Hybridization in Grosbeaks (Pheucticus) of the Great Plains

HYBRIDIZATION IN GROSBEAKS (PHEUCTICUS) OF THE GREAT PLAINS

D^VlD A. WEST

IN bisexual,cross-fertilizing organisms a speciesmay be definedas a groupof populationsthat showsintrinsic reproductive isolation (actual or potential) from othersuch groups of populations(Mayr, 1942). Sym- patticpopulations are easilyaccommodated by this definition,but in al]o- pattic groupsthe existenceof potentialreproductive isolation is moredif- ficult to determine.It is widely believedthat spatialisolation is necessary for speciesformation (Mayr, 1942; Dobzhansky,1951: 205). Once a populationis dividedand geneflow between the isolatesceases, differential selection and the occurrence of mutation and recombination lead to diver- gence.This is classicaltheory of speciationand needsno further elabora- tion. A later rejunctionof thesepopulations wi]] providea naturaltest of the degreeof intrinsic reproductiveisolation that they have acquired. For this reasonsuch rejunctions, or secondarycontacts, are of interest to the student of evolution. There are two basicoutcomes of a secondarycontact: the populations are reproductivelyisolated, or they hybridize. Actually a continuousseries of possibilitiesexists. Sincethe evolutionof speciesis a gradualprocess, its interruptionat any giventime wi]] resultin somethingbetween hybridi- zationand lack of it. In addition,the ecologicaleompatabi]ity of the two populationswill modify this simplified dichotomy. Hybridizatibnis of commonoccurrence in secondarycontacts. Among birds the studiesof Meise (1936), Sib]ey (1950, 1954, and 1958), Dixon (1955), and Mayr and Gi]]iard (1952) may be mentioned.A distinction shouldbe made betweenthe occasionalhybridization between sympatric species(e.g., ducksof the genusAnas), in whichreduced success of the hybridspreserves the integrityof the species,and hybridizationbetween a]]opatricpopulations in recentsecondary contact. In thispaper the latter case is considered. An exampleof a secondarycontact with hybridizationis foundin two North Americangrosbeaks, the Rose-breastedGrosbeak, Pheucticus ludovicianus,and the Black-headedGrosbeak, P. melanocephalus.These birds, hereinafterreferred to as ludovicianusand melanocephalus,are the North Americanrepresentatives of a smallgenus that alsoincludes P. chrysopeplus and P. aureoventrisof Central and South America. They are more closely related than either is to the Latin American species. They occupy com- plementaryranges, ludovicianus east of the GreatPlains, melanocephalus to the west. They are primarilywoodland birds and do not breedwidely in the Plains. They do occur in the riparian woodlandsof, the Plains

399 The Auk, 79: 399-424. July, 1962 400 W•sT, Hybridization in Grosbeaks [ Auk / Vol. 79 rivers,however, and there they interbreed.It has rarely been doubted that the two formsare goodspecies, despite the evidenceof hybridization reportedby Swenk (1936). The similaritiesof nests,eggs, and habits have often beennoted, however, and it appearsthat the differencesin maleplumage have convincedtaxonomists of their specificstatus. This paperis a reevaluationof the statusof the two grosbeaks.Speci- mensfrom the GreatPlains will be described,and evidencefrom hybridization will be consideredin relationto the specificstatus of the two forms. The paper describesthe biologicalsituation, but a taxonomicjudgment will be made. The investigationis part of a studyof avian hybridizationin the Great Plainsinitiated by CharlesG. Sibleyunder grant G-1832 of the National ScienceFoundation. Other studiesin this project includethe following: Indigo and Lazuli buntings,Passerina (Sibley and Short, 1960); Rufous- sidedTowhees, Pipilo (Sibleyand West, 1958); Flickers,Colaptes (Short, 1959); orioles,Icterus galbulaand bullockii(Short, in prep.). This paper is modifiedfrom a thesispresented to the GraduateFaculty of Cornell Universityin partial fulfillmentof the degreeof Doctor of Philosophy.

METI-IODS AND MATERIALS

Between1955 and 1957 field parties from Cornell Universitycollected 308 specimensof Pheucticusin Nebraska,South Dakota, and Colorado. Material was borrowed from other museums for additional information. Most specimenswere males, and little informationwas gatheredon females. All specimenswere taken between mid-May and late July. Standard measurementsare in millimeters. The length of the black throatof adultmales from the chinto the posterioredge of the blackarea was alsomeasured. In specimenswith an indefiniteposterior margin the measurementextended to the point where black and the breast color were equallymixed. The lengthsof terminalwhite spotson rectriceswere also taken. Hybrid index. Two color charactersare sufficientlydifferent in the two grosbeaks to be used in a hybrid index scheme of the sort commonly employed (Anderson, 1949). The characters are: 1. Breast, belly, and underwing color: rose-red in ludovicianus,yellow in melanocephalus. 2. Extent of brown in plumage: present on underparts,rump, hind neck, and other areas of melanocephalus,absent and usually replaced by white in ludovicianus. Five gradations of each character could be derived from specimens,using only males. These are scored as follows: Belly color "0" rose, as in ludovicianus "1" slightly orange or salmon-pink "2" orange Vol.Auk79 ] WrsT,Hybridization in Grosbeal•s 401

yellow with slight salmon tinge rich yellow, as in melanocephalus Brown extent no brown in areas noted, as ludovicianus trace of hind neck collar; tinge of brown on rump and underparts collar obvious; other areas about half of maximum only slightly lessthan melanocephalus,especially in collar and underparts "4" brown as in melanocephalus Thus a male specimenis indexed on a scale of "0" to "8" when the two characters are summed. When given separately (as "l"q-"3"), the first number refers to belly color. Intermediacyin underwingcolor of malesis expressedeither as a uniform color or as a mixture of pinker and yellowet feathers. Becauseof the individual variation in brown head striping of melanocephalus,that area is not included in the scoring of brown. Correlation o] hybrid index characters. Ideally each character of a hybrid index should be independent. That is, it should not be linked with the other characters.If singlegenes, or genecomplexes, are the bases of the characters,then independentassortment is desired,although dif- ferentialsurvival of somecombinations might mask it. Lackingknowledge of the geneticsof the characters,a correlationcoefficient between the two characterswill give some indication of the degreeof linkage. If, in a schemewith three characters,two are closelylinked, any summedindex will be biasedin favor of the two linked characters.Using 67 specimens of obviouslyhybrid origin,a correlationcoefficient of 0.484 was calculated for the two charactersof the hybrid index. The hypothesisof no correlation is rejectedat the 0.001 level. Despite the dose correlationthere are two reasonsfor retaining the two characters. There are many specimens that do not fall into the categoryof perfect correlation.It would be difficult to compressall the variation of the hybrids into a more simple scheme. In addition, since only two charactersare used, there is no danger of introducinga bias into the summedindex. Male plumage. Comparisonsof male plumage are given in Table 1. Descriptions are modified from Ridgway (1901), and they refer to typical specimens. There is considerablevariation, however, and several points concerning such variation should be noted. 1. Several adult male ludovicianusin nuptial plumage have partly obscuredsupra- and postocular stripes and capital stripes. These are identical in placement to the brown marks in melanocephalusalthough lacking the brown pigment of the latter. 2. All adult male ludovlclanus examined have a concealed hind neck collar of subterminal white bars on the otherwise black feathers. Melanocephalus commonly has several black-tipped feathers in its brown collar. 3. Many adult male ludovicianus have partly concealed,white subterminal bars on the black interscapulars. Melanocephalus feathers are brown-barred on rather lanceolate interscapulars. Some individuals of ludovicianus have similar lanceolate interscapularsthough lacking the brown pigment. 4. The extent of black on the throat of ludovicianus is variable. Some individuals 402 WEST,Hybridization in Grosbeaks [ Auk [ Vol. 79

TABLE 1 PLUMAGE OF ADULT MALES

Plumage area ludovicianus melanocephalus

Throat black, sometimesmixed with black for about 5 mm, rest rose (note 4) brown Breast rose brown, some with partly con- cealedyellow Belly center streak rose, sides white brown with yellow center streak Undertail coverts white pale brown Head and neck black (note 1) black with brown collar (some have head stripes, note 2) Back black (note 3) black and brown striped Rump white, with some black tips brown, with some black tips Underwing coverts and axillars rose yellow Tail black, outer three pairs of black with outer two pairs of rectrices with white tips rectrices white-tipped (note 5) (note 5) have the black restrictedto a narrow band at the chin, as is typical of melanocephalus, while others have the black extending onto the breast. Data are given in Table 2. 5. The numberof white-tipped rectricesvaries in each form. Typical melanocephalus has two pairs of rectriceswith large white tips, while ludovicianususually has three such pairs. Table 3 summarizesvariations among "pure" samples of each form. Female plumage. Ludovicianus females usually have salmon-yellow underwing feathers,while in melanocephalusthis area is pure yellow. There is great variation in ludovicianus,however, as has often been noted. Ridgway (1901: 614) says: "... underwingcoverts and axillarsyellow (maize yellow, chromeyellow or light orange yellow) .... "Of adult femalesfrom northeasternUnited States only two have yellow underwings. The others range from slightly pinkish-orangeto rose-red, as in males. A femalewith rose-redunderwings was collectednear O'Neill, Holt County, Nebraska, on 26 June 1955 and had an enlarged ovary. Moyer (1930) described two similar

TABLE 2 LENGTlt O• BLACK TIIROAT IN IV[Iv[ WITH VARIANCES OF 5OIv[E SAIV[PLES

Locality 3-9 10-14 15-19 20-24 25-29 30-34 35-39 s•

Appalachian Region - 2 10 16 8 4 - 23.4 ( ludovicianus) Mich. and Minn. 2 - 10 12 14 9 2 42.4 ( ludovicianus) Platte Transect (hybrid zone) Blair - 2 3 4 4 - - Schuyler 1 1 1 5 9 12 1 Silver Creek 3 1 1 8 7 4 1 Grand Island 8 3 3 5 7 6 1 96.0 Elm Creek 11 2 - 2 - - - Gothenburg 12 3 2 - 1 - - Western U.S. 63 20 4 .... 6.5 ( melanocep halus) Southern Mexico 13 4 ..... ( melanocephalus ) Auk ] WEST,Hybridization in Grosbeaks 403 Vol. 79 J

TABLE 3 TAIL SPOTLENGTIt, IN lVIIV[

Locality Rectrix 0 +-9 10-19 20-29 30mmup

6 - - - 9 29 5 - - - 14 24 Appalachian Region 4 2 - - 22 14 ( ludo vicianus ) 3 19 14 1 4 - 2 32 5 - - - 1 36 1 - - -

6 - - - 27 6 5 - 3 28 2 Western U.S. 4 - 25 - 7 - (melanocephalus) 3 5 25 - - - 2 13 17 - - - 1 25 3 - - -

The white spot on the inner web is measured from the tip to the basal end of the spot about 3 mm from the rachis. An isolated spot is indicated by q-. specimensfrom near Chicago. Rand (1948) reports one from Medicine Hat, Alberta. There is no lack of such individuals, and their occurrence away from the zone of hybridization in Nebraska excludeshybridization as the responsiblefactor. Females differ in two other ways: breast of ludovicianusis usually heavily streaked; that of melanocephalususually clear bully-yellow. The amount of yellow in the plumage differs. Melanocephalus has yellow in the belly streak, head stripes, etc., while ludovicianus does not.

GENERAL BIOLOGY OF THE GROSBEAKS

Plumagesequence and molts. Dwight (1900) describedthe molts and plumagesof ludovicianus. His sequenceincludes partial postjuvenaland prenuptial molts, as well as a completepostnuptial molt. The only part of his sequencerequiring modification is the first prenuptial molt. First nuptial plumageis acquiredthrough a partial prenuptial molt, including body feathers,terriaries, and most wing coverts. Somesecondaries and a few rectricesare replaced. All primaries are retained. Dwight in- correctly states that the tail is replaced at this time and is black (not brownish as in juvenal plumage) in first nuptial plumage. A first-year ludovicianusretains most of his juvenal rectricesuntil the followingfall. Among nine first-year males taken in May and June (after completion of the molt) the following variation was found: 7 May. Pairs 1 and 2 and right 6 replaced; the rest not. 10 May. Pairs 1 and 6 replaced; right 2 emerging. 11, 12, 15 May. Pair 1 replaced. 22 May. Right 1-6 and left 1-2 replaced. 25 May. Pair 1 and right 3 replaced. 27 May. Right 1-6 and left 1-3 replaced. 5 June. Pair 1 dropped; right 2-6 replaced; left 2-3 emerging. Apparently most individualsmolt the central rectricesfirst (the usual 404 WEST,Hybridization in Grosbeaks L[ Vol.Auk 79 passerinepattern). The remainderare molted,however, with considerable irregularity. Ivor (1944) describedthe molt of ludovicianusin captivity as lastingfrom early Januaryto late April. He foundthat while first-year birdsreplace a few remigesand rectricesin the prenuptialmolt, adults retain all thesefeathers, molting them only in their postnuptialmolt. This is confirmed by specimens. There is a distinctwinter plumagein adult male ludovicianus,the head acquiringstripes similar to thoseof first-year malesand of adult melanocephalus.The rosebreast is well veiled with buff and spottedwith black. Adult nuptialplumage is acquiredby a partial prenuptialmolt, in which the body feathersalone are involved. At this time the rosebreast reaches full expression,and, with few exceptions,the back and head becomecom- pletely black. The tips of the secondariesretained from the previousfall becomeworn and have small notcheswhere there were buffy spotsbefore. The plumage sequenceof melanocephalusis much like that of ludovicianus,although there is no distinctive winter plumage in males, the feathersbeing only slightly buffy-edged. Distribution and migration. Generalizedranges are shownin Figure 1. Two banding recoveriesof ludovicianusin its wintering range suggest that the more easternpopulations winter farther south and hence farther east. Two hybrids, presumablyreared in the Great Plains, were taken from the area where the winter ranges overlap, in southern Mexico. These specimensare describedbelow. Habitat preJerence.Ludovicianus is found in deciduousand mixed de- ciduous-coniferouswoodland throughout its range. At the northern end of the breedingrange (northern British Columbia) it is restrictedto the area designatedby Munro and Cowan (1947) as Peace River Parkland. The habitat is characterizedby "... predominanceof mature aspens growingin semi-openstands with occasionaldense groves of smalleraspens, scatteredwhite spruces,white birch and, in depressionsalong seepage courses,groves of willows." Ground cover is dense. The distributionof aspenforest north of the Canadianprairies delimits grosbeak distribution in the area, with the exceptionof the river valleys of southernManitoba, Saskatchewan,and Alberta. In northeasternUnited States deciduouswoods with dense undergrowth are a commonhabitat. At Ithaca, New York, ludovicianusis abundant in wet woodsat the head of Cayuga Lake. Elms (Ulmus americana), Cottonwood(Populus deltoides), Silver Maple (Acer saccharinurn),and Basswood(Tilia americana)are commontrees, and there is a denseunder- story of shrubs,especially Spicebush (Lindera benzoin) and Elderberry ( Sambucuscanadensis) , and vines. In the southernAppalachians ludovicianus breeds at elevationsabove Vol.Auk79 ] WEST,Hybridization in Grosbeaks 405

ludo vic•nus Summer Winter "::/':• rnelanocephalu$ Summer Winter

Figure l. Approximate breeding and winter ranges of the two North American Pheuctlcus. Data from Dr. E. M. Reilly, unpublished. about 1,000 meters(3,000 feet). In northernGeorgia (Burleigh, 1958) it is restricted to areas of scattered rhododendronthickets, and nests were found here and in yellow birchesand chestnuts.Perrygo (1936, 406 W•sT, Hybridizationin Grosbeaks k[ Vol. 79 unpubl.journal in USNM) reportsits breedingin birch-spruceassociations (Cranberry Glades) in West Virginia. Along the Platte and Missouri rivers in Nebraska this grosbeakbreeds most commonlyin grovesof mature cottonwoods,with understoryvegeta- tion two to three meters high. This formation is similar to that found at Ithaca, New York, and seemsto be the one favoredthroughout the range. Melanocephalusis reported from a greater diversity of habitats, although riparian woodland is most frequently mentioned. Munro and Cowan (1947) report it in coastalBritish Columbiain "mixed and second growth forest in the Puget SoundLowlands." In California Grinnell and Miller (1944) state that willow-cottonwoodassociations are especially favored,and that grosbeaksoccur in montaneforest also, especially where intermixedwith deciduousoaks. They conclude(p. 444): "Perhaps an important factor is local diversity of plant growth and extensive'edge' conditions." Although primarily a bird of mountains and foothills in the West, melanocephalusbreeds on the western edge of the Plains and along the rivers to easternNebraska and Kansas. Here it occupieshabitat similar to that of ludovicianus.Breeding birds often wander to open fields and dry, sparselywooded areas of river bottomland alongthe Platte River to feed, but they apparentlyrequire densevegetation nearby for nesting. Nests and eggs. Many authorshave commentedon the similarity of the nestsand eggsof the two grosbeaks.Reed (1904) describedthe eggsof ludovicianusas greenish-blue,spotted especiallyat the larger end with reddish-brown.The eggso.f melanocephalus,he says, are a paler blue, the spotsbeing less distinct. Coues(1874) states,concerning the eggsof melanocephalus:"In nearly a dozenspecimens I can find no reliable dif- ferencesfrom the eggsof the Rose-breastedGrosbeak." The nestsof the two are extremelysimilar, being built of twigsand rootletsand so frail that the eggsare usually visible from below. Vocalizations. Recordingsof the songs and calls of both forms are available and indicate the similarities better than descriptionscan. Melanocephalushas more mewingand slurred notes in its primary song. In addition the commoncall note (a sharp chink) is a little lesssharp in melanocephalus.The songsand calls appear to servesimilar functionsin the two, however,and they are no more different than those of many geographicraces. Plumage pigments. The grosbeaksdiffer in two types of pigments. The brown colorprevalent in melanocephalusis probably a melanin,while the reds and yellowsare carotenoids,most likely xanthophylls,which are commonin red and yellow feathers (Fox, 1953). Several subspeciesor closely related speciesare known in which the differencesin red and Vol.Auk79 ] WF.sT, Hybridization in Grosbeaks 407 yellowfeathers involve simple differences in carotenoids.Test (1942) has investigatedthe pigmentsof the flickers, Colapres,and Kritzler (1943) those of certain African weavers,Euplectes. Among other examplesnot as yet studied at the pigment level are two South American tanagers, Ramphocelus(Sibley, 1958), the two North Americantanagers, Piranga olivaceaand ludoviciana,and two Indian bulbuls,Pycnonotus (Sibley and Short, 1959). In all these casesthere is somehybridization between the forms,and it seemslikely that thesewill prove to involve quantitativedif- ferencesin carotenoids. There is no reason to believe that the grosbeaks are different from Euplectesand Colaptesin this regard either. Evolution of plumagedifferences. It is important to ask what sources of selectionhave operatedin the evolution of the plumagesof the two grosbeaks. Once the original division of the ancestral populations oc- curred, differencesin selectionwould be expectedto alter the genetic characteristicsof the isolates. Although chanceevents may have been important when the original division took place, it is unlikely that they would continueto be important. It is probablethat selectionis responsible for the presentplumage patterns of the grosbeaks.In this connectioninter- action between the ancestralgrosbeaks and sympatric,but perhaps un- related, speciesmay have played a part. Suchinteractions, although they may be theoreticallyimportant, are difficult to demonstrate. A logical extensionof the idea of reinforcementof isolating mechanismsthrough selectionagainst hybridization is that interaction, however slight, can effect similar results. Sibley (1957) has discussedthis possibility more fully. Ludovicianusis strongly sexually dimorphic comparedwith melanocephalus. It has been suggestedsince Darwin that sexualselection is at least partly responsiblefor the more brightly coloredmales of sexually dimorphicspecies through the greater successsof those individualsin stimulating and attracting mates. On the other hand the dull-colored females have evolved under the impact of selection by predators. In ludovicianusit is difficult to fit the facts to this theory. Males of both the grosbeaksincubate part of each day while femalesalone incubateat night. In addition males assistin brood care, and both sexessing while on the nest. Thus it would seemthat any selectiveadvantage accruing to the concealinglycolored females would be cancelledby the presenceof a brightly coloredmale around the nest when predators (small accipitrid hawks, jays, etc.) could eliminate a whole generationof birds. The occurrenceof hybridizationsuggests that whateverit is in male plumageof the two formsto which femalesrespond in pair formationit is not the detailsof plumage,including a particular color, red or yellow. It maybe insteadthe over-allappearance of the males,including the white 408 W•sT, Hybridizationin Grosbeaks [/ Vol. 79 wingand taft patches,which are the samein the two formsand whichare displayedprominently by courtingmales (Ivor, 1944). The great individual variation that both forms show in certain aspectsof plumage furthersuggests that detailsof plumageare not importantin reproductive success of males. HYBRIDIZATION

Previousstudies. Swenk (1956) summarizedthe know]edgeof grosbeak hybridizationin the MissouriValley region,particularly Nebraska, and describedthe hybridspecimens known to him. The first hybridcollected in Nebraskawas taken in 1920 east of Hastings,in Clay County, and otherswere seen during the following15 years. These were a]] adult malesand were nearly intermediate in breastcolor. In June1950 two male m½lano½½phal•swere seen in Lincoln,Lancaster County, and appearedto be mated to female phenotypicl•dovfcfan•s. They were seen feeding young out of the nest. Severalof the local grosbeakswere bandedat the end of summer,and the followingyear a hybrid, althoughunhanded, returned with them. Swenk describedthe bird, and it closelyresembled a first-generationhybrid raisedin captivity by Ivor (see below). Other hybridshave been reported since 1956 (Brooking,1956; Nebr. Orn. Union Report, 1958). Unfortunate]y,very few specimenshave been taken in Nebraska,and most sight recordsare of intermediateindividuals. Slight tracesof hybridizationare not evidentin the field. Specimensare often labeledas oneor the otherspecies and are thusburied unreported in col- lections. In addition hybrid females are difficult to identify. Other hybrids collectedoutside Nebraska before 1955 include the following: 1. Ad. male. Ja]apa,Veracruz, Mexico. January 1888. F. D. Godman Collection,in the British Museum (N.H.). This and the following are the only wintering hybrids known to me. 2. Sex? San Agustin, 5,700•, 1Vfichoacf•n,Mexico. 23 February year? R. T. Moore Collection. I have not seen the specimen,which is reported by Friedmann, Griscom, ½t •L, 1957 as •hybrid." It is presumably a male. 3. •¾oung" male. Stump Lake, Nelson County, North Dakota. 29 July 1902. L. B. Bishop (reported by Swenk, 1939). This bird had •'the generalcolor of the Black-headedGrosbeak but with underwing coverts and axillars salmon pink." 4. Adult male. Rawlins County, Kansas. 28 July 1936. KU1VINH 21523. This specimenscores "2" q- •2" and is the only one known to me from Kansas. Tiemeier (1937) called it "ludov{cfanus." 5. Two adult males. Wheeler, Charles Mix County, South Dakota. 9 June 1937. UMMZ 128894 and 128896. The former specimen scores "2"q-"2," the latter "4" q- "3." Theseare among the very few hybrids known from South Dakota. Up to 1955 at least nine hybridshad beencollected and anotherhalf dozen seen. The earliest known specimenwas taken in Mexico in 1888, Vol.Auk79 ] WEsx,Hybridization in Grosbeaks 409 while the majority were collectedin Nebraskafrom 1920 to 1940.Paucity of specimensfrom the Great Plains accountsfor lack of knowledge of hybridization.

Changesin distributionin historic times. Swenk mapped the rangesof the grosbeaks and documentedan apparent invasion of eastern Nebraska by melanocephalusbe- tween 1915 and 1930. He also discusseda similar westward expansionof ludovicianus a little earlier. His evidence was largely anecdotal, consistingof personal observations of residentsand a few specimens.He recorded that melanocephalushad appeared for the first time in a number of localities in eastern Nebraska during the period mentioned. He was certain that this form had invaded eastern Nebraska, since competent observershad been in the area before the invasion and had reported no breedingrecords of melanocephalusbefore about 1915. At Hastings,Adams County, melanocephaluswas not reported until 1914, but by 1924 it was common. These records may in part refer to hybrids, as no specimens were collected until 1920. At Crete, 120 km (70 miles) to the east, it was first reported about 1930. At Lincoln presumedmigrants were seen as early as 1911, but it was not until 1930 that breeding melanocephalus(a hybrid pair) were reported. It is probable that this was a true invasion, as the appearanceof melanocephalus fairly recently agrees somewhat with the increase in numbers of hybrids, a result, presumably,of greater reproductive contact between the forms. Unfortunately, there were few observersor collectorsin the Platte Valley, where grosbeaksare now com- monest in Nebraska. Thus evidence from that critical area is lacking. Regarding ludovicianusSwenk (1936: 28) states: "It seemsquite certain that the specieshas extendedits range westwardin the past seventy-fiveyears" (that is, since about 1860). The evidence for this assertion is scanty. In 1900 ludovicianus seems to have bred (or hybridized) nearly as far west as it now does along the Platte. A specimenwas collectednear Kearney in June 1901, and at Elm Creek, a few miles west of Kearney, very few pure ludovicianusare now found. Along the Republican River in northwestern Kansas a hybrid was collected in 1936 (listed above) nearly 200 km (130 miles) west of the nearestlocality (Red Cloud, Nebraska) where ludo- vlcianuswas reported by Swenk the sameyear. There are several recent records of ludovicianus in breeding season from the Nebraska-Colorado border along the South Platte River (near Brule, Nebraska; Nebr. Orn. Union Report, 1950 and 1951). The information does not reveal whether these were breeding, or whether they were mated to female ludovicianus,but their presencethere in mid-June is suggestiveof breeding, and most likely with female melanocephalus.There are earlier reports of breeding ludovlcianusin northern Colo- rado (Burnett, 1902), but there is no evidence of a general increase of that form in the western Plaiv, s. In South Dakota Hayden collectedboth forms on the Missouri River at the Bijou Hills on 16 May 1856 (Baird, 1858), but these in part may have been migrants. A female ludovicianus and several melanocephalus were taken. This is the easternmost part of the Missouri Valley in which Hayden encountered melanocephalus. At the presenttime the influenceof ludovicianusdoes not extendfar beyond this point. The hybrids collectedin 1937 were taken just south of the Bijou Hills. Presentstudy. In 1955 a grid of collectinglocalities was visited in the central Plains (Nebraska and South Dakota) to learn somethingof the distributionand extent of hybridizationof grosbeaksand other bird popu- 410 WESt, Hybridization in Grosbeaks Auk Vol. 79

TABLE 4 SAMPLINGGRm, 1955, wII• HYBRm INVEX SCO•S o• MALES

Number Index scoreso! males Locality c•c• •)•) 0 I 2 3 4 5 6 7 8

Nebraska Crete 1 - - - 1 ...... Hastings 3 - 1 i - - - i - - - St. Paul 2 2 - - - 1 .... i Burwell 2 - i ...... 1 Halsey 1 ...... i O'Neill 2 i - i - - i .... Spencer 2 - i - - - i* .... Bassett 2 ...... 2 Chadron 11 i ...... ii South Dakota Chamberlain 3 i ...... 3 Murdo i ...... i Kadoka 6 4 ...... 6 Midland 3 I .... 1' - - - 2 Howes i i 3 ...... i 1 Rapid City 7 2 ...... 7 Promise i ...... Mobridge 5 1 ...... 5

* Sight records; approximate index. lationsin secondarycontact. It then seemeddesirable to run a transect acrossthe Plains through the zone of hybridization. This was done in 1956 and 1957, and campswere establishedat 80-kin (50-mile) intervals from the MissouriRiver, nearOmaha, to the baseof the Rocky Mountains in northern Colorado. This river systemwas chosenfor its direct route acrossthe Plains and becauseof its fine riparian woodlandnearly all the way. In the transect most of the grosbeakhybrids were collected. The localitiesvisited in the three yearsare shownin Figure 2, the town nearestthe collectingstation beingindicated. In Figure 3 is shownthe distribution of grosbeaksin the Plains. Records from North Dakota are largely derived from unpublishednotes at the North Dakota State Uni- versity, Fargo, suppliedby J. F. Cassel. Table 4 lists the localitiesof the 1955 samplinggrid togetherwith the numbersof specimensand the hybrid indices of the males. Platte Transect, 1955-1957. Eleven localitieswere visited along the Platte and South Platte rivers, with one camp on the Missouri River at Blair, Nebraska, north of Omaha. At each locality collectingwas done along severalkilometers of river bottomland. The samplingtechnique consistedsimply of collectingall grosbeakspossible, with equaleffort spent on eachspecimen and no specialattempt to collecthybrids. In the time available (three days) this seemedthe best system. The Grand Island locality was visited in all three years; othersin 1956 and/or 1957. Histo- Vol..•uk 79 '1] WEs:r,Hybridization in Grosbeaks 411 412 W•.sT,Hybridization in Grosbeaks [ Vol.Auk 79

Figure 3. Distribution of grosbeaks in the central Great Plains. gramsof hybrid index scoresof specimenstaken in the Transectare given in Figure 4. The South Platte River rises in the Rocky Mountains of central Colo- rado and flows northeastwardacross the Plains to its junction with the North Platte in westernNebraska. From this point the Platte runs east- ^uk ] WEST,Hybridization in Grosbeaks 413 Vol. 79 /

Blair Ave "o"

O.002/km

• 5 Schuyler Ave "0 13" 0 2 4 6 8

O.012/•m

Silver Creek Ave "1.1" 1 O.024/km

Grand island Ave "3-0" 0 2 4 ,6 8 T O.040/krn

Elm Creek Ave "6'2"

O.015/krn

Gathenburg ß 0 2 4 6 8 Ave.1i'7-4" O.007/krn

Sutherlond , Ave. "8'0" 0 2 • 6 8

Figure 4. Histogramsof hybrid index scoresfor samplesfrom the Platte Transect. The shift per kilometer in averageindex is given. Localitiesare about 80 kilometers (50 miles) apart. ward to enter the Missourisouth of Omaha. There is a continuousriparian woodlandalong this river systemthat spansthe grasslandof the Great Plainsand provides a dispersalroute for woodlandbirds from the Missouri River to the mountains.It alsoprovides habitat for somewestern species, includingsome mammals, although birds have taken a greateradvantage of it. In partsof the Plains(western Kansas for example)the rivershave their watershedwell east of the mountains,leaving a wide gap between their westernends and the westernedge of the Plains. In other areas (Nebraskafor example)there is a continuouswoodland connection across the grassland. 414 W•sT, Hybridizationin Grosbeaks 1_[ Vol. ^uk 79

Severalwriters, especiallyFremont (1848), were consultedfor information on the state of riparian timber in the Plains before white man had changedit drastically. For more details on this point see West (1959). Two pointsseem valid concerningwoodland along the riversof the Plains during the last 150 years. 1. There has been at least a scatteringof woodlandnearly throughout the length of the Platte and South Platte rivers since before white man arrived there. Islands were more thickly wooded than river banks, but therewas undoubtedly suitable grosbeak habitat at leastas far westas the presentNebraska-Colorado border, althoughperhaps not so continuously distributed. 2. There was reductionthrough lumberingin the original amount of riparian woodlandduring the 19th century. The already narrow belt of trees was reducedto a sparsercondition, while trees were being planted around townsand on the flood plain. Sincethe turn of the century, however, there has been an increasein riparian timber, until at the present time thereis undoubtedlymore such woodland along the Platte than there was 150 yearsago. The increasemay be accountedfor by severalfactors. The reductionin water volumein the Platte, throughincreased diversion for irrigation in the western Plains, has apparently substantially reduced the scouringof banks and subsidiarychannels downstream. Near Grand Island, in central Nebraska, it seemsthat old channelsare being over- grown faster than new ones are being cut by the river. Local residents report that the river is dry in many years in early summer; this is confirmed by personalexperience. The annual floodsthat ravagedthe chan- nels during the spring in the last century are reported to have filled the river from bank to bank. The Platte River is no longer subject to such devastationthroughout most of its course. Fire, which is believedby many to have contributedto the prevention of natural forestson the prairie (Bessey,1914; Gleason,1923), is no longera factor along the rivers. Bessey(1897), speakingof the tributaries of the Loup, Blue, and Missouri rivers in easternand central Nebraska, said: "No one who has seen and studied the forest areas of eastern Nebraska will be able to doubt that they are spreadingwhere they are given a fair opportunity and are not prevented by man or his domestic animals." By now man had replacedfire. Tree planting has contributedmany shelterbeltsand timber claims to the Platte Valley. Becauseof their sparse undergrowththese are not usually suitablehabitat for grosbeaks.In someareas the plantingshave been closerto the river, and undergrowthhas been allowed to develop. There is now little lumberingalong the Platte. Hence natural regenera- tion is not seriouslyimpeded, except sometimesby grazing. Vol.Auk79 ] WEST,Hybridization in Grosbeaks 415

Althoughthe river channelsmay be dry for months,there is undergroundwater in the bed, and this providesmoisture for trees along the banks. One often finds pools of cold water where undergroundwater comesto the surfacein an otherwisedry, sandy river bed. Grosbeakswere amongthe commonbirds at all localitieson the Platte River. The braidedchannels provide much suitable habitat in the islands that lie between them. These have a characteristicvegetational structure includingmuch "edge." The sidesof the islandsusually have a wall of understorytwo to three metershigh with mature cottonwoodsabove. Behindthe wall is a clearingor area of scrubbyvegetation. If the island is cut by a disusedchannel there will be anotherwall on eitherside of it, and it will often have a thicket of willows in its bed. The effect of all this is to providelocal diversityin plant cover,a conditionthat attracts grosbeaks. Hybrid zone. The zone of hybridization is shown in Figure 2. In Nebraskahybrids were collectedat five localitieson the Platte Transect, from Schuylerto Gothenburg.It can be seenthat at no locality were hybridsalone collected. The zone,which here is about320 km (200 miles) wide,is actuallyan overlapzone with hybridization.Since it seemsthat ecologicalfactors are not keepingthe grosbeaksseparated, other reasons must be sought to explain the broad overlap without completebreak- down. Three such reasonswere considered:reduced success of hybrids, assortativemating, and differential immigration. Data on successof hybridsare quite inadequate,and only a little informationwas gathered on assortativemating. Five mated pairs were collectedon the Platte and two more near Hastings. The mated pair taken near Silver Creek sug- gestsassortative mating, sinceboth birds appearedto be melanocephalus, while the rest of the samplecollected there was closeto ludovicianus(index of "4" or less). The remainderof the matedpairs, however,shows commonestclasses paired, not classesmost similar. The informationis meager. If there is a tendencytoward assortativemating it would be strength- ened if immigrationinto new parts of the range were accomplishedby both sexes. Several extralimital records of ludovicianus in the western stateshave apparently involved mated pairs, with both birdsludovicianus. The most recent is reported by Rickard (1960). Femalesare difficult to distinguishfrom those of melanocephalus,but it is possiblethat small groupsof ludovicianusmay migrate togetherand lead to intragroup matings.The pair of melanocephalusat Silver Creek couldbe explained this way. The methodsof rangeexpansion in birds are not well known, but it is presumablyfirst-year birds that do the pioneering,remaining where they go for successivebreeding seasons. Michener and Michener 416 W•sT,Hybridization in Grosbeaks [ Vol. 79

(1951) found that melanocephalusmales wandered more than females within the local area. Factorsinfluencing the width of the hybrid zone. Althoughthe zoneof hybridizationextends from Kansasto southernCanada, it is not of equal width everywhere.Some of the factorsresponsible for this variation are discussed below. In Kansasand Nebraska there are severallarge rivers that crossthe Plains and providerelatively broad areasin which contactscan occur. In southernNebraska the zoneis about 300 km wide,in northeasternNebraska about 200 km. Beginningin South Dakota and extendinginto Canada thereis a gap betweenthe easternriver systems,flowing into the James, Red, and Sourisrivers, and the westernstreams flowing into the Missouri. Many riverscross the areawest of the MissouriRiver in the Dakotasand providehabitat for melanocephalus,but eastof the Missourithe habitat is unsuitable,being mostly treelesscountry, and not spannedby rivers. The rangesof the grosbeaksare effectively separatedhere. There is hybridization along the Missouri River in southernSouth Dakota, but the influenceof ludovicianuscan reach westernSouth Dakota only by this route, and not directly acrossthe easternpart of the state. There is a recordof a hybrid in easternNorth Dakota (Bishop,in Swenk, 1939), but there is no evidenceof more than sporadic occurrenceof melano- cephalusgenes east of the Missouri River in the northern Plains. The samegap extendsacross southern Saskatchewan, but in the Cypress Hills, in southwesternSaskatchewan, both grosbeakshave been found breeding(Godfrey, 1950). Althoughno hybrids are reported,it would be surprisingif nonewere there. A hybridizingpopulation could flourishin the CypressHills, where there is suitable habitat, even though the sur- roundingcountry is open grassland.West of the CypressHills the ranges of the two grosbeaksfinally separate,ludovicianus extending north along the westernedge of the Plains and melanocephalussouth into the foothills of the Rocky Mountainsof Montana. Melanocephalusis unreportedfrom Alberta, evenin the westernedge of the CypressHills, but it must surely occur there and perhapsalong the rivers of the area. Measurements. Statistical summariesof mensural data are given in Tables 6-9 and 11 of West (1959). Averagelengths of wing sampleseast and westof the Plains are not significantlydifferent, althoughwithin the zone there is a gradualshift toward longerwing in the westernsamples. Differences in tail length are more consistentbetween the two forms. Data on tail and tarsal length are given in Figure 5. In generalthe two grosbeaksdiffer only slightly in measurements,melanocephalus in the westernPlains being a little larger and having a deeperbill. Hybrid index. Histogramsof hybrid index for the Platte Transect are vol.Auk79 ] WEST,Hybridization in Grosbeaks 417

Bla,r • 00) • 04)

66 70 74 78 82 86 20 22 24 26 TOH Length m mm Torsus Length in mm Figure 5. Statistical summaries of tail length an tarsal length. Horizontal lines show range; vertical line marks the mean; open rectangle shows one standard deviation either side of the mean; solid rectangle shows two standard errors of the mean either side. Numbers of specimens are given in parentheses. givenin Figure4, the shift in averageindex per kilometerbeing indicated. The figureis self-explanatory.The greatervariation in hybrid populations toward the center of the zone should be noted.

EXPERIMENTAl,, HYBRIDIZATION AND GENETICS The geneticsof wild birdsare poorlyknown, largely because of the dif- ficultiesof making large-scalecrosses. As Danforth pointedout (1950), however,this should not deterone from investigatingthe geneticsof wild species.In severalstudies of polymorphism(Southern, 1945; Coochand Beardmore,1959), countsof phenotypesderived from field work have beenused, and in thosespecies commonly kept by aviculturists(estrildines, ducks,pheasants) the resultsof aviary crosseshave beenused as well. The geneticbasis of polymorphismis often oneor two pairsof alleles (Huxley, 1955). Likewise, major gene differencesoften characterize domesticbreeds of fowlsand ducks (Danforth, 1950). The investigation of geneticdifferences between species or geographicraces is not so easy. Danforth believesthat the first differencesto developbetween isolated forms are of major genes,but that theseare modifiedand reinforcedto 418 W•s%Hybridization in Grosbeaks [ Vol.^uk 79 form multiple-factorsystems. Sheppard (1954) finds that the genetic differencesbetween species are of suchmultifactorial systems, but he believesthat differencesdevelop gradually through the accumulationof small changes.Miller (1941) suggestedboth major genesand multiple factors in the genusJunco, a groupof allopatricpopulations showing more or less complete interfertility. Hinde (1956) found considerablevariation in plumageof offspringbetween two sympatricEuropean finchesand con- cluded that the genetic differencesbetween them are multiple-factor systems.He remarks: "... with someexceptions those characters of plumage which the ethologistwould expectto have significanceas socialsignals dependupon multiple factor differences. The fact is circumstantialevi- dencethat they have survival value." One crosshas been made in captivity between the grosbeaks. H. R. Ivor hybridized them in his aviaries in 1943, using a male melanocephalus from Vancouver, B. C., and a female ludovicianusfrom southern Ontario. Sincethe crossis unreported,I am grateful to Mr. Ivor for allowingme to use his journals and report it here. The following account and quota- tions are from his notes.

The male parent came into full song on 30 April 1943 and mated about 20 May. Nest building was in progresson 25 May, and eggs were laid on 28, 29, and 30 May. The nestlingshatched 8, 9, and 10 June. The youngestdied two days later; the oldest left the nest at 12 days. By 3 July both surviving young showed rosy-orange feathers on their breasts,indicating that they were both males. By mid-February 1944 their heads were getting "quite black, a much better black than one sees on the Rose- breast in the first winter . . . breast buff with apricot showing through and a small streak of apricot extendingtowards the abdomen; abdomen white with tinge of buff towards flanks; primaries seem to be the old ones, brown with whitish patches . . . greater wing coverts black and white . . . secondaries with two new black feathers (one each wing)." In early May the more advanced hybrid was described: "Head fairly goed black, with a few buff feathers; nape same; back black and rich buff striping; rump light cinnamon; tail same black as head, three outer rectrices on each side with pure white on ventral surface; chin black with a few buff feathers; breast very rich mahoganybrown, covered by apricot tinging the feathers; apricot pure on lower breast, forming a fairly wide streak; abdomen white tinged with cinnamon.... "The younger hybrid died on 13 June 1944 and was preserved (skin and partial skeleton, ROMZ 71002). This specimenresembles in nearly every detail a first-year male collectedat Grand Island, Nebraska, on 17 June 1957 (CU 27792). The specimen was scored "l"q-"3," or just intermediate in summed index. The surviving hybrid lived until 18 November 1947. Ivor describedthe bird in detail, and it was evidently much like its sibling. These birds are similar to one of thecommon classes of hybridscoilected in Nebraska,although there is considerable variation in the color of breast and underwing feathers of the presumed first-genera- tion hybrids. In addition the hybrids cannot be divided into discrete categories; rather, a gradual series can be demonstrated between the parental types. This sug- gests that the two grosbeaksdiffer in a small number of genetic factors, but that these are part of a multiple-factor system. It appears further that there is some Voi.Auk79 ] WEST,Hybridization in Grosbeaks 419 dominanceof factors controlling rose breast (as against yellow) and brown (against lack of brown). The hybrid index was devised before Ivor's hybrid was seen and was based on the assumptionof no dominance. Although this may be invalid, the index was retained until more evidence could be gathered.

DISCUSSION AND CONCLUSIONS

Severalauthors (Chapman, 1924, and Mayr, 1942, amongthem) have stressedthat the use of allopatry as a speciescriterion is unrealisticand tends to mask the evolutionaryhistory of such forms, even when they have divergedconsiderably in external morphologyor ecology. These authorsfee] that a]lopatricrelatives are often best consideredmembers of polytypic species.This opinion has receivedsupport from severalrecent studiesof avian hybridization betweena]lopatric populationsin secondary contact. Among many examplesmay be mentionedSib]ey (1950, 1954, and 1958), Mayr and Gilliard (1952), and Short (1959). On the other hand,Vaurie (1955) hasstressed the importanceof carefulstudies of allopatric formsand haspointed out that they are sometimesspecies ("pseudo- subspecies").Selander and Giller (1961) have recently describedsuch a case in a North American pair of grackles. It is apparent that differenceper se is not important in effecting reproductiveisolation between species, but that differencesin particularaspects of structureand physiologyare neededto accomplishthis. The differences in plumage that have evolved in the spatially isolated grosbeaksare not effective in preventing hybridization between them when the two have come into contact. This is reasonable,since these differences did not evolve under selectionagainst interaction between the two forms. The differences are fortuitous in relation to the two grosbeaks. The breedingrange of the commonancestor of the grosbeakswas split at sometime in the recent geologicalpast, probably not earlier than the Pleistocene. Since it is not known when Pheucticus first entered North America from the New World tropics,any discussionof the time of the split must remain highly speculative. If the ancestor was in North America before the last glacial maximum (about 17,000 years ago ac- cording to Martin, 1958), it would have been driven south by the ad- vancingice sheet(see Martin, 1958 for recentsummary). For the ancestor of melanocephalussuitable habitat would probably have existedin Mexico; the populationfrom which ludovicianusevolved might have been pushed into the southeasternpart of North Americaor the West Indies. Alterna- tively, the easternpopulations might have been foundedby a small off- shoot of melanocephalusstock into the West Indies or northern South America. The retreat of the ice sheet and amelioration of climate across North America would have permitted the northward expansionof woodland and with it the populations, now partly differentiated, into their 420 WEST,Hybridization in Grosbeaks [ Auk 1. Vol. 79 present rangesand eventually secondarycontact. The opennessof the Great Plains has existedsince this time and would have preventedgene exchangebetween the two expandingpopulations. The founding of ludovicianusby a small offshootof the larger melanocephalusancestral populationcould have provided the conditionsnecessary for a rapid evolution of the easternform (Mayr, 1954), if one assumesthat ludovicianus is more changedfrom the ancestorthan is melanocephalus.Ludovicianus showsfar more sexualdimorphism, a conditionmost probably derived from the monomorphismcommon in many tropical finchesat the present time. It is alsomuch lessa tropical or subtropicalform than is melanocephalus. It is difficult to predict the future o.f the contact in Pheucticus. There are at least three possibilities: 1. Swamping,by which the bordersof the zone of hybridizationwould expand, and the populationsin the zone and to either side of it would becomeblurred through increasedgene flow. 2. Stabilizationo.f the zone,through selection against hybrids outside it, while hybridizationwould continue within it. This could result in a situationlike that of certain Europeancrows (Corvus coroneand cornix; Mayr, 1942). 3. Reinforcementof isolating mechanisms,by which selectionagainst the hybridswould lead to divergencein thoseaspects of plumage,be- havior, etc. that are effectiveas isolatingmechanisms. This would pro- ducea reversecline with the two forms more different where they come into contactthan wherethey are allopatric. (See Sibley,1957, for further discussionof this possibility.) At the moment the developmentof somethingbetween swamping and stabilizationseems most probable. The zoneof hybridizationwill probably expand,as it apparentlyhas in the recentpast, while in the center of the zonethere may still be "pure" individuals,at least for a time. Thesemay be recombinationproducts that resemblethe parental forms, but if the genetic differencesbetween the grosbeaksare multifactorial this seems unlikely. It doesnot seemlikely that better mate choiceor the developmentof isolating mechanismsthrough selection against the hybrids will cause hybridization to ceaseas the zone of overlap widens. Gene flow from ludovicianuspopulations to the east, where reinforcementwould have no selectivebasis, would tend to prevent the populationsin the zone of overlap from evolving intrinsic isolation. In addition there seem to be in- sufficient ecologicaldifferences between the grosbeaksto allow them to becomesympatric. Whatevermay be the future of the zoneof hybridizationin Pheucticus, auk ] WEST,Hybridization in Grosbeaks 421 Vol. 79 J it is evidentthat there is no well-developedbarrier to gene exchangeat the moment. The level of speciesevolution of the two grosbeaksseems best describedbiologically by regarding them as membersof a polytypic species.I chooseto disregardthe Pacific coast race of melanocephalus (maculatus)in view of the doubtful positionof the populationsin Mexico and until a full study of variationthere can be done. The grosbeaksthen become:Pheucticus ludovicianus ludovicianus (Linnaeus), Rose-breasted Grosbeak,and P. I. melanocephalus(Swainson), Black-headedGrosbeak.

ACKNOWLEDGMENTS

I am gratefulto C. G. Sibleywho suggestedthe problemand advised me throughoutthe study. To the other membersof my graduate com- mittee, L. C. Cole and A.M. Srb, I am indebtedas well. The field work was carriedout with the assistanceof the following: W. G. Gibson (1955), A. L. Nordby (1955), L. L. Short,Jr. (1955-1956), C. G. Sibley (1955), F. C. Sibley(1955-1957), J. H. Smith (1955), and D. B. Wingate(1955- 1957). Financialsupport came from severalsources: the above-mentioned National ScienceFoundation grant (G-1832) to Dr. Sibley; a grant from the Frank M. Chapman Memorial Fund of the American Museum of Natural History, 1956; and a Sigma Xi-RESA Grant, 1957. Without this supportthe work couldnot have been done. Vehicleswere provided by the New York State Collegeof Agricultureand equipmentby the De- partment of Conservation,Cornell University. Personsin Nebraska who assistedin variousaspects include Doris Gates, Burr Nelson, William F. Rapp, and officials of the Nebraska Game, Forestationand Parks Commission,especially Lloyd Vance and the late Paul Gilbert. Unpublisheddata were kindly supplied by J. F. Cassel, H. R. Ivor, and E. M. Reilly, Jr. Curatorsof the followingmuseums are thanked for their help: American Museumof Natural History, British Museum (N.H.) (BMNH), Royal Ontario Museum (ROMZ), SaskatchewanMuseum of Natural History, Museum of Comparative Zoology, University of Kansas Museum of Natural History (KUMNH), University of Michigan Museum of Zoology (UMMZ), Universityof MinnesotaMuseum of Natural History, and the United States National Museum (USNM). Abbreviationsare used as indicated. I have benefited from discussionswith Walter Bock, L. L. Short, Jr., F. C. Sibley, P. A. Johnsgard,and R. G. Wolk.

SUMMARY

A collection of 308 grosbeaksfrom the Great Plains, representing Pheucticusludovicianus, melanocephalus, and hybrids between them, is 422 W•sr,Hybridization in Grosbeaks [ Vol.auk 79 described.The two formscome into secondarycontact along many of the river valleysof the central and northernPlains. The zone of overlap and hybridizationis about 320 km (200 miles) wide acrossits widestpoint in Nebraska. The grosbeaksare geographicallycomplementary forms. Habitat prefer- ence, nests,eggs, and vocalizationsare very similar. Although the males differ in certain aspectsof plumagecolors, there is a great similarity in plumagepattern. Someaspects of malesof melanocephalusare represented in many adults of ludovicianusin less-developedform. Female plumages are almost identical. There is overlap in mensuralcharacters investigated. Where their rangesoverlap in the Plains the two forms interbreed,and variation found in hybrids suggeststhat backcrossingand production of secondgeneration hybrids is occurring. Two hybrids reared in captivity by H. R. Ivor closelyresemble one of the classesof hybridscollected in the hybrid zone in Nebraska. There is someevidence of increasedrange overlapduring this century. The two grosbeaksact as membersof a polytypic species;they are not more different than membersof a number of geographicallyvariable species,and they interbreed where their ranges come into contact.

LITERATURE CITED

ANDERSOn,E. 1949. Introgressivehybridization. New York, John Wiley and Sons. ix q- 109 pp. BAIRD, S. F. 1858. Birds, in Pacific railroad reports. Vol. 9. Washington, D.C. lvi q- 1,005 BESSE¾,C. E. 1897. Are trees advancing or retreating on the Nebraska plains? Science,n.s., 10: 768. BESSE¾,C.E. 1914. Plant migration studies. Univ. Nebr. Studies,5: 11-27. Baoo•, Mas. A.M. 1936. Another hybrid grosbeak seen in the Hastings vicinity. Nebr. Bird Rev., 4: 82. BURLEIGH,T.D. 1958. Georgia birds. Norman, Univ. Okla. Press. xxix q-746 BUR•ETT, W.L. 1902. The Rose-breastedGrosbeak in Colorado. Condor, 4: 94. CI•APMAk•,F. M. 1924. Criteria for the determination of subspeciesin systematic ornithology. Auk, 41: 17-29. Cooc•, F. G., and J. A. BEARDMORE.1959. Assortative mating and reciprocal differencesin the Blue-Snow Goose complex. Nature, 183: 1833-1834. Couzs, E. 1874. Birds of the Northwest. Washington, Dept. of the Interior. xi q- 791 DANFORTt•,C.H. 1950. Evolution and plumagetraits in pheasanthybrids, Phasianus X Chrysolophus. Evol., 4: 3Ol-315. D•xo•, K.L. 1955. An ecologicalanalysis of the interbreedingof crestedtitmice in Texas. Univ. Cal. Publ. Zool., 54: 125-206. DOBZHA•$KY,T. 1951. Genetics and the origin of species. 3rd Ed., revised. New York, Columbia Univ. Press. xii q- 364 DWi•I•T, J., JR. 1900. The sequenceof the plumages and moults of the passerine birds of New York. Annals N.Y. Acad. Sci., 13: 73-360. Vol.Auk 79] WEST,Hybridization in Grosbeaks 423

Fox, D. L. 1953. Animal biochromes and structural colours. Cambridge, Cam- bridge Univ. Press. xiv q-379 pp. FREMONt,J. C. 1848. Narrative of the exploring expedition to the Rocky Moun- tains in the year 1842, etc. Syracuse.427 pp. FRIEDMANN,H., L. GRISCOM,R. T. MOORE,and A. H. MInLER. 1957. Distributional check-listof the birds of Mexico. Part 2. Berkeley, Cooper Orn. Soc. 470 pp. GLEASON,H.A. 1923. The vegetational history of the Middle West. Annals Assoc. Amer. Geogr., 12: 39-85. GODV}•E¾,W. E. 1950. Birds of the Cypress Hills and Florten Lake regions, Sas- katchewan. Bull. Nat. Mus. Canada, No. 120, 96 pp. GRINNEnL,J., and A. H. MInLER. 1944. The distribution of the birds of California. Berkeley, Cooper Orn. Soc. 608 pp. HINDE, R.A. 1956. Breeding successin carduelineinterspecies pairs, and an exami- nation of the hybrids' plumage. J. Genet., 34•: 304-310. HUXLEY, J. S. 1955. Morphism and evolution. Heredity, 9: 1-52. IVOR,H. R. 1944. Bird study and semicaptivebirds: the Rose-breastedGrosbeak. Wils. Bull., 36: 91-104. KRITZLER,H. 1943. Carotenoidsin the display and eclipseplumage of bishop birds. Physiol. Zool., 16: 241-255. MARTIN, P.S. 1958. Pleistocene ecolGgy and biogeography of North America. In Zoogeography. Washington, Amer. Assoc. Adv. Sci. x q- 477 pp. MAYR, E. 1942. Systematicsand the origin of species.New York, Columbia Univ. Press. xiv q- 334 pp. MAYR, E. 1954. Change of genetic environment and evolution. In Evolution as a process. London, Allen and Unwin. 367 pp. MAYR, E., and E. T. GILLIAV,O. 1952. Altitudinal hybridization in New Guinea honeyeaters. Condor, 34•: 325-337. MEISE,W. 1936. Zur Systematikund Verbreitungsgeschichteder Haus- und Weiden- sperlingePasser domesticus(L.) und hispaniolensis(T.). J. Orn., 84: 631-672. MmHENER, J. R., and H. M•C}rENER. 1951. Notes on banding records and plumages of the Black-headedGrosbeak. Condor, 33: 93-96. Mn;zE}•,A.H. 1941. Speciationin the avian genusJunco. Univ. Calif. Publ. Zool., 4•4•: 173-434. MOYER, J. W. 1930. Unusual markings on Rose-breastedGrosbeaks. Auk, 47: 567-568. MUNRO, J. A., and I. McT. COWAN. 1947. A review of the bird fauna of British Columbia. Brit. Col. Prov. Mus. Spec.Publ. No. 2: 1-285. NEBRAS•:AORNIT}rO•OOISTS' UNION. 1938. Report on the thirty-sixth annual field day of the N.O.U. Nebr. Bird Rev., 6: 47-48. NEBRAS•:AORNIT•{O•OOISTS' UNION. 1950. Twenty-sixth cooperative spring migration report. Nebr. Bird Rev., 18: 77-79. NEBRAS•:AORNIrHOLOOISTS' UNION. 1951. Twenty-seventh cooperative spring migration report. Nebr. Bird Rev., 19: 51-81. RAND, A. L. 1948. Birds of southern Alberta. Bull. Nat. Mus. Canada, No. 111. 105 pp. REED,C.A. 1904. North Americanbird eggs. New York, Doubleday,Page and Co. 356 pp. Rxc•:Am),W. H. 1960. An occurrenceof the Rose-breastedGrosbeak in southern Nevada. Condor, 62: 140. 424 WEST,Hybridization in Grosbeaks [1. Vol.t•uk 79

RIDGWAY,R. 1901. The birds of North and Middle America. Bull. U.S. Nat. Mus., No. 50, part 1. xxx q-715 pp. SELA>•DER,R. K., and D. R. G•LER. 1961. Analysisof sympatry of Great-tailed and Boat-tailed grackles.Condor, 63: 29-86. SnErrARD,P.M. 1954. Evolution in bisexually reproducingorganisms. In Evolu- tion as a process. London, Allen and Unwin. 367 pp. SaORT,L. L., JR. 1959. Hybridization in the flickers (Colapres) of North America. Ph.D. Thesis, Cornell University. 226 pp. Sm•EY, C. G. 1950. Speciesformation in the Red-eyed Towhees of Mexico. Univ. Calif. Publ. Zool., 50: 109-194. Sm•EY, C. G. 1954. Hybridization in the Red-eyed Towhees of Mexico. Evol., 8: 252-290. Sm•Ez, C. G. 1957. The evolutionary and taxonomic significanceof sexual dimor- phism and hybridization in birds. Condor, 59: 166-191. Sm•Ez, C. G. 1958. Hybridization in some Colombian tanagers, avian genus Ramphocelus. Proc. Amer. Phil. Soc., 102: 448--453. SmnEz, C. G., and L. L. SaORT,JR. 1959. Hybridization in some Indian bulbuls Pycnonotusleucogenys X P. cafer. Ibis, 101: 177-182. SmnEZ,C. G., and L. L. SaORT,JR. 1959. Hybridization in the buntings(Passerina) of the Great Plains. Auk, 76: 443-463. StonEY,C. G., and D. A. WEST. 1959. Hybridization in the Rufous-sidedTowhees of the Great Plains. Auk, 76: 326-338. SOUThERn,H.N. 1945. Polymorphism in Poephila gouldiae Gould. J. Genet., 47: 51-57. SWE•X, M.H. 1936. A study of the distribution,migration and hybridismof the Rose-breastedand Rocky Mountain Black-headed grosbeaksin the Missouri Valley region. Nebr. Bird Rev., 4: 27-40. Sw•>•x, M.H. 1939. The speciesof Hedy•nelespresent in the Devil's Lake region of North Dakota. Nebr. Bird Rev., 7: 22. TEST, F. H. 1942. The nature of the red, yellow and orange pigmentsin wood- peckersof the genusColapres.' Univ. Calif. Publ. Zool., 46: 371-390. Tm•vrEr•R,O. W. 1937. Summer birds of Rawlins County, Kansas. Trans. Kans. Acad. Sci., 40: 397-399. VAtmrs,C. 1955. Pseudo-subspecies.Acta XI Congr.Int. Orn., 1954: 369-380. WEST,D.A. 1959. Hybridizationin the grosbeaks(Pheucticus) of North America. Ph.D. Thesis, Cornell University. 95 pp. Department o/ Conservation,Cornell University, Ithaca, New York. Present address:Sub-department of Genetics,University of Liverpool, Liverpool,England.