Short Communications
and Commentaries
The Auk 109(2):383-385, 1992
Further Information on the Geneticsof Bill Crossingin Crossbills
JEFFREY
G. GROTH•
Museumof Vertebrate
Zoology,Universityof California,Berkeley,
California94720,USA
Ticehurst(1910) was among the first to note that
White-wingedCrossbills(Loxialeucoptera)
with rightcrossinglower mandibles substantiallyoutnumber
left-billed birds, while in the Red Crossbill (L. cur-
leucoptera
(sexes
combined),
218wereleft-billed,while
under a 3:1 ratio only 196would be expected.These
data gave X2 = 3.29 (df = 1, with P < 0.10; not P >
0.10 as indicated by Benkman 1988:578),which in-
virostra)the ratio is nearly 1:1. Jameset al. (1987)
found that in L. leucoptera
the ratio of right- to leftbilled birds is approximately3:1. In a large sample,
dicates a substantial
Benkman(1988) found a 3:1 ratio in L. leucoptera
and
suggestedthat this ratio reflectssimple Mendelian
differentfrom 3:1(72 expected,X2 = 6.0,df = 1, P <
0.02), but actuallycloserto 2:1 (96 expected).
In supportof his hypothesisof right-crossingdom-
inheritance.
The 3:1 Mendelian ratio is well-known in genetic
deviation
from 3:1. Benkman
di-
vided his sampleby sexand found 90 left-billed females out of 288, which is not only significantly
inance, Benkman(1988) observedseven right-billed
fed by four right-billed adults
systemsof one locuswith two allelesand complete juvenile L. leucoptera
dominance.This ratio is expectedin the specialcase (presumablythe parents)in the field. This observaof the phenotypesof F• progenyin matingsbetween tion is consistentwith a right-dominant system,but
two heterozygotes.
Benkman(1988)proposedthatthe it doesnot falsify other interpretations.Verification
genefor right-crossingis dominant,thereforegiving of the right-dominant mechanismwould have been
riseto threetimesasmanyright-billedbirds.He fur- more conclusiveif offspringof two left-billed birds
ther proposedthat the frequenciesof the two alleles were examined, becausethese matings would theomustbe 0.5 and 0.5 in order for the speciesasa whole reticallyproduceonly left-billed offspring.
to show the 3:1 Mendelian
ratio.
I observedmatingsbetweentwo left-billed L. curAn obviousflaw in this propositionis that there is virostrain captivity,which producedsevenoffspring
no particularreasonto expectentire populationsor (in three broods),two of which were right-billed (one
speciesto show a 3:1 Mendelian ratio, even if the each in the first and last broods).This should not have
geneticmechanismsfor morphismare inherited in a occurredunder a right-dominantmodel,becauseleftMendelian
fashion.
In addition
to the "Mendelian"
ratio in L. leucoptera,
Benkman (1988:578)cited two
other examplesfrom the literatureof polymorphisms
occurringin "simple Mendelian ratiosin wild populationsof birds."In oneof the examplescited,Cooke
and Cooch(1968)did not discussmorph frequencies
in populations,but ratherfrequencies
in progenyfrom
crossesbetween morphs. In a secondstudy, Smith
(1987) measuredfrequenciesof morphs in 10 geographic regions.Eachregion had a different morph
frequencyand had closeto "Mendelian" 3:1 ratiosin
only one or two regions.
Benkman's(1988) calculationsof chi-squarevalues
for morph frequenciesin L. leucoptera
were in error
(apparentlyonly oneof the two compartments
in each
contingencytablewascounted),leadingto his acceptance of the 3:1 ratio as true.
I reexamined
his data
and recalculatedthe chi-squarestatistics.Of 784 L.
• Present address:Department of Ornithology,
American Museum of Natural History, New York,
New York 10024, USA.
383
billed birds would be homozygousrecessiveand all
resultingprogenywould be left-billed. However, an
alternative,left-dominant systemwould not be falsified by this result. If both parentswere heterozygous,an averageof one right-billed offspringshould
occur out of every four.
The
left-dominant
model
is the reverse
of Benk-
man's,yet it also is consistentwith his observation
of right-billed adults with right-billed juveniles. In
the left-dominant model, right-billed birds would be
homozygousrecessiveand, therefore,would produce
only right-billed offspring. Of course,it is conceivable that the two specieshave different geneticmechanisms for bill crossing.However, without further
information,it is mostparsimoniousto hypothesize
that sucha mechanismevolved only once in Loxia.
Basedon Benkman's(1988)morph frequenciesand
assumingHardy-Weinbergequilibrium in the species,a left-dominantmodelpredictsgenefrequencies
in L. leucopterato be 0.150 (left) and 0.850 (right).
There is no reasonthesefrequenciesshould be less
likely than 0.5 and 0.5. It is clear that the (roughly
approximate)3:1phenotypicratio producedin nature
384
Short
Communications
andCommentan•es
has nothing to do with "Mendelian" 3:1 ratiosand is
only coincidentally similar.
I recordedbill-crossingdirection in 3,647 specimens of North
American
L. curvirostra. Sexes com-
bined, 1,895were left-billed and 1,752were rightbilled,which issignificantlydifferentfroma 1:1ratio
(X2 = 5.61, df = I, P < 0.02). I found that 1,182 of
2,256 males (52.4%, X 2 = 5.17, df = I, P < 0.05) were
left-billed, and 713 of 1,391 females (51.3%, X 2 = 0.88,
df = 1, P > 0.3) were left-billed,showinga slight
excess
of thismorphin bothsexes.Usingthe original
phenotypicfrequenciesin L. curvirostra
(and HardyWeinberg assumptions),under the as yet unfalsified
left-dominant model, the frequenciesof the two alleleswould be 0.693(right) and0.307(left). Therefore,
it would be likely that two left-billed L. curvirostra
chosenat random(suchas in the first mating of captives discussedabove) would be heterozygous,becauseheterozygoteswould compriseabout 78%of all
left-billedbirds.Furtherbreedingexperimentsshould
[Auk,Vol.109
ing it lessprofitablefor individuals to forageon cones
previouslyvisited by the samemorph. He postulated
that revisitationis commonwhen birds forageon
conesof Pinus,
while it is notcommonor unimportant
for other cones(usedby L. leucoptera),
suchas those
on spruce (Picea)and larch (Lan'x),which are removed
from branchesor not revisited. This hypothesisis
testable,becausepopulationsof crossbills
vary in conifer usage.Forexample,onewouldexpecta 1:1morph
ratio in pine-dependentL. leucoptera
megaplaga
of Hispaniola.
If frequency-dependentselectionis important, a
secondexpectationis that populationsof L. curvirostra
not associated with Pinus cones, such as small-billed
forms,would have fluctuationsaway from 1:1, as in
L. leucoptera.
To testthis,I estimatedbill-crossingfrequenciesusinga subsetconsistingof all L. curvirostra
specimens(fully-grown birds only) with lower-mandible widths of 9.0 mm or less (n = 814; most of these
specimens were labeled L. c. minor or L. c. sitkensis).
be performedto assessother unfalsifiedgeneticor
In thissample,411wereleft-billedand403wererightbilled. Thesefrequenciesare very similar to those
Severalothergeneticmechanisms
for bill crossing found in the total sampleof L. curvirostra.
can be evaluated with the existing data. Although
The developmentalmechanismresponsiblefor billsexesdiffered slightly in morph frequenciesin both crossingdirectionin Loxiaremainsincompletelyunspecies,neither right- nor left-dominant models of derstood.Severaltypesof abnormalbill crossingin
nongenetic hypotheses.
sex-linked
inheritance
would
be
consistent
with
domesticfowl have genetic bases(Landauer 1938),
but none of theseappearrelated to crossingmechamodeof inheritanceisknown (e.g.the GouldJanFinch, nismsin Loxia.Jameset al. (1987)speculatedthat nonChloebia
gouldiae
[Southern1945]and nestlingsof the geneticdevelopmentalresponsesby juvenile birds to
Mute Swan,Cygnusolar[Munro et al. 1968]),the sexes cone-scalespiralingdirection(phyllotaxy)is the cause
differ markedly in their phenotypic frequencies.In
of crossbillbill-crossingratiosin the wild. Their hythe four possible 6ne-locus/two-allele/sex-linked
pothesispredicted that cone-morphfrequenciesin
systemsfor crossbills(left and right dominantin each food treesof L. leucoptera
shouldmatchthoseof bird
species),the phenotypic frequenciesin the females bills; however, they admitted that available data did
(the heterogameticsex)would need to equal the hy- not supporttheir hypothesis.This contradictiondoes
potheticalgene frequenciesfor the entire population not prove that crossingdirection in crossbillshas a
(the males would have Hardy-Weinberg propor- geneticbasis.No currentdatadisprovethe hypothesis
tions). Yet, in none of the four cases for Loxia do
that directionality has the potential for environmencalculationsusing these modelsfit the data.
tal modificationduring the critical stagewhen the
Even if the two speciesof crossbill have different
mandible tips of juveniles grow enough to cross.
geneticmechanismsand right dominanceholds for Asymmetrical muscle development on the sides of
L. leucoptera,
Benkman(1988)did not explainwhy two the head and behavioral "handedness"(seeKnox 1983)
alleles should exist at frequenciesof precisely 0.5. for example, potentially could develop only after
One hypothesisis that these frequenciesare main- crossingdirection has been determined. It would be
tained becausethe polymorphism is balanced by se- worthwhile to study theseauxiliary changesin prelection, or there is selectionfavoring he•erozygotes crossingjuveniles. Onset of thesechangesbefore the
(Ford 1945).Instead,Benkman(1988:579)implied that developmentof bill crossingwould favor the hyselectionon bill-crossingdirection is not operative in
pothesisof geneticdeterminismin crossingdirection.
L. leucoptera.
Without selection,and only chanceand Nongenetic (sensuJameset al. 1987) or random facgenetic drift operating, there is no reasonto predict torsmight explainthe 1:1phenotypicfrequencyseen
that gene frequenciesshould approach0.5; instead, in L. curvirostra, but this would leave the skewed disbecauseof finite populationsize, the chancesof any tribution in L. leucoptera
unexplained. Another quesavailable
evidence.
For birds
in which
a sex-linked
population remaining at a 50:50 gene frequency become lesseach generation (Wallace 1981).
On the other hand, Benkman(1988:579)did suggest
tion is whether there may be similarities between
crossbill
"handedness"
and handedness
in human
populations,which consistentlyshow a great pre-
that L. curvirostra
experiencesfrequency-dependent dominance of right-handers. Human handednessis
selection because birds revisit cones, therefore
mak-
not inherited in a simple Mendelian fashion, and is
April1992]
Short
Communications
andCommentaries
thought to be related to functional asymmetryin
hemispheresof the brain (Annett 1985).
The captivebirdswere providedaviaryspaceat the
Field Stationfor BehavioralResearch,Universityof
California, Berkeley.I thank curatorsand staffat the
following institutionsfor providing loans or access
to specimens:Museum of Vertebrate Zoology, California Academyof Sciences,SanDiego Natural HistoryMuseum,Universityof Californiaat LosAngeles,
CaliforniaStateUniversityat Long Beach,University
of Californiaat SantaBarbara,University of Arizona,
DelawareNatural History Museum,WashingtonState
University, National Museum of Canada, Louisiana
StateUniversity Museumof Zoology,CarnegieMuseumof Natural History, University of Kansas,University of Washingtonat Seattle (Burke Museum),
Denver Museum of Natural History, University of
Nebraskaat Lincoln, University of Wisconsin,Santa
BarbaraMuseum of Natural History, Royal British
Columbia Museum, University of Puget Sound,Cornell University, American Museum of Natural History, and Universityof MontanaZoologicalMuseum.
DouglasA. Bell,Ned K. Johnson,ThomasBatesSmith,
and JohnTrochetgaveusefulcommentson the manuscript.
BENKMAN, C.W.
385
1988. A 3:1 ratio of mandible
cross-
ing direction in White-winged Crossbills.Auk
105:578-579.
COOKE,
F., ANDF. G. COOCH.1968. The geneticsof
polymorphism in the goose Anser caerulescens.
Evolution
22:289-300.
FORD,E. B. 1945. Polymorphism.Biol. Rev. Camb.
Philos.
Soc. 20:73-88.
JAMES,P. C., T. W. BARRY,A. R. SMITH, AND S. J. BARRY.
1987.
Bill crossover ratios in Canadian
crossbills
Loxiaspp. Ornis Scand.18:310-312.
KNOX, A. G. 1983. Handedness in the crossbills Loxia
and the Akepa Loxopscoccinea.
Bull. Brit. Ornithol. Club
103:114-118.
LANDAUER,W. 1938. Notes on cross-beakin fowl. J.
Genet.
37:51-68.
MUNRO, R. E., L. T. SMITH, AND J. J. KUPA. 1968. The
geneticbasisof color differencesobservedin the
Mute Swan (Cygnusolor). Auk 85:504-506.
SMITH,T. B. 1987. Bill size polymorphism and intraspecificniche utilization in an African finch.
Nature
329:717-719.
SOUTHERN,
H. N. 1945. Polymorphismin Poephila
gouldiaeGould. J. Genet. 47:51-57.
TICEHURST,
C. B. 1910. Dimorphism in the crossbill.
Brit. Birds 3:261-262.
LITERATURE CITED
WALLACE,
B. 1981. Basicpopulation genetics.Columbia Univ. Press, New York.
ANNErr, M. 1985. Left, right, hand, and brain: The
right shift theory. Lawrence Erlbaum, London.
Received
18 February1991,accepted
13 January1992.
The Auk 109(2):385-389, 1992
Estimating Absolute Densities of Flying Seabirds
Using Analyses of Relative Movement
LARRY SPK4R, NADAv NUR, AND DAVID G. AINLEY
PointReyesBirdObservatory,
StinsonBeach,California94970,USA
The 300-m-band-transect
samplingmethodis currently the most widely used method for counting
seabirdsat seabecauseit provides an estimateof density (birds/unit area;reviewed by Tasker et al. 1984,
Haney 1985; see also Burnham et al. 1980). Density
likely that the conversionof raw countsof all birds
estimates
seen (birds/unit time) to bird densities will ever be
derived
from these counts often are affected
by variation in detectionratesof seabirdscausedby
severalfactors,including bird size, color and behavior, aswell asweatherand observerability. However,
the most critical
bias results from the effect of move-
ment by flying birds (reviewed by Taskeret al. 1984).
Counts of flying seabirdsare actually a measureof
bird "flux" and, thus, are usually an overestimation
of absolute(i.e. true) density(J.A. Wiens, D. Heine-
man, and W. Hoffman, 1978 unpubl. report to National Oceanic and Atmospheric Administration,
Boulder, Colorado). Due to the effects of bird movement, Tasker et al. (1984) concluded that it was "un-
possible,"and Haney (1985) concludedthat calculationsof absolutedensitieswould not be possiblewithout "considerableadditional qualifications." Yet, a
standardizedapproachto seabirdcensuses
is essential
for valid comparisonsbeweenstudies,an important
considerationin view of the recentupsurgeof seabird
studiesat sea.Accuratedensityestimatesare particularly critical for calculatingenergy fluxesand food