A Hybrid Red Crossbill-Pine Siskin (Loxia Curvirostra X Carduelis Pinus

Home , Carduelis, Crossbill, Pine siskin, Red crossbill, Tibetan serin

January1984] ShortCommunications 155

HILLS, M. 1978. On ratios--a response to Atchley, nov, Cramer-Von Mises and related statistics Gaskins and Anderson. Syst. Zool. 27: 61-62. without extensive tables. J. Amer. Stat. Assoc. 69: SAS INSTITUTE.1982. SAS user's guide: basics.Cary, 730. North Carolina, SAS Institute, Inc. ZAR, J. H. 1974. Biostatisticalanalysis. Englewood SHAPIRO,S.S., & M. B. WILK. 1965. An analysis of Cliffs, New Jersey,Prentice-Hall, Inc. variance test for normality (complete samples). Biometrika 52:591-611. Received3 March 1983,accepted 6 September1983. STEPHENS,M.A. 1974. Use of the Kolmogorov-Smir-

A Hybrid Red Crossbill-Pine Siskin (Loxia curvirostra x Carduelis pin us) and Speculations on the Evolution of Loxia

DAN A. TALLMAN • AND RICHARD L. ZUSI 2 'Departmentof Mathematics,Natural Sciences and Health Professions, Northern State College, Aberdeen,South Dakota 57401 USA; and 2National Museum of NaturalHistory, SmithsonianInstitution, Washington, D.C. 20560 USA

On the morning of 27 December1981, a strange streaksweakest on lower throat and belly and dark- finch appeared at Tallman's feeder in a residential est and best defined on flanks and crissum. backyardin Aberdeen,Brown County, South Dakota. Upperparts dusky olive streaked or spotted with Alone and in the companyof Pine Siskins,the bird dark gray. Feathersof forehead and crown dark with consumedsunflower seeds.It fed on the ground and whitish or yellowish edges, giving spotted effect. alsocracked seeds while perchedon a sunflowerhead Longer feathers of nape, neck, and back dark gray hung from a clothesline.Tallman noted that this finch, borderedwith dusky olive laterally, giving streaked when approached,did not fly with a small siskin effect. Rump contrastswith back in having feathers flock but flew alone, silently, in a straightline from with paler olive centersand yellowish lateral edges. the feeder. The bird exhibited aggressiontoward sis- Upper tail coverts uniform grayish olive without kins feeding near it. The finch was obviously not a pronounced center streaks, darker than rump but regular North American species,being most like a paler than back.Superciliary line whitish with dusky siskin but larger, less boldly streaked, and with a streaks.Auricular patch dusky grayish olive. proportionately larger bill and head. Immediately Wings sooty; median and greater wing coverts with upon recognizing the bird as an oddity, Tallman pale, olive-whitish tips (and faint brownish cast)pro- opened Potter traps and mist nets and eventually ducing two narrow wing bars--the anterior one caughtit in a Potter trap. The bird was then prepared somewhat broken and the posterior one continuous. as a museum skin and sent to the National Museum Narrow yellow edgingson all primaries, exceptout- of Natural History, SmithsonianInstitution, where er, extending distally to point of emargination.Sec- Zusi identified it as a hybrid between the Red Cross- ondaries with short yellowish outer border toward bill (Loxiacurvirostra) and Pine Siskin (Carduelispi- tip (tips bilobed), becoming broader and whiter on nus).Apparently no previous reportsof a hybrid be- inner, superficialsecondaries. Inner vanesof prima- tween Loxiaand any other genusexist (Cockrum 1952, ries and secondarieswith broad silvery yellowish in- Gray 1958). ner border that narrows distally and ends proximal The absenceof damageto the bill, feet,or plumage to tip (at point of emargination in primaries). suggeststhat the bird had not been caged.We judged Tail strongly notched; sooty with yellow outer the bird--a female--to be at leasta year old from the edges. Edges least pronounced on outer and most completely ossified skull and the shape of its rec- pronouncedon inner feathersand broadestbasally. trices. The latter are more rounded at the tip and Inner vanes edged narrowly with whitish. have broader inner vanes than those of juvenile Label data include the following: bill dark slate crossbills and siskins. We therefore compared it gray, iris dark brown, legsand feet dark brown, gape mainly with adult females taken in December and whitish, skull ossified, little to moderate fat, no molt, January. ovary tiny. The specimen(USNM 582513)may be describedas Measurements of wing, tail, tarsometatarsus,and follows (seeFig. 1). Underpartswhitish, streakedwith bill are given in Table 1. They are compared with gray.Sides and flanksshading to grayisholive. Patchy thoseof 10 specimensof Carduelispinus pinus and 10 yellowish wash on lower throat, breast,belly, and of Loxiacurvirostra sitkensis--the smallest recognized crissum. Streaks neither bold nor sharply defined, North American race of the species.If wing length formed from longitudinal center stripes on feathers; is taken as an index of body size, the hybrid is closer 156 ShortCommunications [Auk,Vol. 101

Fig.1. Photographsof a hybridLoxia curoirostra x Carduelis pinus and the parental species. From left to rightin eachphotograph: Carduelis p.pinus, female, USNM 442040; Loxia curvirostra x Carduelis pinus, female, USNM 582513;Loxia curoirostra sitkensis, female, USNM 269830.

in size to the smallest crossbill than to the siskin. edgeof the uppermandible is smoothlyconcave from Thetail of thehybrid is longer in proportionto wing baseto tip; the left edgeis somewhatangled near the length than that of the crossbilland similar to that middle of the bill. In the lower mandible the left side of the siskin.Tarsal length of the hybridand of the is smoothlyconcave and the right sideangled. The crossbillis relatively shorter than that of the siskin. tip of the uppermandible is deflectedslightly to the The lengthof the bill of the hybridis greaterthan right and the tip of the lower mandibleslightly to that of siskinsand is alsogreater in proportionto the left. These features are similar to those seen in wing length (13.6%in hybrid, 12.3%in siskin).Com- exaggeratedform in right-crossedcrossbills. Other parable measurements of the bill in Loxia could not crossbillfeatures are the dark auricularpatch, yel- be made. lowish rump, and spottedpattern on the forehead Several features indicate that Loxia is one of the and crown. parentsof the hybrid. On casualinspection, the bill The White-wingedCrossbill (Loxia leucoptera) is showsnone of the asymmetryof a crossbilland re- unlikely as a parent,because the wing barsof the semblesthe conicalbill of a siskin.Under magnifi- hybrid are narrower, not broader, than those of a cationof a dissectingmicroscope, however, the right siskin,and the uppertail covertsare olive ratherthan and left tomial edgesof each mandibleare different- blackish. ly shapedin dorsalor ventralview (Fig. 2). The right Charactersof the hybridthat indicateCarduelis pi-

TABLE1. Measurements(mm)of adult female Carduelis p.pinus, Loxia curvirostra sitkensis, andthe hybrid Loxia curvirostrax Carduelispinus.

Carduelisp. pinus Loxiac. sitkensis (n = 10) (n = 10) Measurements Mean SE Hybrid Mean SE Wing length 70.8 0.42 78.5 81.4 0.43 Tail length Inner rectrix 32.4 0.43 36.5 37.2 0.57 Outer rectrix 39.4 0.37 43.3 42.6 0.42 Tarsometatarsallength 14.2 0.05 15.1 15.5 0.19 Bill length from nostril 8.7 0.12 10.7 -- -- January1984] ShortCommunications 157

are noted among carduelinesfor their habit of feeding in a variety of positions,often head down or back down, and for holding conesor other food sources with one foot while extracting seeds with the bill (Newton 1967, Palmer 1968). The bills of Carduelis pinus,C. spinus(Siskin), C. carduelis(European Gold- finch), Loxia curvirostra,and L. leucopteraare laterally compressedfor much of their length and are provid- ed with fine, sharp points on both mandibles. Their bills differ from the thicker bills of most other species of the Carduelinae. The tweezer-like bills of the three speciesof Carduelismentioned above are used in part to extract seeds from cones and cone-like structures, in which scalesmay be forced apart by gaping. The depressor mandibulae muscle and the retroarticular processof the lower jaw are enlarged in C. carduelis, C. spinus,and C. pinusas part of an adaptive complex for gaping, but this is not true of other speciesof Carduelis,in which gaping does not occur (Newton a b 1967). These anatomical features are also found in Fig. 2. Outline drawingsof mandiblesof a hybrid Loxia,where they serve in a variety of forceful move- Loxia curvirostrax Carduelispinus (USNM 582513). a, ments of the lower jaw. Thus, these three speciesof upper mandible in dorsal view; b, lower mandible Carduelisappear to be preadapted for opening cone in dorsal view. Dotted median lines added to em- scalesby methods other than gaping and for opening phasize asymmetry. tougher and larger cones. In an ancestral form, lateral prying, as suggestedfor the hybrid, might favor the evolution of asymmetrical mandibles and jaw musclesand the further specializations of structure nus to be one of the parents are the extensively found in crossbills.An increase in bill and body size streaked,whitish underparts,the narrow (and faintly would probably have accompanieda shift to feeding brownish) wingbars, yellow outer edges to the bases on larger cones, as it has apparently in Palearctic of the primaries and toward the tips of the second- speciesand races of Loxia (Lack 1944a, b). The North aries, the broad whitish outer edgings on the inner- American races of Loxia curvirostra show more of a most secondaries,and the yellowish inner vanes of difference in body size than existsbetween the Pine the primaries and secondaries. Siskin and the smallest of the races of Loxia curvirostra The pattern of wear on the bill tips of the hybrid (compare Table 1 with measurements in Griscom is intriguing. Under magnification,the tips of both 1937), and a correlation with cone size is probable, mandibles were noted to be smoothly rounded by although it has not been thoroughly documented. wear, and the gonys and left surfacenear the tip of The feeding adaptationstypified in Carduelispinus the lower mandible showed wear in the form of a may have evolved more than once within the car- slight depression with irregular longitudinal stria- duelines, but among living speciesthey are known tions. This wear suggeststhat the bird had used its to occur only in the three speciesof Carduelismen- bill in a manner for which its structure was not whol- tioned above.These few speciesmay serveas a model ly suited, such as separating cone scalesby inserting for an intermediate stage in the evolution of Loxia, the laterally compressedbill between them, with the and they may indeed be the closestliving relatives left side of the bill toward the median axis of the of the crossbills.This suggestion is consistent with cone, and then prying to the left with the partially Raikow's (1978) placement of Loxia and Carduelison opened lower mandible, or by using the tip of the a separatebranch of a cladogram based largely on lower mandible as a fulcrum against the inner cone appendicular myology. The Tibetan Serin (Serinus scaleand prying open the outer scalewith the upper thibetanus)has a slender bill and is a cone feeder, but mandible by rotating the head. The tips of both man- the extent of its adaptation for gaping is unknown. dibles may have been worn by pounding on tough This speciesis widely regarded as a member of Car cones. Red Crossbillsare known to open pine cones duelis,but it was transferred to Serinusby Wolters by prying laterally with the opened lower mandible (1967) on the basisof its plumage pattern. (Tordoff 1954). We thank Allan R. Phillips for his comments on Certain speciesof the genus Carduelisexhibit fea- the hybrid specimen and Richard C. Banks for his tures and behavior that might have been present in constructivecriticism of the manuscript. The photo- the ancestor of Loxia. Members of Carduelis and Loxia graphs were prepared by Victor Krantz. 158 ShortCommunications [Auk,Vol. 101

LITERATURE CITED PALMER,R. S. 1968. Spinuspinus (Wilson), Pine Sis- kin. Pp. 424-447 in Life historiesof North Amer- COCKRUM,E.C. 1952. A check-list and bibliography ican cardinals, grosbeaks, buntings, towhees, of hybrid birds in North Americanorth of Mex- finches,sparrows, and allies. U.S. Natl. Mus. Bull. ico. Wilson Bull. 64: 140-159. 237, part 1. GRAY,A. P. 1958. Bird hybrids. A check-list with RAIKOW,R.J. 1978. Appendicularmyology and re- bibliography.Farnham Royal,Buckinghamshire, lationshipsof the New World nine-primaried England, Commonwealth Agricult. Bur. oscines(Aves: Passeriformes). Carnegie Mus. Nat. GRISCOM,L. 1937. A monographicstudy of the Red Hist. 7. Crossbill. Proc. Boston Soc. Nat. Hist. 41(5): 77- TORDOFF,H. B. 1954. Social organization and be- 210. havior in a flock of captive, nonbreeding Red LACK,D. 1944a. Ecologicalaspects of species-for- Crossbills. Condor 56: 346-358. mation in passerinebirds. Ibis 86: 260-286. WOLTERS,H. E. 1967. •ber einigeasiatische Car- 1944b. Correlation between beak and food duelinae. Bonn zool. Beitr. 18: 169-172. in the crossbill, Loxia curvirostra Linnaeus. Ibis 86: 552-553. Received9 May 1983, accepted30 August1983. NEWTON,I. 1967. The adaptive radiation and feeding ecologyof someBritish finches.Ibis 109:33- 98.

The Effect of Burrow Loss on Mate Choice in the Leach's Storm-Petrel

DOUGLASS H. MORSE • AND STEPHEN W. KRESS2 •Divisionof Biologyand Medicine, Brown University, Providence, Rhode Island 02912 USA;and 2NationalAudubon Society/Cornell University Laboratory of Ornithology,159 Sapsucker WoodsRoad, Ithaca, New York 14850 USA

Long-lived seabirdsoften nest for many yearswith metal screening (1- x 1-cm mesh) was placed over the same mate at the same site (Fisher and Lockley each shingle and securedwith 15-cm wire spikesin 1954, Lack 1968). Repeated pairings might result each corner to prevent accessto these burrows. The either from strong mate fidelity or from a tendency remaining burrows were not covered. In June 1981 to return to the same burrow or nest site, here called the 37 burrows not covered in April were censused site tenacity (Hild•n 1965, Morse 1980). Retentionof as in 1980, and inhabitantsof neighboringburrows a site sometimesexceeds retention of a mate (e.g. were capturedin order to locateas many of the birds Allan 1962, Morse and Buchheister 1979), which ten- displacedfrom the experimentalburrows as possible. tatively suggeststhat site tenacity is primarily re- In all, over 500 unnumbered burrows, those closest sponsiblefor repeated pairings in these instances.If to the 62 numbered burrows, were searched for dis- repeated pairing is a consequenceof site tenacity, placed birds. birds deprived of their nest sitesshould remate ran- Mate retention was strongly site-dependent.A sig- domly; if repeatedpairing resultsfrom mate fidelity, nificantly higher proportion of undisturbed individ- birds deprived of their nest sites should retain their uals retained their mates from 1980 than did dis- previous mates. placed individuals (Table 1) (P < 0.001 in a one-tailed We know of no controlled experiments that assess Fisher Exact Probability Test). the roles of mate fidelity and site tenacity in retain- Two groupsof birds could not be usedin this anal- ing matesfrom one breedingseason to the next. Here, ysis (and thus are not included in the two right col- we demonstrateexperimentally that retentionof the umns of Table 1), but nevertheless are consistent with samemate during successivebreeding seasonsby the this conclusion: (1) both members of six undisturbed Leach'sStorm-Petrel (Oceanodromaleucorhoa) is large- pairs changed their burrows, and all but one of these ly or totally dependent on site tenacity. pairs nested with a different mate; and (2) members This experiment was carried out on Little Duck of five displacedpairs managedto burrow under the Island, Hancock County, Maine, an island of 34 ha, screeninginto their original burrows, in spite of the located 15 km from the mainland. Pairs in 62 num- efforts taken to exclude them, and all of them nested beredburrows were capturedand bandedduring June with their mates from 1980. In addition to these two 1980. Before the arrival of birds at the colony in April groups,neither member of 10 displacedpairs or 7 1981, 25 of these burrows were randomly selected, undisturbed pairs was recovered, a significantly and their entrances were covered with a 25- x 15-cm higher proportionof displacedthan undisturbedpairs wooden shingle. Then, a 30- x 30-cm piece of coarse (X2 = 4.48, df = 1, P < 0.05).