Hybridization in the Red-Eyed Towhees of Mexico: the Populations of the Southeastern Plateau Region

THE AUK

A QUARTERLY JOURNAL OF ORNITHOLOGY

VOL. 81 OCTOBER,1964 No. 4

HYBRIDIZATION IN THE RED-EYED TOWHEES OF MEXICO: THE POPULATIONS OF THE SOUTHEASTERN PLATEAU REGION

CHARLES G. S1BLEY AND FRED C. SIBLEY 1

THE extensiveand complexpatterns of variation resultingfrom hybridization betweenthe Rufous-sidedTowhee (Pipilo erythrophthalmus)and the CollaredTowhee (Pipilo ocai) in Mexico have been describedin three previouspapers (Sibley, 1950, 1954; Sibleyand West, 1958). The evolu- tionaryand taxonomicsignificance of hybridization,with particularrefer- enceto birds,has beenconsidered in threeadditional papers by the senior author (1957, 1959, 1961). Other papers,forming a related seriescon- cerned with avian hybridization, are Sibley (1958), Sibley and Short (1959a, 1959b, 1964), Sibley and West (1959), Short (in press), and West (1962). The presentpaper is basedupon new material collectedin 1958 in an attempt to clarify the hybrid towheesituation in easternand southeastern Mexico.

ACKNOWLEDGMENTS

We are especiallygrateful to Roger B. Clapp and S. Craig Smith, who assistedthe junior author in the collectionof the material. The support of the Faculty Research Grants Fund and the New York State College of Agriculture of Cornell University made the field work possible. Field equipment purchased with funds provided by N.S.F. Grant 1845 was utilized.

WEIGIITS AND MEASLIREM2ENTS

All the specimenscollected in 1958 were weighedto the nearest tenth of a gram. Linear measurementsfrom the skins were taken with dial calipers to the nearest tenth of a millimeter as follows: wing, the chord; tail, from the insertion of the two middle rectrices;tarsus, from the joint between the tarsus and the tibia (the heel) to the midpoint of the distal margin on the most distal undivided scute; bill, from the anterior margin of the nostril to the tip of the maxilla.

The authors are not close relatives.

479 The Auk, 81: 479-504. October, 1964 480 SIBLE¾AI•D SiBLE¾, Hybridization in Towhees [ Vol.Auk 81

HYBRID INDEX

The hybrid index methodof analyzinghighly variable populationswas first developedby Anderson(].949), in a seriesof studieson plant hybrids, and has since been used in many other studies. This method provides a quantitative evaluation of the complicatedcolor patterns resulting from extensivebackcrossing. The two speciesof towheeswhich hybridize in Mexico have been de- scribedand illustratedin a previouspaper (Sib]ey, 1950). The following synoptic descriptionpresents the six major plumage differencesused in this hybrid index.

1. Pileum color: chestnutin ocai, black in ½ryrhrophrhalmu3.2. Back and wingspots: white spotson scapularsand wing covertsin ½ryrh•ophrhal- mu3, theseareas unspotted in ocai. 3. Back color: greenin oca{,black in ½ryth•ophrhalm•3.4. Throat color: white in oca{, black in ½•yrh•ophrhal- m•3. 5. Flank color: brownish olive in ocai, rufous in ½ryrhrophr/•alm•3. 6. Tail spots: white spots on three outer rectricesin ½•yrh•ophrhalm•3, white spotsabsent in oca{. For each of these six charactersfive gradationsfrom pure ocai to pure ½ryt/•rop/•t/•al•$ may be distinguished.These gradations are scoredas follows:

0. When expressedas in pure P. ocai. 1. Mainly as in ocai but showing some influence from erythrophthalmus. 2. Intermediate between ocai and erythrophthalmus.3. Mainly as in erythrophthalmusbut showinginfluence from ocai. 4. When expressedas in pure P. erythrophthalmus. A pure ocai specimenwill thus be scored 0 for all six characters for a summatedscore of 0. A pure erythrophthalmuswill score4 for each of the six charactersfor a summatedscore of 24. A more detailed description of this hybrid indexis givenin the originalpaper on hybridizationbetween thesetwo species(Sibley, 1950). Weight is the one characterwhich most closelyfollows the variation of the hybrid index as seenin a comparisonof Figures 1 and 2. The weights of individual birds generallygive a good indication of the hybrid index (Correlation coefficient: .87). None of the other four charactersmea- sured has such a high correlation with the hybrid index although the means follow much the same pattern as those for weight. The measurements are larger in oca• than in erythrophthalmus,but due to the large overlap birds cannot be separatedby one of these measurementsalone. There is usually no overlap in weight between the two speciesat a locality where they occur in sympatry. Oct. SmLEY A•D SmLEY, Hybridization in Towhees 481 1964

Tlaxco I1 Av. 22.0

Teziutl&n 5.2

20 km. N. Perote Av. 11.6

6 km. W. LasVigas A¾. 7.3

Cofre de Perote Av. 2.2

'.? 15 Av. 1.5

Cuaulotolapan • 20

• 2•

Av. 1.4

12 mi. N. Mt. Orizaba

Av. 1.2 Av. 22.2 •5

Tlachlchuca 20 0 3 6 9 12 15 18 21

Figure 1. Histograms of some of the eastern plateau populations. Average hybrid indexes are connected by the lines between samples. Except for the top sample the distancesbetween base lines are proportional to the actual air line distancesbetween populations.

SUMMARY OF PREVIOUS RESULTS

In previous papers (Sibley, 1950, 1954; Sibley and West, 1958) the distributionsand hybrid indexesof many of the towheepopulations have been described. In Figures 3 and 4 this previouswork, plus the material describedin this paper, is presented.Figure 3 gives the presentlyknown distributionof Pipilo ocai and P. erythrophthalmuswith the hybrid index valuesof the populationswhich have been sampled. Figure 4 givesa more detailed view of the eastern plateau populations. Figure 5 presentsa diagrammaticsynopsis of the easternplateau populations. A seriesof hybrid populationsat the western end of the plateau con- nectsP. erythrophthalmusin the Sierra Madre Occidentalto P. ocai in the southwest. The populationsin the trans-plateauarea form a clinal series 482 SIBLEYAND SIBLEY, Hybridization' in Towhees [ Vol.Auk 81

22.0 (6) • Tlaxco 15.2 .• Teziutlc•n (32) • 11.6 20km. N. Perote ' (12)

•,,r'r• (12)7.3 6 kin.W. Las¾igas

15 • • 2.2 CofredePerote ' (15)

.10

I I I 1.5 Cuaulotolapan (8)

20.4• I I j 1I 1.4 12mi. N.Mt. Orizaba

22.2 (7),, [ !'• • [ • i/ 1.2(9) Tlachichuca (16) , ' (13)

Weight of malesin grams

35 40 45 50 55 60 65 70

Figure 2. Statistical analysis of weights of male red-eyed towhees from the eastern plateau. See Figure 6 for explanation of symbols. 1964Oct.] S•BL]•¾AND SmLE¾, Hybridization in Towhees 483

i /-- : 484 Smoky^•l) SmokY, ttybridization' in Towhees [ Vol.Auk 81 1964Oct. ] SmLE¾A•D SmL•¾, Hybridization in Towhees 485

Pachuca Tacopan( 13 km.W. Teziutl•n) 15.5 .•,,(•.,-,,,.,,•Teziutl•n(areaaverage • 15.2) ?'•'•0•Zaqu__•,•,•20i;lmC •N(• p•ir;•. Teziutl•n, 11.6 • TJaxcc •7.3 6kin. W. Las Vigas Las Mesas '•? ( )2.2 CofredePerore 14mi.• '0 •1.5 2mi. E.Cuaulotolapan alinche ElS•o-20.4•1.4 6mi. E. Huecapa Mr.Ixtaccihuatl• Seco ==.=•'• •=mi.•.•.O•i==b= 19.8 • ' _•'• 1:210 mi. ESE. Tlachichuca ..e.....,..,•;•Mt.Popocatepetl •. 23.'/•,,3 •t OrJzaba Trans-Plateau Gradient . 23•erto Morelos ..... • .5 Tehuacanß •

occasionalsympatric hybrids situation not well known •Zoqu•tl•n

interrupted distribution 23.0(• 0 • probablycontinuous distribution but populationsmall R(oSanto Domingo • • largepopulation bothspecies present andsyrupattic

? ? (23.6)• • 0 • Statute Miles

0 10 20 30 Cerro San Fellpe Figure5. The patternof hybridizationand genemigration in the red-eyedtowbees in eastern Mexico. of populationsconnecting erythrophthalmus in the east(Hidalgo; Puebla) to ocai in the west (Jalisco; Colima). Anotherhybrid cline connectsPipilo erythrophthalmusin the Tlaxco area with P. ocai on the Cofre de Perote and Mount Orizaba. On Mount Orizabathe two speciesexist in $ympatryalthough there is evidenceof somehybridization between them. Since1946 the two specieshave beenknown to exist sympatrically,ap- parentlywithout interbreeding, on CerroSan Felipein Oaxaca. However, it was not known how far west Pipilo erythrophthalmusextended toward Guerrero,whether Pipilo ocai occurredon Cerro Yucuyacua,or if any hybridizationoccurred there. A previousattempt to reachthis area in 486 S•LE•rAND SmLmr, Hybridization in Towhees [ Vol.^uk 81

1954 failed becauseof poor roads. In the eastern plateau area there was no known connection of suitable habitat between Mount Malinche and Mount Orizaba, although sucha connectionwas indicatedby the previous data. The extent of the range of Pipilo erythrophthalmusbetween Mount Orizaba and the Cofre de Perote was known from only a few specimens. The field work in 1958 was therefore planned to collect in the area between Mount Orizaba and the Cofre de Perote, the area between Mount Malinche and Mount Orizaba, and on Cerro Yucuyacua in Oaxaca.

F•ELD STUD• •N 1958

The field party, composedof Roger B. Clapp, S. Craig Smith, and Fred C. Sibley, collectedduring June and July at the following localities (the number of adult towhee specimenscollected at each locality is given in parentheses): 12 miles N Mr. Orizaba, t0,000 feet, Puebla, 14-18 June (12); 14 miles NE E1 Seco,8,500 ft., Puebla, 18-19 June (4); 2 miles E Cuaulotolapan, 9,500 ft., Veracruz, 20-22 June (14); Cofre de Perote, t0,000 ft., 5 miles E Tenextepec, Veracruz, 23-25 June (20); 6 miles E Huecapa, 10,500 ft., Puebla, 26-27 June (t0); Route 140, 3 miles SW El Seco, 8,000 ft., Puebla, 28-29 June (52); t0 miles ESE Tlachichuca, 10,000 ft., Puebla, 30 June- 2 July (40); Cerro Yucuyacua, 10,800 ft., 8 miles S Tlaxiaco, Oaxaca, 5-10 July (63). The populations from 12 miles north of Mount Orizaba and 6 miles east of Huecapa are only three miles apart and are treated as one population in this paper. The specimens of Pipilo erythrophthalmus from Cerro Yucuyacua are combined with the t0 collected near Tlaxiaco in 1954.

ECOLOGICAL DISTRIBUTION

Pipilo erythrophthalmusinhabits a drier, warmer, and more openhabitat than doesP. ocai. In the eastern plateau area this means that erythrophthalmuswill occur where there is a mixture of brush and pine or oak. Such habitat occursmainly betweenapproximately 9,000 and 10,000 feet in this area but there is considerable local variation and under suitable conditionserythrophthalmus occurs as low as 7,000 feet and up to 11,000 feet. P. ocai is mainly restricted to the cooler, moister fir and pine forests of the higher elevationsabove the usual range of erythrophthalmus. Al- though mainly found between9,500 and 12,000 feet ocai occursat lower elevationsif cool, moist conditionsprevail, as at Puerto Morelos. Thus, to someextent, the two speciesreplace one another both ecologicallyand altitudinally but they overlap in many places.

MOUNT MAL•NCI•IE--MOuNT ORIZABA CONTACT

Mount Malinche rises to 14,636 feet and is coveredwith an open pine forest from approximately 9,500 to 13,300 feet. Collecting in 1954 re- vealed the presenceof Pipilo erythrophthalmus in limited numbers and the presenceof suitablepathways of dispersalto the Tlaxco area approximately 25 miles to the north (see Figure 4). Towhees were also found on 1964Oct. ] SIBLEYA>•D SZBT•EY, Hybridization in Towhees 487

1.2 ,,r-r'd'•'•,,, , Tlachichuca ' (13)

{16}/ o 23.0 ,,r._d•,, I J J J 0.5 PuertoMorelos lO ' (13)

•- 20 (38) • 30 22.4 •-J'•t•• • El Seco (35) '

Mt. Malinche 22.0(24) ,.r...t_f,.,.,.,•,, •

Weight of males in grams

35 40 45 50 55 60 65 70

Figure6. Statisticalanalysis of the weightsof malered-eyed towhees from the area between Mount Orizaba and Mount Malinche. Numbers in boldface type are average hybrid indexesbased on color; numbersin parenthesesindicate the numberof specimensin the sample.Horizontal lines represent range; open rectangles indicate one standarddeviation; solid b!ack rectanglesindicate twice the standarderror of the mean; the mean is indicatedby a vertical line. The distancesbetween samples are proportionalto the actual air line distancebetween populations. a small volcanic cone about 10 miles SE of Mount Malinche, but it was not known if there was a contact between Mount Malinche and Mount Orizaba. Sucha contactappeared unlikely from the nature of the vegetation in the intervening area. Threemiles SW E1 Seco.--Thislocality is locatedat 8,000 feet just off Route 140. The vegetationis composedof low oaks and thick brush with taller pinesand madrones,and is restrictedmainly to the northern sideof the low ridgebetween Mount Malincheand Mount Orizaba. Suit- able towheehabitat is about a mile wide at the widest point and extends at least five miles west along the ridge toward Mount Malinche. It con- tinueseastward along the road betweenE1 Secoand Atencofor a short distanceand then extendseastward along a seriesof low hills to the north of this road. Towheeswere unusuallyabundant and the 52 specimens collectedhere have hybrid indexes between 20 and 24 with an averageof 22.5 for the series. This value is intermediatebetween the averagehybrid indexof the Mount Malinchebirds (22.0) and the Mount Orizababirds (23.4). The averageweight (Figure 6) alsofalls between those of Mount Malinche and Puerto Morelos. Measurementsare given in Table 1. 488 SiBLF.¾A•) SIBLE¾, Hybridization in Towhees [ Vol.Auk 81

TABLE 1 Mr•ASU•E•r•?S OV ?• SA•VLr• OV P. •rr•t•ov•tr•,•t•s F•o• T•r• MILr•S SW or EL S•.co

Item Sex Number of Mean with Standard Coefficient specimens standard error deviation of variation

Weight • 35 45.1 ___0.4 2.5 5.5 9 17 44.0 ñ 1.0 4.0 9.1 Wing • 35 86.0 q- 0.4 2.3 2.7 • 17 81.4 q- 0.6 2.3 2.8 Tail • 31 101.7 q- 0.7 4.0 3.9 • 13 96.2 q- 1.0 3.7 3.8 Tarsus • 32 28.7 ñ 0.2 0.9 3.1 •2 16 27.8 q- 0.2 0.7 2.5 Bill from • 35 9.9 q- 0.1 0.4 4.0 nostril • 15 9.9 q- 0.3 1.2 12.2

This contact,then, is apparentlyproviding a pathway for gene flow from the trans-plateaugradient, via Mount Malinche, to the birds on Mount Orizaba and is partly responsiblefor the ocai genes which are presentin the erythrophthalmuspopulation on Mount Orizaba. The birds of Mount Orizaba exhibit lessevidence of ocai genesthan birds from E1 Seco. The original dispersalof P. erythrophthalmusto the south may have been along this contact, which still exists between Mount Orizaba and Mount Malinche. Present knowledge of the distribution of P. erythro- phtha.lmusrules out dispersalin the past from the Cofre de Perote area or from the west throughGuerrero. Dispersalalong either of theseroutes would probably have resultedin present-daypopulations existing there. Instead,at presentwe find P. erythrophthalmusextending into theseareas from the south and east respectively.

MOUNT ORIZABA--COFRE DE PEROTE CONTACT Previous work showeda hybrid cline to the north (Teziutlgn) of the Cofre de Perote and both speciesoccurring, with little hybridization,on Mount Orizaba (Sibley, 1950; Sibley and West, 1958). Previouscollec- tions had been made on the SW slope of Mount Orizaba and two speci- mensfrom the Cofre de Perote were indexedat 2 and 3. No specimens were available from the 35 miles betweenthese two peaks. In the sum- mer of 1958, 95 specimenswere collectedat five localities betweenthese peaksto determinethe northernlimit of P. erythrophthalmuson the ridge and to determinethe pattern of hybridization and introgressionin the intervening populations. Ten milesESE of Tlachichuca,10,500 feet.--This locality is on the NW side of Mount Orizaba some 600 feet below a saddle between it and a 1964Oct. ] SIBLE¾AND SIBLE¾, Hybridization in Towhees 489

TABLE 2

MEASUREMENTS OF TIlE SAkVIPLEOF P. 0C•4I FROM 10 MILES ESE o•r TLACIIlCIIUCA

Item Sex Number o! Mean with Standard Coe/ficient specimens standarderror deviation o! variation

Weight • 13 58.2 +- 0.5 2.0 3.4 • 8 56.2 -+- 0.7 1.9 3.4 Wing • 13 86.1 +- 0.8 2.9 3.4 • 8 81.8 ___0.8 2.3 3.8 Tail • 8 102.5 ___0.9 2.5 2.4 • 7 95.1 +- 2.2 5.9 6.2 Tarsus • 12 31.0 +- 0.2 0.7 2.2 9 7 30.0 -+- 0.8 2.0 6.7 Bill from • 13 10.6 -+- 0.2 0.6 5.7 nostril 9 8 10.6 -+- 0.1 0.4 3.7 small peak to the north. Pines start to occurat 9,000 feet, are joined by firs at about 9,500 feet, and disappearentirely around 10,000 feet. The fir forestcontinues to about 10,800feet on the ridgeand higheron Mount Orizaba. The top part of the ridge is coveredwith sparsepines and bunch grass. This seriesof vegetationalzones is found along the entire ridge between Mount Orizaba and the Cofre de Perote. Almost all of the area is planted to potatoes although there are still numerouspatches of brushand fir aroundthe edgesof the fieldsor on the steeperparts of the fields. Also the ravinesand the forest above 10,500 feet are relatively undisturbed. P. ocai and P. erythrophthalmuswere about equally abundant and both were frequently found in the same isolatedpatches of vegetation,although no mixed pairs were observed. P. ocaioccurred in a wider rangeof habitats,being common in the thick, wet, fir forest, while P. erythrophthalmusoccurred only on the edge of this area and in the scatteredpatches of vegetationin the fields.

TABLE 3 MEASUREMENTS OF TIlE SA2VIPLEOF P. ERYTHROPHTHALMUS FROM 10 MILES ESE o• TLACIIICIIUCA

Item Sex Number o! Mean with Standard Coe//icient specimens standarderror deviation o! variation

Weight • 16 45.1 ___0.8 3.3 7.3 9 2 47.1 Wing • 16 84.3 -+- 0.6 2.4 2.8 9 3 79.4 Tail • 15 98.6 -+- 0.6 2.4 2.4 9 3 92.3 Tarsus • 15 28.7 +- 0.2 0.7 2.4 9 1 27.2 Bill from • 15 10.0 ___O1 0.6 6.0 nostril • 2 9.8 490 Sm•E¾AND 8IBLEY, Hybridization in Towhees [ Vol.Auk 81

TABLE 4

MEASUREMENTS OF SAMPLES OF P. OCA1 FROM 12 M•LES N oF MOUNT ORIZABA AND FROM 6 M•LES E OF HUECAPA

Item Sex Numbero! Meanwith StandardCoefficient specimens standard error deviation o! variation

Weight (• 9 57.6 q. 1.6 4.8 8.3 2 1 57.1 Wing (• 9 86.0 q. 1.0 3.0 3.5 • 1 79.8 Tail (• 9 100.2 q. 1.8 5.4 5.4 • 1 92.9 Tarsus 8 9 31.3 q- 0.5 1.4 4.5 • 1 30.1 Bill from (• 9 10.5 q. 0.2 0.6 5.7 nostril • 1 10.2

Of the 40 specimenscollected here 21 were ocai with hybrid indexes ranging from 0 to 6 (average, 1.3). The 19 erythrophthalmusrange from 21 to 24 (average, 22.5). These scoressuggest that occasionalmixed pairs may occur but none was seen and no specimensranked between 6 and 21 were collected. Measurementsare given in Tables 2 and 3. This situation is rather similar to that on the SW face of Mount Orizaba (Sibley, 1950), although a larger percentageof the specimensshow evidenceof hybridizationon the NW slope (78 per cent as opposedto 16 per cent on the SW slope). Twelve miles N of Mount Orizaba, 10,000 feet; and 6 miles E of Huecapa, 10,500 feet.--These two localities are only three miles apart and are consideredtogether. The first is approximatelyeight miles north of the locality discussedjust above. Cultivated fields extend up to 10,300 feet and from here open fir forest reachesto 10,800 feet, with the final 200 to 300 feet to the top of the ridge being open pine and grass. The east side of the ridge did not seemto have any suitable habitat, even along the streambeds, when examinedthrough a binocular. No towheeswere found in the open fir woodsand they were not commonin any of the surrounding area. The 12 birds collectedwere taken from hedgerowsor abandoned fields which had becomecovered with brush 5 to 10 feet high. In one such area 7 birds were collected and several others seen. The ocai-like birds occurred around the edge of this area and the erythrophthalmus-likebirds occurred both here and in the center of the area. The secondlocality of thesetwo was only a few hundred feet below the top of the ridge. Again towheeswere uncommon,apparently due to the lack of suitable habitat. P. ocai occurred in the firs along the edges of fields and streamsand in the thicker areas of fir near the top of the ridge. P. erythrophthalmusoccurred both along the borders and in the centers 1964Oct. ] SIBLE¾^NB Smt,E¾, Hybridization in Towhees 491

TABLE 5 MEASUREMENTS OF SAMPLES OF P. ERYTHROPHTHALMUS FROM 12 MILES N OF MOUNT ORIZABA AND FROM 6 MILES E OF HUECAPA

Item Sex Numberof Meanwith StandardCoefficient specimens standard error deviation of variation

Weight c• 7 47.1 -+- 0.5 1.4 3.0 9 4 45.3 c} 1 (index 13) 50.2 Wing c• 7 85.4 -+- 0.9 2.4 2.8 • 4 80.5 c• 1 (index 13) 86.9 Tail c• 7 100.6 -+- 1.3 3.5 3.5 9 4 89.7 c} 1 (index 13) 101.6 Tarsus c• 7 28.9 ñ 0.5 1.3 4.5 • 4 28.4 c} 1 (index 13) 28.7 Bill c• 7 10.1 -+- 0.1 0.3 3.0 from • 4 10.1 nostril c• 1 (index 13) 10.0 of abandonedfields although it was rare above 10,000 feet due to the lack of suitable habitat. To the north the ridge is lower and flatter on top and as a result is extensivelycultivated. All of the trees and brush have been removedand there is very little towhee habitat. Of 22 birds taken, the 10 ocai-typespecimens range from 0 to 6 (average, 1.4) and the 11 erythrophthalmus-typespecimens range from 17 to 23 (average, 20.5). In addition there is a single intermediatespecimen with an index of 13. These specimensindicate that hybridization does occurin this area althoughthe frequencyis not high. The measurements for thesespecimens are presentedin Tables 4 and 5. Two milesE o! Cuaulo.tolapan,9,500 ]eet.--This locality lies atop a low ridge which is situated between the Cofre de Perore and the. northern extensionof the ridge from Mount Orizaba, and is approximately10 miles north of the previouslocalities. Suitablehabitat is limited to the brush and treeslining the streamsand this occursmostly above 9,300 feet. There is a gap of about three milesbetween the last suitablehabitat here and the nearestsuitable habitat on the Cofre de Perore. The three days collecting producedonly 14 birds and the sightingof 3 others. This probablyrepre- sentsa large percentageof the populationalong about three miles of the ridge. Of the 14 specimens12 are of the ocai type and have an averagehybrid index of 1.5 and a range of 0 to 4. One bird has a hybrid index of 14 and one of 12. These last two specimensare treated separatelyin the analysisof measurementsin Table 6. 492 SIBLE¾Am) SmLE¾, Hybridization* in Towhees [ Vol.Auk 81

TABLE 6 MEASUREMENTS OF THE SAMPLE OF P. OCAI AND I-IYBRIDS FRO•Vr 2 MILES m OF C UAULOTOLAPAN

Item Sex Numberof Meanwith StandardCoefficient specimens standard error deviation of variation

Weight • 8 58.1 q- 1.3 3.6 6.2 9 4 59.1 9 2 index 12 and 14 47.8 Wing • 8 84.8 q- 0.6 1.7 2.0 • 4 80.8 9 2 index 12 and 14 79.7 Tail • 6 101.3 q- 1.0 2.8 2.8 9 4 92.6 • 2 index 12 and 14 91.6 Tarsus • 8 30.4 q- 0.4 1.2 3.9 9 4 29.6 9 2 index 12 and 14 29.3 Bill from • 8 10.5 q- 0.1 0.3 2.8 nostril • 4 10.4 9 2 index 12 and 14 10.4

Cofre de Perote.--This locality is on the southwestside of the peak of the Cofre de Perote and only eight miles north of the previous locality. Pines appear at about 9,500 feet and continueto 9•800 feet. Here they begin to be replacedby fir which continuesto approximately11,000 feet, where it gives way to open pine forest. Between9,800 and 10,000 feet there are many patchesof brush and fir and thick hedgerowswhich are inhabited by towhees. Above 10,000 feet the vegetationis composedof broad belts of forest separatedby equally broad expansesof grass and lupine. These belts run parallel to the slope, following the drainage pattern. Towhees were very abundant in this habitat especially along the edgesbetween the forest and the more open areas. The 20 specimenshave an averagehybrid index of 2.2 with a range of 0 to 9. However, Figure 1 showsthat this large spreadis due mainly to one bird with a hybrid index of 9. Measurementsare given in Table 7. Previouswork (Sibleyand West, 1958) had revealedthe seriesof hybrid populationsbetween the Cofre de Peroteand Teziutlfinto the north (Fig- ures 3 and 4) and it was proposedthat the reasonfor the sharp change in hybrid index in the interveningpopulations was the large size of the terminal populations. The present series of birds, indicating a large, slightly hybrid population on the Cofre de Perote, further supportsthis hypothesis. The Teziutlfin population is also large and has an average hybrid index of 15.2. Fourteen miles NE of El Seco.--At this locality there are two volcanic coneswhich rise to a height of some9,500 feet. These were the only small peaks near the Mount Orizaba-Cofrede Perote ridge which seemedto Oct.] SmLE¾.aND SmilEY, Hybridization in Towhees 493 1964 ]

TABLE 7 MEASUREMENTS OF THE SAMPLE OF P. OCAI FROM THE COFRE DE PEROWE

Item Sex Numbero/ Meanwith Standard Coe//icient specimens standarderror deviation o/variation

Weight • 15 56.4 ñ 1.2 4.6 8.2 •2 5 53.1 ñ 2.1 4.8 9.0 Wing • 15 85.3 ñ 1.0 4.0 4.7 9 5 77.6 ñ 1.7 3.7 4.8 Tail • 15 101.1 ñ 1.2 4.5 4.5 9 5 90.6 ___2.6 5.7 6.3 Tarsus 8 15 30.6 ñ 0.3 13 3.6 9 5 29.3 ñ 0.4 0.9 3.1 Bill from • 14 10.6 ñ 0.1 0.4 3.8 nostril 9 5 10.1 ñ 0.2 0.5 4.9 havesuitable towhee habitat. Pinesstart at about 9,000 feet and continue to the top where firs can be found in the wetter and cooler areas. Only four adultsand one immaturewere collectedalthough several other adults were seen. This population is interesting in being separated from the nearestpopulation to the northwestby 17 miles and from suitable habitat on the Mount Orizaba-Cofrede Peroteridge by about 12 miles. The birds all have a hybrid index value of 23 and all showthe sameevidence of ocai influence,a small white throat patch. It would appear most likely that the populationhad originallyinvaded this peak from the northwestand not from the Mount Orizaba-Cofrede Peroteridge. It wouldbe of interest to determineif this is actually a stable isolatedpopulation as it now ap- pears from the limited number collected. It would appear that this area was invaded by a small number of birds and that there has been very little gene flow into the population. Measurementsare given in Table 8. Commentary.--The past history of the Mount Orizaba-Cofre de Perote contact seems fairly clear from the knowledge of present distribution (Figure 4). Pipilo erythrophthalmus from the north spread toward Teziutl•tn and south toward Mount Malinche. The birds which reached the Teziutlgn-Cofre de Perote regionhybridized with P. ocai and estab-

TABLE 8 MEASUREMENTS OF THE SA2VIpLEOF P. ERYTHROPHTHALMUS FROM 14 M•LES NE O•r EL SECO

Item Sex Numberof Mean Extremes specimens

Weight • 4 43.0 42.0- 43.9 Wing • 4 84.3 81.7- 85.6 Tail • 4 99.0 94.0-102.1 Tarsus • 4 27.9 27.3- 29.1 Bill from nostril • 4 10.1 9.9- 10.4 494 S•BZE¾^•) SmLE¾,Hybridization in Towhees [ Vol.Auk 81

TABLE 9 MF.•SUREIV/•ENTSOF TIlE SAZV/•PLEOF P. 0C.4l FROM CERRO YLICUYACUA

Item Sex Numberof Meanwith StandardCoefficient specimens standarderror deviation of variation

Weight • 37 61.0 ___0.6 3.4 5.6 • 15 58.2 -4- 1.1 4.1 7.1 Wing • 36 88.9 ___0.4 2.5 2.8 • 16 84.6 -4- 0.6 2.6 3.1 Tail • 31 105.8 ñ 0.8 4.3 4.1 • 8 100.1 ñ 1.4 4.0 4.0 Tarsus •' 36 31.3 q- 0.1 0.7 2.2 • 14 30.7 q- 0.2 0.9 2.9 Bill from • 38 10.8 ñ 0.1 0.5 4.6 nostril • 13 10.9 q- 0.1 0.4 3.7

lisheda seriesof hybridpopulations from the Teziutl&nregion to the Cofre de Perote. Thesepopulations then preventedfurther advancealong this route. Gene flow would be limited by the small amountsof habitat present on eachside of the Teziutl/tn area. Sinceit is unlikely that an individual bird would crosseven half the width of the hybrid zone in its dispersal from its point of origin (the fact that the hybrid zone existswould support this), the only way P. erythrophthalmuscould spreadinto this new area would be the geneticswamping out of P. ocai. But as mentioned above the seriesof hybrid populationsand their small size would make this a very slow process.Also, becauseof the large populationon the Cofre de Perote,this effect wouldnot spreadto any great extent beyond this point. The erythrophthalmuswhich spreadsouth to Mount Malinche probably crossedto the west side of Mount Orizaba and the Puerto Morelos area. These birds evidently were able to invade in large enoughnumbers to avoid frequenthybridization with ocai and the formationof hybrid populations.P. erythrophthalmuxthen spreadnorth towardthe Cofre de Perote. However, the restrictedamount of suitable habitat, which per- mits only very small populationsof either species,has produceda situation whichpermits detection of the resultsof occasionalhybridization, al- thoughthere has not been a completebreakdown anywhere along the ridge. This situationis unusualin that this is the only area known in which both pure parental types and intermediatehybrids occurat the samelocality. A similarsituation is known in the genusPasserina in the Great Plains of the United States (Sibley and Short, 1959b).

OAXACA POPULATIONS Cerro Yucuyacua.--Thispeak is located about 60 miles west of the sympatricpopulations on Cerro San Felipe. Previousattempts to reach 1964Oct.] SmLE¾^>m SmLE¾, Hybridization in Towhees 495

TABLE 10 MEASUREMENTS OF TI-IE SAMPLE OF P. ERYTHROPHTHALMUS FROM CERRO ¾UCUYACUA

Item Sex Numberof Meanwith StandardCoefficient specimens standard error deviation of variation

Weight $ 15 44.0 q- O.6 2.3 5.2 2 4 43.1 Wing $ 15 84.8 q- O.6 2.4 2.8 2 4 8o.6 Tail • 13 97.3 q- 0.9 3.3 3.4 9 4 94.4 Tarsus 6' 15 28.3 q- 0.2 0.9 3.2 2 4 28.0 Bill from • 15 10.0 q- 0.2 0.7 7.0 nostril 2 4 10.0 this mountain,in 1954, were not successful,although it was established that P. erythrophthalmuswas very commonin most of the surrounding country. In 1958, pack animalswere hired and it was possibleto establish a camp at 10,800 feet, about 300 feet below the summit. P. erythrophthalmusoccurs along all the streams,around the corn fields, and in any patchesof pine or brushfrom 7,000 feet up. However,as the streamsbecome narrower the amount of habitat decreasesgreatly until at 9,000 feet the habitat occursonly in small patches. At 9,300 feet two pairs were seenand three of thesebirds were collected. One of thesepairs was in the samelarge patch of brush occupiedby a pair of P. ocai with youngout of the nest. From here to 10,000 feet habitat is limited for both species.Above 10,000 feet only one place was found where P. erythrophthalmusoccurred. This wasin the brushyedges around a seriesof fields at 10,500 feet. Six adults were collectedat this locality, which was sur- roundedon all sidesby suitable P. ocai habitat. P. ocai was found in only one place below approximately 10,000 feet. Above this altitude the slope became more gradual and the amount of brush increasedgreatly. The forest cover on this part of the mountain has been heavily cut over and densebrush has grown up almost every- where. The cool,wet climate has produceda much more luxuriant growth than that found at lower elevations. As a result P. ocai occurred almost everywherein great abundanceand the 54 specimenscollected represented only a fractionof the birds in the small area coveredby our hunting. The 54 P. ocaihave an averagehybrid indexof 0 and the 19 P. erythro- phthalmushave an averagehybrid index of 23.6 with a range of 22-24. These are the same values found on Cerro San Felipe, which presentsa situation similar to the presentone. Measurementsare given in Tables 9 and 10. 496 S•BL•¾^•D SmL•¾, Hybridization' in Towhees [ Vol.Auk 81

Commentary.--It seemslikely that the ocai characterswhich are found in the P. erythrophthalmuspopulations in Oaxacaare the result of intro- gressionfrom the north and not from local hybridization. The Rio Santo Domingohas been proposed as a barrierbetween the towheesin the Mount Orizaba-Zoquitl/•narea and thoseon Cerro San Felipe (Sibley, 1954). This barrier probably is highly effectivein relation to P. o.caibut does not prevent the dispersalof P. erythrophthalmusbecause of its tolerance of warmer and drier areas. As a result there would be introgressionof ocaigenes from the northinto. the erythrophthalmuspopulations in Oaxaca but no tracesof erythrophthalmusgenes in the Oaxacaocai populations. It has previouslybeen suggestedthat thesetraces of o.caigenes might be similarto thosefound in towheepopulations in the United Statesand are merely "ancestral"genes. However,the higher percentageof ocai char- actersand the geographiclocation of the Oaxacapopulations would seem to favor the introgressionhypothesis. It thus appearsthat the populationsof P. ocai in Oaxaca,and probably also in Guerrero,have been able to maintain reproductiveisolation from P. erythrophthalmuseven in areas in which the two occur in sympatry. In the more than 120 specimensof P. ocai from Oaxacawhich have been examined there is no evidenceof hybridization. If there had been any previoushybridization, or any recentexchange of geneswith the populations of ocai to the north of the Rio Santo Domingo, one would expect to find tracesof it in the Oaxacaocai populationstoday. At present P. erythrophthalmusmay be spreadingtoward Guerrero, althoughits westernlimit is still not known. It appearslikely that there are other peaks near Cerro Yucuyacuawhere the two speciesmust come into contact. It also appearsunlikely that collectingin theseareas will reveal any hybridization or that future collecting on Cerro Yucuyacua will reveal hybridization.

MAN•S INFLUENCE The effectsof agriculture,deforestation, afforestation, irrigation, and other humanactivities upon native animalsare often profound. Because speciationcan proceedonly whenextrinsic barriers prevent gene flow between two divergingconspecific populations it is not surprisingto find evidence that human activities sometimesdestroy such barriers before speciationis completeand permit secondarycontacts between divergent, but still interfertilepopulations. In Colombia,for example,the destruc- tion of virgin rain forest and the establishmentof extensive coffee and bananaplantations has permitteda secondarycontact between two brush- dwellingforms of the tanagergenus Ramphocelus. The resultinghybridi- zation has produceda cline in color betweentwo strikingly different- 1964Oct. ] SIBI•EYAND SIBLEY, Hybridization in Towhees 497 lookingpopulations (Sibley, 1958). Similarly,across the Great Plains of North America the planting of trees has provided an avenue for increased geneexchange between eastern and westernpopulations of severalwood- land birds. Hybridization between the eastern and western forms of the buntings,Passerina (Sibley and Short, 1959b); the orioles,Icterus (Sibley and Short, 1964); the flickers, Co•aptes(Short, in press); and the grosbeaks,Pheucticus (West, 1962) has resulted. Additional exampleshave been noted by Sibley (1961). In most of thesecases the human activities responsiblefor the break- downof the barrier have beenquite recent,usually within the last century. In Mexico, however,it is highly probablethat human activities modified the ecologicalbarriers betweenPipilo ocai and Pipilo erythrophthalmu• at least 300 and possiblyas much as 3,000 years ago. There is evidence of continuousoccupancy of the Valley of Mexico dating back to ancient times. De Terra et al. (1949) believethe TenexpanMan belongedto a hunting culture which existedabout 10,000 to 9000 B.C. After this there was a seriesof hunting and fruit-gatheringcultures which they call the Chalco-CultureComplex. The archaicculture which followedthis, some- times calledthe Middle Culture, was a well developedagricultural culture (Vaillant, 1941). This is generallyplaced by most Mexican archeologists from 2000 or 1500 B.C. to 500 or 400 B.C. (Sears,1952). Therefore, by 1500 B.C. there was a culture which probably clearedland by burning, as in most primitive societies,and provided the disturbed habitat which would increasethe likelihood of hybridization. By 900 A.D., Vaillant (1941) believes,the TeotihuacanCulture which was by then inhabitingthe area was forcedto move becauseof crop failure and drying up of the streamsdue to deforestation. The tremendousquan- tities of lime cement used in building their temples and cities and the wasteful methodsof obtaining charcoal to reduce the limestone were the major reasonsfor this deforestation. Cook (1949a) reported evidence of heavy erosiondating back to this time. The sameauthor (Cook, 1949b) estimatedthe populationof the Teotlalpan (roughly the northern part of the State of Mexico north of the latitude of Pachuca, an area of 2,300 squarekm) in 1000 A.D. as being 200,000 comparedwith a 1940 population of 110,000. Thus in the period around 1000 A.D. there is evidenceof extensivealteration of the forest cover and, in some areas, of a higher population than at present. Cook estimatesthat by 1520 the population of Teotlalpan was 500,000 after a period of reducedpopulation. Cook and Simpson (1948) estimatedthat the population of Central Mexico was 11,000,000in 1520 droppingto 6,400,000 in 1540 and to 2,500,000 in 1600. Accordingto an official censusin 1793 the populationwas approximately 3,700,000. From these figuresit becomesevident that there must 498 S•BLE'ZA•r•) S•rBLE'Z, Hybridization in Towhees [ Vol.Auk 81

Statute Miles

No. of specimens Oct.] S•BL•¾Am) SmLE¾,Hybridization in Towhees 499 1964 / have beentremendous changes, in the periodfrom 1000 to 1500 A.D., in the forestcover. Thesechanges were probablyas extensiveas thosepro- ducedby the Spanishin their introductionof livestock(Simpson, 1952) and in their miningactivities (West, 1949). Cook (1949a) says:"The important cycleof erosionand depositiontherefore accompanied intensive land useby hugeprimitive populations in centralMexico, and had gone far toward the devastationof the countrybefore the white man arrived. Spanishagriculture, deforestation, and livestockgrazing merely finished what the red man had alreadynearly carriedto completion."In connection with the erosionalevidence he states: "If we wish to give credenceto the testimonyof the land itselfwe mustconcede the passageof 3,000 to 5,000 yearssince the first establishmentof the corn-cultivating,pottery- making cultureson the plateau." Fromthis evidence it may beassumed that the extensiveareas of second- growthbrush and resultinginterdigitation of ocai and erythrophthalmus habitatsthat seemimportant in increasingthe probabilityof hybridization have existedfor at least a thousandyears and possiblylonger.

DISCUSSION This studyhas clarified several aspects of the distributionand variation of the red-eyedtowhees in easternand southeasternMexico. Of particular interestis the remarkablesituation existingin the region that includes Mount Malinche, Mount Orizaba, the Cofre de Perote, and Teziutl/•n. The new material collectedin 1958 now makes it possibleto present a fairly clearpicture of the effectsof hybridizationand the pathwaysof genemigration in this areawhich was so puzzling in the past (Sibleyand West, 1958: 101). Referenceto Figure 5, as well as to Figures1, 3, 4, and 7 shouldmake the followingdiscussion clear. The largepopulations of Pipilo erythrophthalmusin the SierraMadre Oriental(hybrid index24.0) are in contactwith the Pachucapopulation (22.4) which showsthe effects of introgressionvia the Tlaxco (22.0) area. There seems to be little or no direct contact between the Pachuca area and the populationon mountsPopocatepetl and Ixtaccihuatl(19.8) exceptvia the Mount Malinchearea. Thus the Mount Malinche-Tlaxco area (22.0) receivesa strongerythrophthalmus influx from the north and also someocai genesfrom the west via the long trans-plateaugradient (Sibley,1950: 169-176). In addition,erythrophthalmus genes enter the Malinchepopulation via E1 Seco(22.5) and ocaigenes via the Teziutl•n (15.2) population.Thus the Malinche-Tlaxcopopulation stands at a crossroadsin the networkof genemigration pathways, receiving and ex- tendingeffects in four directions. 500 SIBLEYAND SIBLEY, Hybridization inTowhees [ Vol.Auk 81

Before examiningthe Orizaba-Perotearea in detail it is important to considerthe fact that erythrophthalmusgenes also come into this region from the south. The essentially"pure" populationon Cerro San Felipe is apparentlyin limited contact,via dispersingjuveniles, with the Zoquitlgn area. As notedpreviously, both the characterevidence and the topography indicatethat juvenile erythrophthalmuscan and do occasionallycross the barrier of the Rio SantoDomingo lowlands. It is apparentlythis exchange route that accountsfor the occurrenceof chestnutcrown streaksin a percentageof the erythrophthalmusin Oaxaca. This gives them an average hybrid index value of 23.6 although the sympatric ocai show no evidence of hybridization. This populationhas been considered"pure" (24.0) becausethe ocai charactersdid not result from sympatrichybridization. The introgressionof erythrophthalmusgenes from Oaxaca into the Zoquitl/tn-PuertoMorelos (23.0) area is apparently effective in shifting the erythrophthalmusof the west side of Mount Orizaba to 23.4 in spite of the 22.5 influx from E1 Secoand the still lower values (20.4 and 22.2) producedby the limited hybridization and introgressionwith ocai in the area from Mount Orizaba to the Cofre de Perote. The seriesof populationsfrom Tacop/tn (15.5) and Zaquiolca (13.0) and on south (11.6, 7.3) to the Cofre de Perote is of specialinterest because these populations form the bridge between the strongly erythro- phthalmusTlaxco (22.0) area and the nearly pure ocai populationsfrom the Cofre de Perote (2.2) on south along the ridge (1.5, 1.4, 1.2) to Mount Orizaba (.43), Puerto Morelos,and Zoquitl•n. In addition, along the ridge betweenMount Orizaba and the Cofre de Perote, erythrophthalmus and ocai populations,both contaminatedby introgression,are sympatric and occasionallyhybridize. These facts lead to the most puzzling question: why do not the sym- pattic populationsalong the Orizaba-Perote ridge hybridize much more freely and form a singlehybrid population,as in the Teziutl/tn area? This is the questionthat stimulatedthe 1958 field work (Sibley and West, 1958: 101). The 1958 field study showedthat the populationsof both speciesalong the Perote-Orizaba ridge were sparseand that suitable habitat, for both species,has been severelyreduced by human activities. There is little or no habitat intermediate between the types optimal for the two species, thus there is no "hybrid habitat," in Anderson's(1948) sense,that would providea favorableenvironment for a hybrid populationas, for example, the Teziutlgn area affords. The lack of an intermediate habitat will, theoretically,select against hybrids and could be an important factor in reducingthe incidenceof sympatric hybridization becausesuch selection will act to reinforce interspecific isolating mechanisms(Sibley, 1957, 1964Oct. ] SmLE¾^ND SmLE¾, Hybridization in Towhees 501

1961). This explanationis attractive but only speculative.To determine whetheror not it is correctwill requireeven more detailedfield studies. The large terminalgene pools at eachend of the Cofre de Perote-Mount Orizaba ridge must also have profound effects on the small intervening populations.The large erythrophthalmus(23.4) populationon Mount Orizaba easily swampsany ocai genesreceived by introgressionvia the Perote-Orizabaridge. Similarly the ocai populationson the ridge and the Cofre de Perote absorb the trickle of erythrophthalmusgenes from the south. The tenuouscontact between the north end of the erythrophthalmus seriesand the ocai on the Confre de Perote is a partial barrier, erected by humanagriculture, that further reducesthe opportunityfor introgression. Nevertheless,there is some gene exchangebetween the Cofre de Perote population and the erythrophthalmuspopulations of the Orizaba ridge as shownby the fact that the hybrid indexesof thesepopulations form a gradientpattern from 20.4 at 12 miles north of Mount Orizaba to 22.2 at Tlachichucaand 23.4 on the west slopeof Mount Orizaba. Some ocaigenes are alsocontributed by sympatrichybridization but the gradient clearly indicatesintrogression between the Cofre de Perote ocai (2.2) and the erythrophthalmusof the Orizaba area. Becausethe populationsof erythrophthalmusalong the Perote-Orizaba ridge are so sparsethe ocai effect is quickly swampedin the large erythrophthalmusgene pool of the Mount Orizaba-PuertoMorelos area. Figure 5 summarizesthis situation. The gradientin hybrid indexesof the seriesof ocai populationsbetween the Cofre de Perote and Mount Orizaba also testifies to the fact that introgressionin this area is more effective than sympatrichybridization in determiningthe characteristicsof the populations. In spite of the occasionalsympatric hybrids the gradientprogresses smoothly from 2.2 on the Cofre de Perote to 1.5 eight miles to the south near Cuaulotolapan. Twelve miles farther south the index is 1.4 and it dropsto 1.2 in another eight or nine miles (at 10 miles ESE of Tlachichuca). On Mount Orizaba the average index is .43. Thus, in a distance of about 32 miles, the average index changes1.77 hybrid index units or .05 units per mile. The slope of this gradient (see Figure 1) is characteristicfor the sectionof an intro- gressivecline near oneof the terminalspecies. Several examples are known in Jaliscoand Michoacgn(Sibley, 1954). The introgressivegradient from the Mount Malinche-Tlaxco area through the Teziutlgn region to the Cofre de Perote, Mount Orizaba, Puerto Morelos, and Zoquitlgn is also of interest for taxonomicreasons. This series of populationsgradually bridges the morphologicalgap betweenthe pure erythrophthalmuspopulations of the Sierra Madre Oriental and the virtually pure ocai at Puerto Morelos and Zoquitlgn. In this re- spectit presentsexactly the sameproblem as found in the trans-plateau 502 SmL•¾AnD SmL•¾, Hybridization in Towhees [ Vol.Auk 81 gradient (Sibley, 1950) and in the several erythrophthalmus-ocaigradi- entsin Jaliscoand Michoac/tn(Sibley, 1954). The similarityto the cline from Mesa de los Puercos(23.5) to Cerro Gordo (22.8), Cerro Grande (22.6), CerroViejo (13.7), CerroGarcia (8.0), Mazamitla (2.6), and the mountainsof southwesternJalisco (.17) is striking. The Teziutl•n populations (average15.2) providean interestingparallel to the Cerro Viejo (13.7) population. Both are at the centerof the gradientand both show the expandedvariability expectedfrom the recombinationof genesresult- ing from a fertile interspecificcross. The Cerro Viejo sampleof 76 specimensranges from 6 to 20, covering15 index units. The Teziutl•n sample of 55 specimensranges from 9 to 21, covering13 indexunits. The populationsin the centerof the trans-plateaugradient (P/ttzcuaro, 7.8, and Mil Cumbres,13.5) exhibitthe samevariational pattern (Sibley, 1950). The taxonomicdifficulty presentedby this type of situation was first encounteredduring the study of the trans-plateauseries. Becauseof the sympatricpopulations on Cerro San Felipe and Mount Orizaba it was de- cided to maintain ocai and erythrophthalmusas separate speciesand to view the long introgressivegradients as elongatehybrid zones. For purely nomenclatural purposes the division between ocai and erythrophthalmus in the trans-plateau gradient was placed at 10 miles east of Morelia, Michoac,•n. In the presentsituation this divisionpoint would fall between the area 12 mi. north of the town of Perote and Teziutl/tn, judging from the average hybrid indexesfor the samplesfrom 12 mi. north of Perote and from Zaquiolca (----3 mi. W Teziutl•n). South of this point the populations shouldbe referred to ocai, north of it to erythrophthalmus.How- ever: "This must be reco•o•izedas a practical solution to the incompati- bility of Linnaeannomenclature in this specialsituation and shouldnot be regardedas a synopsisof the biologicalsituation" (Sibley, 1950: 176). The 1958 field work has provided answersto severalof the questions posedby Sibleyand West (1958) but, as always,additional questions are exposed.It would be instructiveto know more about the actual numbers of birds in each of the populationsin the easternplateau area and to determine, in considerabledetail, their exact limits of distribution. Thus censuses,transect studies, and a detailed ecologicalsurvey of the entire area would be appropriateas a next step. Collectingshould be continued to fill in the gaps in the known gradients and to locate additional avenues of genemigration. The situationat Zoquitlfi.nand the.precise nature of the Rio SantoDomingo lowlands need to be explored. In additionit would be useful to know more about the interspecificinteractions that must occurwhere the two speciesare sympatric.Observations of their behavior, Oct.] Sm•.E¾A•D SmLE¾,Hybridization in Towhees 503 1964 .I especiallywith regard to the nature of interspecificisolating mechanisms, competition, etc., should be instructive.

SUMMARY

Studyof a collectionof 215 adult specimensof Pipilo erythrophthalmus and Pipilo ocai obtainedin 1958 in the southernplateau regionof Mexico showsthe following: 1. The Mount Malinchepopulation is in contactwith the populations on the west slopeof Mount Orizaba via the E1 Secoarea. 2. The complexsituation betweenMount Orizaba and the Cofre de Perote (Sibley and West, 1958) is clarified. It seemsto involve: (a) sparse,sympatric populations of the two species,each nearly pure but both tainted by introgression;(b) occasionalsympatric hybrids, which (c) may be selectedagainst because of the lack of intermediatehabitat; (d) large terminal gene pools that quickly swampthe effects of introgression. 3. The populationof P. ocai on Cerro Yucuyacuain Oaxacais sympatric with P. erythrophthalmusbut no evidenceof hybridization is evident. Man's role in providingthe ecologicalconditions favorable for hybridizationbetween the two speciesof towheesis reviewed.Evidence that such conditionsmay have beenproduced more than a thousandyears ago is discussed.

LITERATURE CITED

A•DERSO•,E. 1948. Hybridization of the habitat. Evolution, 2: 1-9. A•DERSO•,E. 1949. Introgressivehybridization. New York, John Wiley and Sons. Coo•:, S.F. 1949a. Soil erosionand populationin central Mexico. Ibero-Americana. Vol. 34. Berkeley, California, University of California. Coo•:, S.F. 1949b. The historicaldemography and ecologyof the Teotlalpan. Ibid., Vol. 33. Coo•:,S. F., A•D L. B. Sr•so•. 1948. The populationof Mexicoin the sixteenth century. Ibid., Vol. 31. DE TERR^,H., J. ROMERO,AND T. D. STEWART.1949. TenexpanMan. Viking Fund Publ. Anthr., no. 11. SEARS,P. B. 1952. Palynologyin southernNorth America. I. Archeologicalhori- zons in the basinsof Mexico. Bull. Geol. Soc. Amer., 63: 241-254. SaoRT, L. L., JR. (In press) Hybridization in the flickers (Colapres) of North America. Sm•;E¾,C.G. 1950. Speciesformation in the Red-eyedTowhees of Mexico. Univ. California Publ. Zo51., 50: 109-194. Sm•;E¾,C.G. 1954. Hybridization in the Red-eyedTowhees of Mexico. Evolution, 8: 252-290. 504 SIBLE¾AlffD SiBLE¾, Hybridization in Towhees [ Vol.Auk 81

SXBLE¾,C. G. 1957. The evolutionary and taxonomic significanceof sexual dimor- phism and hybridizationin birds. Condor, 59: 166-191. SmLE¾,C. G. 1958. Hybridization in some Colombian tanagers,avian genusRam- phocelus. Proc. Amer. Phil. Soc., 102: 448-453. SIBrE¾,C. G. 1959. Hybridization in birds: taxonomic and evolutionary implica- tions. Bull. Brit. Orn. Club, 79: 154-158. SIBrE¾,C. G. 1961. Hybridization and isolating mechanisms.Pp. 69-88 in Verte- brate speciation (W. F. Blair, ed.). Austin, University of Texas Press. St•LE¾, C. G., AX• L. L. SX•ORT,JR. 1959a. Hybridization in some Indian bulbuls PycnonotuscaJer X P. leucogenys.Ibis, 101: 177-182. Si•E¾, C. G., AXDL. L. SX•ORT,JR. 1959b. Hybridization in the buntings (Passerina) of the Great Plains. Auk, 76: 443-463. SIB•E¾,C. G., AXDL. L. SX•ORT,JR. 1964. Hybridization in the orioles (Icterus) of the Great Plains. Condor, 66: 130-150. SIBLE¾,C. G., AXD D. A. WEST. 1958. Hybridization in the Red-eyed Towhees of Mexico: the easternplateau populations. Condor, 60: 85-104. SIBLE¾,C. G., AND D. A. WEST. 1959. Hybridization in the Rufous-sided Towhees of the Great Plains. Auk, 76: 326-338. Si•esoN, L. B. 1952. Exploitation of land in central Mexico in the sixteenth century. Ibero-Americana. Vol. 36. Berkeley, California, Univerisity of California. VAIrLANT, G. C. 1941. Aztecs of Mexico: origin, rise and fall of the Aztec nation. Garden City, New York, Doubleday, Doran. WEST,D.A. 1962. Hybridization in the grosbeaks(Pheucticus) of the Great Plains. Auk, 79: 399-424. WEST, R. C. 1949. The mining community in northern New Spain: the Parral mining district. Ibero-Americana. Vol. 30. Berkeley, California, University of California.

Department of Conservation,Cornell University, Ithaca, New York, and Pacific Project, U.S. National Museum, Washington,D. C.