemsin consequences.Academic Press,London. primates. Pages 589-600 in Sperm competition BIRKHEAD,T. R., J. PELLATT,AND F. M. HUNTER. 1988. and the evolution of mating systems(R. Extra-pair copulation and sperm competition in L. Smith, Ed.). Academic Press, Orlando, Florida. the Zebra Finch. Nature 334:60-62. HARVEY,P. H., ANDM. PAGEL.1991. The comparative BRISKIE,J.V. 1990. Behavioural and anatomical cor- method in evolutionary biology. Oxford Univ. relatesof spermcompetition in birds. Ph.D. the- Press, Oxford. sis, Queen's Univ., Kingston, Canada. HOWARTH,B., JR. 1974. Sperm storage:As a function BRISKIE,J. V. 1992. Copulation patterns and sperm of the female reproductive tract. Pages237-270 competition in the polygynandrous Smith's in The and its functions (A.D. Johnson Longspur.Auk 109:563-575. and C. W. Foley, Eds.).Academic Press,New York. BRISKIE,J. V. 1993. Smith's Longspur. In The birds KENAGY,G. J., AND S.C. TROMBULAK. 1986. Size and of North America, No. 34 (A. Poole, P. Stetten- function of mammaliantestes in relation to body heim, and F. Gill, Eds.). Academy of Natural Sci- size. J. . 67:1-22. ences, Philadelphia, and American Ornitholo- LAKE,P.E. 1981. Male genital organs.Pages 1-61 in gists' Union, Washington, D.C. Form and function in birds, vol. 2 (A. S. King BRISKIE,J. V., AND T. R. BIRKHEAD. 1993. A review and J. McLelland, Eds.). Academic Press,London. of the methods used to study the anatomy of LESSELLS,C. M., AND T. R. BIRKHEAD. 1990. Mecha- avian sperm storage.Ornis Scand.24:323-329. nismsof sperm competitionin birds: Mathemat- BRISKIE,J. V., ANDR. MONTGOMERIE.1992. Sperm size ical models. Behav. Ecol. Sociobiol. 27:325-337. and sperm competition in birds. Proc. R. Soc. MARTIN, P. A., T. J. REIMERS,J. R. LODGE,AND P. J. Lond. B 247:89-95. DZlUK. 1974. The effects of ratios and numbers BROWNELL,R. L., JR., AND K. RALLS. 1986. Potential of spermatozoa mixed from two males on the for spermcompetition in baleenwhales. Rep. Int. proportion of offspring. J. Reprod. Fert. 39:251- Whal. Comm. (Spec.Issue 8):97-112. 258. BURKE, T., N. B. DAVIES, M. W. BRUFORD,AND B. J. MCKINNEY,F., K. M. CHENG,AND D. J. BRUGGERS.1984. HATCHWELL.1989. Parentalcare and mating be- Sperm competition in apparently monogamous haviour of polyandrousDunnocks Prunella mod- birds. Pages523-545 in Sperm competition and ularisrelated by paternity by DNA fingerprint- the evolution of animal mating systems(R. L. ing. Nature 338:249-251. Smith, Ed.). Academic Press, Orlando, Florida. CARTAR,R. V. 1985. Testis size in sandpipers:The MOLLER, A.P. 1988a. Female choice selects for male fertilization frequency hypothesis.Naturwissen- sexualtail ornamentsof the monogamousSwal- schaften 72:157-158. low. Nature 332:640-642. DAVIES,N. B. 1983. Polyandry, cloaca-peckingand MOLLER,A. P. 1988b. Testessize, ejaculatequality 888 JAMESV. BmSrdE [Auk,Vol. 110

and sperm competition in birds. Biol. J. Linn. Soc. SIMS,M. E., G. F. BALL,AND M. F. CHENG. 1987. Sperm 33:273-283. competitionafter sequentialmating in the Ringed MOLLER,A.P. 1991. Sperm competition,sperm de- Turtle-dove. Condor 89:112-116. pletion, paternal care, and relative testissize in SMrrH, R.L. 1984. sperm competition. Pages birds. Am. Nat. 137:882-906. 601-659 in Spermcompetition and the evolution NAKAMURA,M. 1990. Cloacal protuberanceand cop- of animal matingsystems (R. L. Smith,Ed.). Ac- ulatory behaviorof the Alpine Accentor(Prunella ademic Press, Orlando, Florida. collaris). Auk 107:284-295. SMrrH,S. M. 1988. Extra-pair copulationsin Black- PARKER,G.A. 1970. Sperm competitionand its evo- capped Chickadees:The role of the female. Be- lutionary consequencesin the .Biol. Rev. haviour 107:15-23. 45:525-567. TRIV•S, R.L. 1972. and sexual QUAY,W.B. 1986. Cloacal protuberanceand cloacal selection.Pages 136-179 in Sexualselection and sperm in passerinebirds: Comparativestudy of the descent of man (B. Campbell, Ed.). Aldine quantitativerelations. Condor 88:160-168. Press,Chicago. SALT,W. R. 1954. The structure of the cloacal pro- WEATHERHEAD,P. J., AND R. J. ROBERTSON.1980. - tuberance of the Vesper Sparrow (Poocetesgra- ual recognitionand anticuckoldrybehaviour in mineus)and certainother passerinebirds. Auk 71: SavannahSparrows. Can. J. Zool. 58:991-996. 64-73. WOLFSON,A. 1952. The cloacal protuberance--A SHUGART,G. W. 1988. Uterovaginal sperm-storage meansfor determiningbreeding condition in live glandsin sixteenspecies with commentson mor- male passerines.Bird-Banding 23:159-165. phologicaldifferences. Auk 105:379-384. WOLI•SON,A. 1954. Notes on the cloacal protuber- SIBLEY,C. G., AND J. E. AHLQUIST. 1990. Phylogeny ance, ,and possiblecopulatory and classificationof birds: A study in molecular organin malepasserine birds. Bull. Chic.Acad. evolution. Yale Univ. Press, New Haven, Con- Sci. 10:1-23. necticut.