Short Communications
and Commentaries
The Auk 111(3):703-710, 1994
SerrateTomia: An Adaptation for Nectar Robbing in Hummingbirds?
JtJAN FRANCISCOORNELAS
Departmentof Ecologyand Evolutionary
Biology,Universityof Arizona,
Tucson, Arizona 85721, USA
Hummingbirds are distinguished from all other
groups of birds by the range of bill-shape and billlengthvariation.Suchbill configurations
are thought
to match particular floral morphologies(e.g. Grant
and Grant 1968, Stiles 1978, Brown and Kodric-Brown
1979, Stein 1992). Considerable research has been
conductedon bill shapeand length variation;in contrast, serrations have not received much attention.
In
their extensivereview of the literature dealing with
bills (and tongues)of nectarivorousbirds, Paton and
Collins (1989) make no mention of bill serrations. Bill
serrations,a variable feature among hummingbirds,
have been studied only in a few species.Only two
genera, Androdonand Ramphodon,
have been consistently reportedashaving bills with distinctlyserrated
edges(e.g.Ridgway1890,1911,Johnsgard1983,Hilty
and Brown1986).Accordingto Ridgway(1911),however, this characteris present in other genera,and it
consists of a variable
number
of minute
serrations
in
the terminal portion of the maxillary and mandibular
tomia. In this paper I describebill serrationsin hummingbirds at the generic level and discusstheir functional significance.
straight or curved mandibles. Bill serrationsare very
conspicuouson both tomia in Androdonand Ramphodon,but in Glaucisand Sappho
serrationswere detected
only on the maxillary tomium. The serrationsare
presentonly on one-thirdto one-halfof thebill starting at thetip. Thosetowardthe tip usuallyarelonger
and slant posteriorlyat an angle of 45ø.
In the 24 generalacking hookedbill tips but having
serratedtomia, the serrationsat the tip of the bill are
not as enlarged as in the previous group. Sixteen
generahave serrationsonly on the maxillary tomium
and 8 generahave serrationson both tomia (seeTable
1).
Given the condition of many older specimensat
museums,it was often impossibleto countaccurately
the number of serrations.However, most genera have
about 20 to 30 serrations along the cutting edge of
the bill; serrationsare most abundant distally and
diminish proximally. A unique situation prevails in
Schistes
geoffroyi
in which serrationswere mainly observed
at the middle
of the bill.
Individuals
of An-
drodonand Ramphodon
may have more than 50 very
long and conspicuousserrations (Fig. 1). In ChrysuMethods.--I
studied skin collections at five museronia,Cynanthus,Goldmania,Oreonympha,Taphrolesbia,
bill serrationsare
ums (see Acknowledgments),examining skin speci- and in somespeciesof Chlorostilbon,
mensof 107 genera(98%of total of 110 genera)and weakly developedwhen present (seeTable 1). In such
311 species(92% of total of 320 known species)of
cases,the existenceof serratetomia mustbe regarded
hummingbirds. I checked at least 10 specimensof
with uncertainty.
each speciesfor bill serrationsusing a stereomicroThe shape of the bill is variable even among humscope and/or a hand lens. Most museum skins are mingbirds that lack a hooked tip. Most of the humprepared with the mouth closed (often tied shut). mingbirds with serrations have very pointed bills.
When this was the case,I opened the bills carefully This very sharp cutting edge resemblesa needle in
for examination.I recorded the following informa- Doryfera,Augastes,
Chrysolampis,
and Damophila.
In two
tion: (1) shapeof bill tip (e.g. hooked,wedge shaped); generathe terminal portion of the bill is very much
(2) the number and shapeof serrations(when pres- compressed(i.e. wedge shaped;Heliothryxand Auent); and (3) the distribution of the serrationsalong gastes;Ridgway 1911). The bill of the Fiery-tailed
the bill (e.g. serrationsonly on maxillary tomium).
Awlbill (Avocettula
recurvirostris)
is upturned with serResults.--Twenty-eightgenera (23% of humming- rationson both tomia (Fig. 1).
bird genera) have specieswith serratedbills (a total
Discussion.-The parallel bill structures in Androof 69 species;Figs.1 and 2). Four of thesegeneraalso don,Glaucis,Ramphodon,
and Sappho
and 24 other genhave hooked bill tips (Table 1). In the Tooth-billed
erahavenot been exploredcarefully.In the following
Hummingbird (A ndrodonaequatorialis),
both the upper
sections,I propose two possible functional explanaand lower bill (maxilla and mandible) are hooked,
tions for such morphological characteristicsin humand the edge of both tomia finely toothed (ca. 50 mingbirds.
serrationsalong 10 ram) or fringed near the tip (RidgIt has been hypothesizedthat the function of bill
way 1911, Johnsgard1983; Fig. 1). The genera Ram- serrationsis to facilitate the holding of insects(Ridgphodon,Glaucis,and Sapphohave hooked maxillasbut
way 1890, Johnsgard1983, Hilty and Brown 1986,
7O3
704
Short
Communications
andCommentaries
[Auk,Vol.111
A
i
Irnrn
I
1
tnlrn
D --'----I
I
•rnm
Fig. 1. Serratetomiain lateralview. Drawingswere madeby examiningseveralmuseumspecimens
(see
Acknowledgments).
Left (unhookedbills): (A) Colibricoruscans;
(B) Anthracothorax
prevostii;
(C) Thalurania
furcata;and (D) Avocettula
recurvirostris.
Right (hookedbills): (E) Androdon
aequatorialis;
(F) Glaucis
aenea;(G)
Sappho
shanganura;
and(H) Ramphodon
naevius.
Notechanges
in size(dataavailableonlyfor unhooked),
shape,
and orientationof serrationsamongspeciesin both groups.
Gosner1993).The hook and bill serrationsmay form habitats(open canopyand forestinterior, respectivepart of a feeding structureof hummingbirds that spe- ly; Ch•vez-Ramlrezand Dowd 1992,Chavez-Ramirez
cialize on hard-bodied arthropods.Such a morpho- and Tan 1993). This suggeststhat an associationof
logical device may serveto secureprey by increasing bill setrationswith foragingtechniqueand/or habitat
the coefficientof friction when catching or holding type is unlikely.
prey.
Little is known about how hummingbirds handle
Most hummingbirdsroutinely consumearthropods different types of arthropod prey. In aviaries, humasa sourceof protein (Young 1971,Remsenet al. 1986, mingbirdsseemincapableof manipulatingprey items
Ch•vez-Ramlrez and Dowd 1992, Ch•vez-Ramlrez and
with the bill or even holding them for more than a
Tan 1993). Mobbs (1979) describedfive methodsused few seconds(pers.obs.,Thompson1974).Mobbs(1979)
in captivity among hummingbirds when capturing observedthat "when a hummer capturesan insectin
insects and attributed
such variation
to habitat difflight (hawking), its forward movement forces the
ferences (e.g. open areas vs. forest interiors) or to prey so far to the rear of the gape that it is already
substrate (usually leaves or bark). Foraging tech- swallowed." It is possible,however, that highly inniques include flycatching and gleaning from flow- sectivoroushummingbirdsusethe bill serrationsfor
ers,leaves,and spiderwebs(Young 1971,Mobbs 1979). manipulating food items.
Two specieswith setrationson both tomia,the GreenGosner(1993) suggestedthat scopatetomia in birds
throated Carib (Sericotesholosericeus)
and Purple- (i.e. brushlike ridgeson cutting edgesof mandibles)
throated Carib (Eulampis
jugularis;Table 1), use dif- are adaptationsfor handling hard-shelledprey. If this
ferent arthropod foraging methods(flycatchingand is true, one would expect that hummingbirds with
flower gleaning,respectively)and foragein different serratedbills shouldfeed on aerial hard-bodiedprey.
July1994]
ShortCommunications
andCommentaries
A
705
TABLE1. Generaof hummingbirdswith minute serrations
on tomia.
Serrations
!
• 1•
Genera
Only
Hook
maxilla
Known
Both to rob
tomia
nectar
Aglaiocercus
B
Androdon
X
x
Anthracothorax
Augastes
A vocet tula
Chalybura
Chlorostilbon
i
i 1ram
x
Chrysolampis
Chrysuronia
Xa
Colibri
½
x
Cynanthus
Damophila
Doryfera
•
1 lmm
Elyira
x
Eulampis
Eupherusa
x
Glaucis
Goldmania
D
•
Heliothryx
Polytmus
Oreonympha
Ramphodon
Sappho
• lmm
x
X
X"
x
Schistes
x
X"
X
X
x
Sericotes
E
x
x
x
x
Taphrolesbia
x-
Thalurania
Trochilus
x
x
x
x
ßSerratetomia. when present,weakly developed(see text).
I
i lmrn
of hummingbirds,thosewith serratedbills specialize
on hard-bodied arthropods, nor that when humFig. 2. Photographsshowingonly upper mandi- mingbirdsforagetheyuseconsistentlythe sametechbles. Smooth tomia: (A) Campylopterus
hemileucurus nique and/or forageon the samesubstrata.A detailed
(UA, 8 972). Serrate tomia: (B) Cohgricoruscans
(U^,
studyis neededof the kind of arthropodsconsumed
13 897); (C) Thaluraniaridgwayi(UA, 8 025); (D) An- by bothhermitsandnonhermits.The arthropod-feedthracothorax
prevostii(UA, 13 010); and (E) Colibrico- ing hypothesisfor serratetomiaand hookedbills can
tuscans(UA, 13 898).
be evaluated as information accumulates on (1) dif-
Hummingbirdswith a hookand/or setrationson their
bills would be capturingand holding different types
of prey and usingdifferent foragingtechniques(e.g.
gleaning vs. flycatching)than birdswithout hooksor
setrations
on their
bills.
Little
is known
about the
types of arthropods hummingbirds consume, but
Remsen et al. (1986) found that hermits take a much
higher proportionof soft-bodiedspidersthan do most
nonhermits.Basedon this possiblepattern,one would
expectto find bill serrationsdistributedmostlyamong
nonhermit species.This seemsnot to be the case.
Accordingto the classificationof Sibley and Monroe
(1990), 25% of the genera in each group of hummingbirds have serrated bills (6.8% of all hermit speciesand 23.1%of all nonhermit species).
There is no evidence then that, among either group
ferencesin arthropoddietsof hummingbirds,and (2)
differencesin arthropod-foragingtechniques.
Nectar robbing is a behavior exhibited by various
speciesof hummingbirds and passerinebirds. Nectar
is obtainedthrough holes made near the basesof the
corolla tubes, in a manner generally circumventing
contactwith the sexualparts of the flowers (Inouye
1983).I hypothesizethat bill serrationsmay enhance
the exploitation of resourcessuch as long-tubed corollas and tough-tissuecorollasby facilitating:access
to protectednectaries;thegraspingof toughandwaxy
blooms;and the cutting of flower tissue.
There are several morphological similarities between passerine flower-piercers (Diglossa)and the
hummingbirds that have serratetomia. The maxilla
of a typical speciesof Diglossasweepsslightly upwardsfrom its baseand endsin a distinctstronghook
706
ShortCommunications
andCommentaries
[Auk, Vol. 111
T^I•I.•2. Nectarrobbingin the family Trochilidaeand plantsutilizedby nectar-robbing
hummingbirds.
Hummingbird species
Plant species(reference')
Anthracothorax
nigricollis
Primary nectar robbers
Tabebuia
serratifolia(3)
A. viridis
Malvaviscusarboreus(9)
Chlorostilbon
maugaeus
Erythrinaberteroana
(9), Neurodolphia
volubilis
(5), Malvaviscus
arboreus
(9),
Pitcairniaangustifolia
(9), P. bromeliifolia
(5), Tabebuia
haemantha
(5), T.
rigida(9), T. schumanniana
(5)
C. mellisugus
Plant not identified (7)
Barlena cristata(9), Tabebuiarosea(9)
Salviaspp. (8)
Plant not identified (9)
C. ricordii
Colibri thalassinus
Eulampisjugularis
Eupherusa
eximia
Plant not identified (13), Justiciaaurea(2), Malvaviscus
palmarus(2), Drymoniarubra(2), Drymoniaconchocalyx
(2), Raziseaspicata(2), Poikilacanthus macranthus(2)
Heliothryxbarroti
Erythrinasp. (10), Aphelandra
golfodulcensis
(6), Raziseasp. (12), Heliconia
Thaiuraniaridgwayi
Trochiluspolytmus
spp. (13)
Plant not identified (8)
Plant not identified (9)
Panterpeinsignis
Secondary nectar robbers
Plant not identified (11), Centropogon
talamancensis
(1)
Nectar
thieves
Plant not identified (9)
Marcgraviaspp. (12)
Calliphloxevelynae
Heliodoxa
jacula
Brugmansia
sp. (4)
Plant not identified (14), Aphelandra
golfodulcensis
(6), Justicia
aurea(6)
Centropogon
valerii(1)
Calatheaspp. (13)
Lesbiasp.
Phaethornis
longuemareus
Selasphorus
fiammula
Threnetes ruckeri
• References:
(1) Colwell et al. 1974;(2) Feinsingeret al. 1987;(3) ffrench 1973;(4) Gould 1861;(5) Kodric-Brownet al. 1984;(6) McDadeand
Kinsman1980;(7) Meyerde Schauensee
and Phelps1978;(8) pets.ohs.;(9) Quesada-Tyrrell
andTyrrell 1990;(10) Skutch1973;(11) Stiles1983;
(12) Stiles 1985; (13) Stiles and Skutch 1989; (14) Wetmore 1968.
(Vuilleumier 1969, Bock 1985) as in Androdon,Glaucis,
Ramphodon,
and Sappho.
Immediatelybehind the hook,
there is a notch in the maxillary tomium that is followed by two to four setrationsin Diglossa
(but these
are weak or almostabsentin somespecies;Bock1985).
Similarly, among hummingbirds,28 genera have the
edge of the tomia finely toothed near the tip, so that
the tip of the bill resemblesthat of a flower-piercer
bill (pets. obs.,Ridgway 1890).
The hook and setrations of the upper jaw, the flat
cutting edge of the mandibular tip, and the incomplete mandibulartube in Diglossa
speciesare all parts
of a specializedadaptive complexassociatedwith the
unusual method of nectar feeding used by flowerpiercers (Bock 1985). The bird robs nectar first by
holding the side of the flower with its hooked upper
jaw and cutting through the corolla wall with the
mandibular tip. With the corolla held in place, a longitudinal slit is cut into its near wall with the mandibular tip. Finally, the bird's tongue is protruded
through this slit into the flower to obtain nectar
piercing by short-billed hummingbirds(Beal 1880,
Ridgway1890,Grant1952,Skutch1954,Colwell1973,
Gentry1974,Janzen1975,Ingels1976,Feinsingerand
Colwell 1978,Snow 1981,Inouye 1980a,1983;for more
references, see Table 2).
Four behavioral categoriesamong nectar robbers
can be recognized(after Inouye 1980b):(1) highly
specializedprimary nectarrobbersperforatethe base
of long-tubedcorollas(gamepetalouscorollas)pollinated mainly by other hummingbirds;(2) secondary
nectar robbers utilize flowers perforated by other
hummingbirdsor passefineflower-piercers;(3) nectar
thieves use the opening utilized by pollinators but,
without biting or making holes, collect nectar from
flowersmorphologicallyadaptedfor pollinationby a
different classof visitors;and (4) baseworkers obtain
nectarby reachingbetweenthe petals.This behavior
is performedto flowers with polypetalouscorollas,
thereby bypassingthe opening used by pollinators
and gaining more direct accessto the nectary at the
baseof the corolla.Forexample,Ch•vez-Ramlrezand
(Skutch 1954, Vuilleumier 1969).
Dowd (1992) observedDominican caribsprobing be-
The Diglossaflower-piercersare perhaps the bestknown examples of nectar robbers among birds
(Skutch1954,Lyon and Chadek 1971,Colwell et al.
tween the petals(polypetalouscorolla)to takenectar.
1974, Snow 1981, Kodric-Brown
Inouye 1980b);however,carlbsalsoexhibitbehaviors
of a nectar thief, as observedby Ingels (1976). Ac-
et al. 1984, Arizmen-
di 1994), but several studies have reported flower
The observation indicates that caribs, which have ser-
rated tomia, fit the definition of base workers (after
July 1994]
ShortCommunications
andCommentaries
cording to Inouye (1980b),however, a baseworker
would not cut the flowerswith its bill, but the opening usedby pollinatorsis not usedeither. Carlbsapparently do not cut the flowers with their bills, but
would be better able to probebetweenpetalsdue to
the serrated tomia on their
bills. Sometimes
these
hummingbirds would collect nectar that has leaked
between the petals.
Someprimary nectarrobbers(after Inouye 1983)
are mostlyreferredto in the literature as"marauders"
(seeFeinsingerandColwell 1978),andsecondarynectar robbersand nectarthievesare consideredas parasitoids.However, I follow the categorizationmade
by Inouye (1980b) to minimize confusion. Because
nectarthievesprobablydo not greatly influencethe
activitiesof pollinators(Inouye 1980b,Stiles1985),I
do not include
them in further
discussion.
Hummingbirds exert strongselectivepressureson
corollamorphology.Forexample,Fenster(1991)suggestedthat corollalength is a floral specializationdue
to competition for pollinators. By increasingcorolla
length, plantsare discriminatingamonga setof pollinators(i.e. long-billed hummingbirds)that would
evolveas highly specializedon commonlyrich-nectar, long-tubed corollas(e.g. Wolf et al. 1976).
Once a mutualistic relationship has arisen, there is
a high probability that a third speciesmight evolve
to take advantageof mutualistic partnerships(Boucher et al. 1982,Inouye 1983,Sober6nand Martinez
del Rio 1985).Nectar robbersare probablythe bestdocumentedexampleof cheatingbehavior (Sober6n
and Martinez del Rio 1985) deriving benefitsfrom
flowers without pollinating, while competingwith
the pollinatorsfor nectar(McDade and Kinsman1980,
Arizmendi 1994). Some hummingbird flowers are
structurallyprotectedagainstnectarrobbers(Inouye
1980b),but most are unprotectedand are regularly
exploitedby nectarrobbers,which significantlyaffect
the amountof nectaravailableto the pollinators(Snow
1981,Arizmendi 1994).However, the ecologicalimpact of nectar robbing on both plant reproductive
fitnessand pollinators'foraging efficiencyremains
controversial (Hawkins 1961, Heinrich and Raven
1972, Koeman-Kwak 1973, McDade and Kinsman 1980,
Sober6n and Martinez del Rio 1981, Roubik 1982,
Inouye 1983, Kodric-Brown et al. 1984, Arizmendi
1994).
It has been assumedthat short-billed hummingbirds cannotfeed on flowerswith long corollas(Colwell et al. 1974),but thesespeciesoften obtain nectar
from flowers with long corollasby making perfora-
707
An obvious prediction is that serrated tomia should
be moreprevalentamongshort-billedhummingbirds
than long-billed hummingbirds.
There
are several immediate
benefits
to the cheater
when it performsthis behavior. First, nectarrobbing
is potentially more energy efficient than legitimate
flower visitation (Darwin 1889). Furthermore, it has
beendemonstratedwith beesthat the costof foraging
decreasesif nectar is obtainedby circumventingthe
flowerentrance(Weaver1956,Free1968,Inouye 1980a;
but see Sober6n and Martinez del Rio 1985). Second,
nectar robbershave accessto highly rich resources
(i.e. nectarin long-tubed corollas)that otherwise are
restrictedonly to long-billed pollinators.This behavior is advantageousto nectar robberswhen their legitimate accessto nectar in short-tubed corollas is
restricted(e.g. by competition,or by scarcityof accessible flowers).
One trade-offfor beingopportunistic(i.e. obtaining
nectarby bypassingthe opening usedby pollinators)
is rarity (Brian 1957, Stiles 1975, Kodric-Brown and
Brown 1979, Sober6n and Martinez del Rio 1985).
Individuals of a third speciestake advantageof mutualisticpartnershipsand becomedensitydependent
on the mutualistic interaction. Rarity among nectar
robberswould be the evolutionary outcomeof plant
specializationon a particular set of true pollinators
(e.g. long-billed hummingbirds) and the improvementof the robbingtechniquesof the excludednectar
exploiters (i.e. hummingbirds with short, serrated
bills). As predicted, nectar-robbing hummingbirds
seemto be rare in nature(Stiles1985,Hilty and Brown
1986)and/or erraticon a regionalscale.When present
at a smaller scale,however, nectar robbersare fairly
common(pers.obs.,Stiles1985,Hilty andBrown1986,
Feinsinger et al. 1987, Collar et al. 1992, Ornelas in
press).
Since nectar robbing lets the nectar exploiter become essentiallyindependent of floral morphology
(Stiles 1985), some behavioral attributes should be
associated
with this behavior.For example,Stiles(1983)
observed that Fiery-throated Hummingbirds (Panterpeinsignis)utilize flowers perforatedby passefine
flower-piercers and may follow the flower-piercer
from flower to flower, usingthe flower-piercers'holes
to extractthe nectar. This observationsuggeststhat
an inherent
difference
in behavior
should
exist be-
tween this nectar robber and a nonnectar-robbing
species(Brian 1957). In contrastto most hummingbirds that visit flowers by the obvious and correct
entrance,nectar robbersshould be expectedto have
tions at the base (Darwin 1889, Skutch 1973). This
a more plasticand lessstereotypedset of behaviors.
behaviorindicatesthat flowerswith long corollasse- In contrastto mostlegitimate visitors,nectarrobbers
crete nectar attractiveto short-billed hummingbirds need to keeptrackof the flowersthat havebeen puncsolong asthey areableto extractit (e.g.Kodric-Brown tured by the flower-piercers and have to deal with,
and Brown 1979, Inouye 1983, Koctric-Brownet al. for example,flowerswith well-protected
nectariessuch
1984, Feinsinger et al. 1987). Hummingbirds may as mostbromeliads.This suggeststhat nectarrobbers
pierce the base of flowers with long corollaswith a shoulddiffer from legitimatevisitorsin their tenacity
short,serrate,hooked,and/or exceptionallysharpbill.
and observationalskills when obtaining food. An-
708
ShortCommunications
andCommentaries
[Auk,Vol. 111
other differencebetween thesetwo groupsof hum-
brles-plantas
y el ladr6nden•ctarDiglossa
baritula
mingbirds is that individuals of most nectar robbers
(Passeriformes: Aves). Tesis Doctorado, Univ.
Nacional Aut6noma de M•xico, M•xico, D.F.
foragealone,and rarely defend territories(Wetmore
1968,ffrench 1973,Meyer de Schauensee
and Phelps BEAL,W.J. 1880. Fertilization of flowers by hum1978, Stiles 1985, Hilty and Brown 1986, Stiles and
mingbirds. Am. Nat. 14:126-127.
Skutch1989).Typically,nectarrobbersdo not estab- Boche,W.J. 1985. Is Diglossa(?Thraupinae)monophyletic?Pages319-332 in Neotropicalornithollish territories,but they are notablybellicose;they
are able to withstand attacks of most territorial humogy (P. A. Buckley,M. S. Foster,E. S. Morton, R.
mingbirds (Stiles 1985,Stiles and Skutch 1989).
S. Ridgely,and F. G. Buckley,Eds.).Ornithol.
Monogr. 36.
The rhamphothecal
featuresof the flower-piercers
are adaptations
for nectarivoryon flowerswith long BOUCHER,D. H., S. JAMES,X•qD K. H. KEELER. 1982.
corollas(Bock 1985). However, the flower-piercing
The ecologyof mutualism.
Annu.Rev.Ecol.Syst.
13:315-347.
method of nectarfeeding on flowers with deep corollas is not restrictedto the flower-piercersamong Bmx•, A.D. 1957. Differences in the flowers visited
passerines,but is found in two other genera (Coereba
by four speciesof bumblebeesand their causes.
J. Anim. Ecol. 26:71-98.
and Conirostrum)
and in severalgeneraof hummingbirds (e.g. Skutch 1954,Bock 1985,Stilesand Skutch BROWN, J. H., AND A. KODRIC-BRowN. 1979. Con1989;seeTable 2). This suggeststhat the complexof
vergence,competition,and mimicry in a temperate community of hummingbird-pollinated
featuresof the bill for nectarrobberyhas evolved
morethan oncein birdswith suchmorphology.Alflowers.Ecology60:1022-1035.
though a phylogenyof the Trochilidaeis not yet CHAVEz-RAMIREZ, F., AND M. DOwD. 1992. Arthroavailable, the existence of serrated tomia in otherwise
pod feeding by two Dominican hummingbird
species.Wilson Bull. 104:743-747.
dissimilarhummingbirdspeciesleadsme to suggest
that this feature has evolved severaltimes indepen- CHAVEZ-RAMiREZ, F., AND S.S. TAN. 1993. Habitat
dently within the family.
separationand arthropod resourceuse in three
With the evidencepresentedabove,it appearsthat
LesserAntillean hummingbirds.Condor 95:455458.
bill serrationsare an adaptationin somehummingA. MADROiqO
birds for nectarrobbing, but further supportfor this COLLAR,N. J., L. P. GONZAGA,N. KRABBE,
NIETO, L. G. NARANJO,T. A. PARKERIII, AND D.
ideais requiredfrom ontogenetic,phylogenetic,and
behavioral
studies.
C. Wœc;œ. 1992. Threatened birds of the Amer-
Acknowledgments.--I
thank Felipe Chfivez-Ramirez,
icas.TheICBP/IUCN red databook,3rd ed.,part
W. Mark Roberts,GaryD. Schnell,and LarryL. Wolf
for helpful discussion
and review of the manuscript.
2. International Council for Bird Preservation,
Ornito16gicadel Institutode Biologlade la Univer-
Diglossa
plumbea,
and its evolutionaryimplica-
Cambridge,United Kingdom.
Lucinda McDade, Stephen M. Russell, Peter Scott, COLWELL,
R. K. 1973. Competition and coexistence
SusanWethington, and Yaron Ziv provided invaluin a simpletropicalcommunity.Am. Nat. 107:
737-760.
able suggestionsconcerningearlier versionsof the
manuscript.I thank the following curatorsfor allow- COLWELL,R. K., B. J. BETrS, P. BUNNELL,F. L. CARing me to examine specimens under their care: G.
}'ENTER,
AND P. FEINSIIqGER.
1974. Competition
Barrowclough(American Museum of Natural Histofor the nectarof Centropogon
valeriiby the humry, AMNH); P. Escalanteand L. Navarijo (Colecci6n
mingbirdColibrithalassinus
and the flower-piercer
sidad
Aut6noma
deM•xico,IBUNAM);
T'HueIs(University of Arizona, UA); A. Navarro (Museo de Zoologia de la Facultad de Ciencias de la Universidad
Aut6nomade M•xico, MZFC); and A. T. Petersonand.
D. Willard (Field Museumof Natural History,FMNH).
tions. Condor
DARWIN, C.
1889.
76:447-452.
The effects of cross and self-fer-
tilization in the vegetablekingdom.D. Apleton,
New
York.
FEINSINGER,P., AND R. K. COLWELL. 1978. Commu-
Kris Sondereggerpreparedthe illustration.Funding
nity organization among Neotropical nectarfeeding birds. Am. Zool. 18:779-795.
wasprovidedby scholarship56254from ConsejoNaP., J. H. BEACH,Y. B. LINHART,W. H. BUSBY,
cionalde Cienciay Tecnologla(CONACyT-M•xico), FEnqSINGER,
and by a grant from the Universityof Arizona ReAND G. MURRAY.1987. Disturbance,pollinator
searchTraining Group (RTG) in the Analysisof Bipredictability,and pollination successamong
ologicalDiversification.
CostaRicancloudforestplants.Ecology68:12941305.
FENSTER,
C.B. 1991. Selectionon floralmorphology
by hummingbirds.Biotropica23:98-101.
LITERATURE CITED
FmlENCH,R. 1973. A guide to the birds of Trinidad
and Tobago.LivingstonPublishingCo.,WinneARIZMENDI,M. C. 1994. Interacciones eco16gicas
wood, Pennsylvania.
mfiltiples: E1 casodel sistemamutualistacoli- FREE,
J.B. 1968. Thebehaviorof beesvisitingrunner
July1994]
ShortCommunications
andCommentaries
709
beans(Phaseolus
multifiorus).
J. App1.Ecol.5:631-
MCDADE,L. A., ANDS. KINSMAN.1980. The impact
of floral parasitismin two Neotropical hummingbird-pollinatedplant species.Evolution 34:
GENTRY,
A.H. 1974. CoevolutionarypatternsinCen638.
tral American Bignoniaceae.Ann. Missouri Bot.
Garden
61:728-759.
944-958.
MEYER DE SCHAUENSEE,R., AND W. H. PHELPS. 1978.
GOSNER,
K. L. 1993. Scopatetomia: An adaptation
for handling hard-shelledprey?Wilson Bull. 105:
316-324.
GOVLD, J. 1861. Introduction to the Trochilidae or
family of humming-birds.Printed (for the author) by Taylor and Francis, London, United
Kingdom.
GRANT,V. 1952. Isolation and hybridization between Aquilegiaformosaand A. pubescens.
Aliso
2:341-360.
A guideto the birdsof Venezuela.PrincetonUniv.
Press,Princeton, New Jersey.
MoBBs,A.J. 1979. Methodsusedby the Trochilidae
when capturing insects.Avic. Mag. 85:26-30.
ORNELAS,
J. F. In press. Capltulo II. Las especies
extintasy amenazadassecci6n2. Aves:familia
Trochilidae.In VertebradosMexicanosextintosy
en peligro de extinci6n (G. Ceballosand D. Navarro, Eds.).Univ. Nac. Aut6n. M•x., M•xico, D.F.
PATON, D.C.,
GRANt,K. A., ANDV. GRANt. 1968. Hummingbirds
and their flowers. Columbia Univ.
Press, New
York.
HAWKINS,R. P. 1961. Observationson the pollination of red clover by bees:I. The yield of seeds
in relation to numbersand kinds of pollinators.
Ann. App1. Biol. 49:55-65.
HEINRICH,B., ANDP. H. RAVEN.1972. Energeticsand
pollination ecology.Science176:597-602.
HILTY,S. L., ANDW. L. BROWN.1986. A guide to the
birds of Colombia. Princeton Univ. Press, Prince-
ton, New Jersey.
INGELS,
J. 1976. Observationson the hummingbirds
AND B. G. COLLINS. 1989. Bills and
tonguesof nectar-feedingbirds:A review of morphology,functionand performance,with intercontinental comparisons.Aust. J. Ecol. 14:473506.
QUESADA-TYRRELL, E., AND R. A.
TYRRELL. 1990.
Hummingbirds of the Caribbean. Crown Publishers, New York.
REMSEN,J. V., F. G. STILES,AND P. E. SCOTT. 1986.
Frequencyof arthropodsin stomachsof tropical
hummingbirds.Auk 103:436-441.
RIDGwAY,R. 1890. The hummingbirds. Rep. Am.
Mus. Nat. Hist., New York.
RIDGWAY,R. 1911. The Birds of North and Middle
Orthorhynchus
cristatus
and Eulampis
jugularisof
America, part 5. Smithsonian Institution Press,
Martinique (West Indies). Gerfaut 66:129-132.
Washington, D.C.
INOVYE,D. W. 1980a. The effect of proboscisand ROUSlIC,
D.W. 1982. The ecologicalimpact of nectarcorollatube lengthson patternsand ratesof flowrobbing beesand pollinating hummingbirdson
er visitation. Oecologia45:197-201.
a tropical shrub. Ecology63:354-360.
INOVYE,D. W. 1980b. The terminologyof floral lar- SIBLEY,C. G., AND B. L. MONROE. 1990. Distribution
ceny. Ecology61:1251-1253.
and taxonomyof birds of the world. Yale Univ.
INOUYE,D.W. 1983. The ecologyof nectarrobbing.
Press, London.
Pages 153-173 in The biology of nectaries (B. SICUTCH,A. F. 1954. Life histories of Central AmerBentleyand T. Elias,Eds.).ColumbiaUniv. Press,
ican birds, vol. 1. Pacific Coast Avifauna Series
New
York.
JANZEN,
D.H.
31.
1975. Ecologyof plantsin the tropics.
Edward Arnold,
London.
Crown Press, New York.
JOHNSGARD,
P.A. 1983. The hummingbirdsof North
America. Smithsonian Institution
Press, Wash-
ington, D.C.
KODmC-BRowN,A., AND J. H. BROWN. 1979. Com-
petition betweendistantly related taxa in the coevolution of plantsand pollinators.Am. Zool. 19:
115-127.
KODRIC-BROwN,
A., J. H. BROWN,G. S. BYERS,AND D.
F. GORI. 1984. Organization of a tropical island
community of hummingbirdsand flowers.Ecology 65:1358-1368.
KOEMAN-KwAK,
M. 1973. The pollination of Pedicularispallustrisby nectar thieves (short-tongued
bumblebees). Acta Bot. Neederlandica 22:608-615.
LYON,D. L., ANDC. CHADEK. 1971. Exploitation of
nectar resourcesby hummingbirds, bees (Bornbus),and Diglossa
baritulaand its role in the evolution
of Penstemon kunthii. Condor
SICUTCI-I,
A. F. 1973. The life of the hummingbird.
73:246-248.
SNOw,D.W. 1981. Coevolutionof birds and plants.
Pages169-178 in The envolving biosphere(P. J.
Greenwood, Ed.). Cambridge Univ. Press,Cambridge.
SOBERON,J., a2q'DC. MmlTINEZ DEL RIO. 1981. The
dynamics of a plant-pollinator interaction. J.
Theor. Biol. 91;363-378.
SOBERON,
J., AND C. MARTINEZDELPdo. 1985. Cheat-
ing and taking advantagein mutualisticassociations.Pages192-216 in The biology of mutualism (D. H. Boucher, Ed.). Croom Helm, London.
STEIN,B. A. 1992. Sicklebill hummingbirds, ants,
and flowers.
Bioscience
42:27-33.
STILES,
F.G. 1975. Ecology,floweringphenologyand
hummingbird pollination of some Costa Rican
Heliconiaspecies.Ecology56:285-301.
STILES,
F. G. 1978. Ecologicaland evolutionary im-
710
ShortCommunications
andCommentaries
plications of bird pollination. Am. Zool. 18:715-
VUILLEUMIER,
F. 1969. Systematicsand evolution in
Diglossa
(Aves,Coerebidae).Am. Mus. Novit. 2381.
727.
STILES,
F.G. 1983. Speciesaccounts.Birds.Pages593594 in CostaRican natural history (D. H. Janzen,
Ed.). Univ. ChicagoPress,Chicago.
STIEES,
F.G. 1985. Seasonalpatternsand coevolution
in the hummingbird-flowercommunityof a Costa Ricansubtropicalforest.Pages757-787 in Neotropical ornithology (P. A. Buckley,M. S. Foster,
E. S. Morton, R. S. Ridgely, and F. G. Buckley,
Eds.).Ornithol. Monogr. 36.
STILES,
F. G., AN• A. F. SKU'rCH.1989. A guide to the
birds of Costa Rica. Cornell Univ. Press, Ithaca,
New
[Auk, Vol. 111
WEAVER,
N. 1956. The foraging behavior of hon-
eybees
onhairyvetch.I. Foraging
methods
and
learning to forage. InsectsSociaux3:547-549.
WEtMOP,
E, A. 1968. The birds of the Republic of
Panama.Smithson.Misc. Coil., vol. 150, part 2.
WOLF, L. L., F. G. STILES,AND F. R. HAINSWORTH. 1976.
Ecologicalorganizationof a tropical highland
hummingbird community.J.Anim. Ecol.45:349379.
YOUNC;,
A.M. 1971. Foragingfor insectsby a tropical
hummingbird. Condor 73:36-45.
York.
Received
2 December
1993,accepted
22 January1994.
THOMPSON,
A. L. 1974. A new dictionary of birds.
McGraw-Hill,
New York.
The Auk 111(3):710-713, 1994
Divergence in the Mitochondrial DNA of Empidonax traiIIii and
E. alnorum, with Notes on Hybridization
•
WINY•R •
Academyof Natural Sciences,
BenjaminFranklinParkway,Philadelphia,
Pennsylvania
19103,USA
The Empidonax
flycatchersrepresenta morphologically conservativespeciesassemblage.This conservatismhasmaderecognitionof specieslimits difficult,
and there hasbeena recenttrend toward recognizing
separatespeciesamong populationsformerly considered conspecific
(e.g. Stein 1958,1963,Johnson1980,
Johnsonand Marten 1988).Empidonax
flycatchersrepresent a group where molecular systematicscan addressa variety of heretofore unansweredquestions,
especiallythoseregarding degreesof distinctivehess
and gene flow between populationsor species(see
alsoHewitt 1988,Avise and Ball 1991).Sibling species
such as the Willow Flycatcher(E. traillii) and Alder
Flycatcher(E.alnorum)
are particularlyinterestingwith
regard to the evolution of intrinsic isolating mechanisms,an important componentof the speciationprocess.The degreeto which intrinsic reproductiveisolating mechanisms
(i.e. ability to discriminatebetween
con- and heterospecifics,
resulting in assortativemating) have arisenbetweentwo allopatricpopulations
will affecthow distinctthesegene poolswill remain
following secondarycontact.
A scarcityof recognizedhybridsamongEmpidonax
flycatcherssuggeststhat, althoughthey are morphologically conservative,their intrinsic reproductive
isolatingmechanisms(or cohesionmechanisms;
Tern-
pleton 1989)are remarkablywell developed.Perhaps,
however, it is our inability to recognize Empidonax
hybrids, rather than their true rarity, that leads to
their apparent scarcity.Two of the five recognized
tyrannid hybrids (seeShort and Burleigh 1965,Phillips 1966, Phillips and Short 1968)are intergeneric,
suggestingthat congenerichybrids are being overlooked.Further,becauseindividualsof the genusEmpidonax(including E. traillii) have produced hybrids
with heterospecifics(see Short and Burleigh 1965,
Phillips 1966), hybridization might be predicted between the closely related E. traillii and E. alnorum.
Seutin and Simon (1988) arrived at the same conclu-
sion in a different manner, suggestingthat the close
phenotypic similarity and habitat preferencesof E.
alnorum
andE.traillii,togetherwith the extensivezone
of sympatry,makeshybridization betweenthem likely. These authors failed to find evidence of hybridization, however, and concluded that these species
were reproductivelyisolatedin southeasternCanada.
They sought evidence of hybridization using allozyme electrophoresis,although this technique had
already revealed no fixed allelic differencesbetween
the two speciesin Minnesota(Zink and Johnson1984).
Without a fixed-allelicdifference,the ability to detect
hybrids is compromised.Another problem with the
study was that it did not include traillii from allopatric
populations.
* Present address: National
Museum
of Natural
His-
My study consistedof three parts:(a) estimating
tory, Division of Birds, SmithsonianInstitution, the level of divergencein the mitochondrialDNA
Washington,DC 20560, USA.
(mtDNA) of E. traillii and E. alnorum;(b) finding spe-