And Golden-Crowned (R

ECOMORPHOLOGY OF THE NORTH AMERICAN RUBY-CROWNED (REGULUS CALENDULA ) AND GOLDEN-CROWNED (R. SATRAPA) KINGLETS

ALLEN KEAST AND $ARI SAUNDERS Departmentof Biology,Queen' s University,Kingston, Ontario K7L 3N6, Canada

ABSTRACT.--Weanalyzed body morphology relative to habitat utilization and locomotory feeding movementsin the Ruby-crowned(Regulus calendula) and Golden-crowned(R. satrapa) kinglets. The Ruby-crownedKinglet hasa relatively longer, wider, and shallowerbill, longer rictal bristles, and a longer humerus and ulna. The capture of aerial prey and the plucking of insectsfrom foliage,which require somewhatcomplicated aerial maneuvers,made up 24 and 27%of the Ruby-crownedKinglet's locomotoryfeeding movements;and they made up 2 and 13%of the Golden-crownedKinglet's feeding movements. The Golden-crownedKinglet has relatively shortertarsi and longer toesthan the Ruby-crowned,and highly groovedsoles of the feet. Golden-crownedKinglets hung or clung from foliagein 43%of prey attacks, whereasthe Ruby-crownedKinglet hang-gleanedin only 7%.The Golden-crownedKinglet fed predominantly in conifers, the Ruby-crownedin deciduoustrees. In somemajor patternsof divergenceand specialization,the two North American species are comparableto those found by Leisler and Thaler (1982) for the PalearcticR. regulusand R. ignicapillus.Received 17 April 1990, accepted1 April 1991.

THE GENUSRegulus, a group of small-bodied of the bill at the skull; length and width of mandible; insectivorouswarblers, comprises two pairs of sternum length and width; lengths of femur, tibio- species;one pair occursin the northern forests tarsus,tarsometatarsus, and middle toe (foot digit 3) of the Nearctic and the other in the Palearctic without claw; lengths of humerus, ulna, carpometa- forest. Leisler and Thaler (1982), in an inno- carpus,hand digit 3 (phalanx 1 and phalanx 1 +2). (2) vative study involving quantificationof mor- Externalmorphology (from museumskins)--bill length, tip to rear of nares; bill width and depth at rear of phological characters,feeding patterns, and nares;length of rictal bristles;lengths of wing chord, habitat utilization patterns, compared the Pa- tail, centralmost and outermost rectrices, tarsometa- learctic Goldcrest(R. regulus)and Firecrest(R. tarsus,middle toe (digit 3) without claw, middle toe ignicapillus).Differences were found in bill claw, hallux without claw, hallux claw, footspanwith width, length of rictal bristles, length of hu- claw;lengths of alula, primaries10-5, and secondaries merus,shape of wing and tail, length of alula, 1 and 2. Becausethe skeletonsand skins necessarily and morphologyof the solesof the feet. These representeddifferent individuals, the two data sets features reflected differencesin foraging and are treated independently and some measurements habitat-use patterns. were repeatedfor each series.Body weight, wing aspect ratio (ratio of length to breadth of wing), and We made a comparative study of the two wing loading (massper unit area of wing; g.cm-2) North American species, the Ruby-crowned were obtained from live birds trapped during north- Kinglet (Reguluscalendula) and the Golden- ward migration at Prince Edward Point, Ontario. crowned Kinglet (R. satrapa).We analyzed the Drawings of the feet of these birds were made. Some morphologyof eachspecies and related differ- of the measurementsare not independent and over- ences to locomotory feeding movements and lap (see also Leisler and Thaler 1982). Examplesare substrateutilization patterns. the bill measurements,and "footspan with claws" relativeto "hind toelength," and "hind clawlength." METHODS Footspanis an adaptive complex reflecting the individual's ability to perch on substratesof different di- Morphometricmeasurements.--We measured 8 skel- ameters;hind toe length reflects its ability to cling etons and 10 museum skins of males of each species from vertical surfaces.The units of the foot are po- as follows: (1) Skeleton--we measuredtotal length of tentially adaptive individually and as a member of a fused vertebral column (thoracic, lumbar, and sacral complex.The additional measurementsare included vertebrae combined); length of cervical vertebrae;to- on this basis. tal skull length; maximum skull width and depth; bill Morphological variables were grouped into func- length, frontal-nasal hinge to tip; width and depth tional units related to specific biological functions,

88O The Auk 108: 880-888. October 1991 October1991] KingletEcomorphology 881 following Oxnard (1973, 1975),and Leislerand Wink- load stronglyand relatively evenly on factor 1 let (1985).The morphologicaldata were analyzedwith (Table 1A). Femur, tibia, and tarsus function as factoranalysis to resolvedifferences between species, a group and appearto representa size-of-leg supplementingabsolute measurements. Results pre- factor which separatesthe two species. The sentedfor factoranalyses are factorcorrelations (load- Ruby-crownedKinglet hasthe longer leg (Fig. ings)for orthogonallyrotated variax solutions, to best 1A). The middle toe (digit 3) does not appear suit the data,using MacintoshStatView SE-Graphics, to be associated with the other three variables. V.1.02. We were interested in analyzing groups of charactersrelated to specificbiological functions log- Factor2 explains99% of the variance in digit 3 ically or through previousstudies in the literatureso but doesnot clearlyseparate the kingletspecies. that any relative differencesin these could be com- (B) Legand foot dimensions,museum skins (tar- pared directly with the ecological data we have a sometatarsus,middle toe [digit 3], middle claw, reason to think should be related. hind toe [hallux], hind claw, footspan with Factor1 wasdesignated as an overall approximation claw). The two speciesseparate along a factor of sizevariation within the functionalcomplex (which of tarsus relative to hind toe and hind claw is generallytrue) when (1) all loadingson that factor (Tables 2 and lB, Fig. lB). The Ruby-crowned were in the same direction correspondingto a joint increase (or decrease) in all measures, (2) the coeffi- Kinglet has relatively longer tarsi and shorter cients were of similar magnitude, and (3) the factor hind toesand hind claws.Factor 2 loadsstrong- representeda large proportion of the variance in the ly for footspanwith claw and doesnot separate data. When these three conditions were met, the sub- the species. sequentfactors were interpreted asaxes of shapevari- (C) Wingaxis, skeletal (humerus, ulna, carpo- ation within the functional complex (see Mosimann metacarpus,phalanx 1, phalanx 1+2). The spe- and James 1979, Bookstein et al. 1985, Grant 1986). cies are separatedalong factor 1 or "size of When these criteria were not met, the essence of a wing," which loadspositively for all elements factor representing "size" in that complex of char- (Table1C, Fig. 1C).Factor 2, interpretedas distal acters is noted. relative to proximal elementsof wing, alsosep- Locomotoryfeeding movements and substrateutilization.--Data were obtained in southeastern Ontario at aratesthe two speciesbut to a lesserextent than Little CataraquiCreek, near Kingston,and Prince Ed- factor 1. There is a trend towards longer distal ward Point, near Picton, between 15 April and 10 relative to proximal elements in the Golden- May, 1989. The sites contained a diversity of both crowned Kinglet. conifer and deciduous(bare and newly leafed) trees (D) Flight features,museum skins (alula, pri- and shrubs,which permitted the speciesto express maries 10-5, secondaries 1 and 2, tail length, any feedingsubstrate and habitat-usepreferences. Data tail fork depth). There doesnot appear to be a on percentagesof foraging individuals that fed in generalsize factorrelating thesevariables. Fac- coniferousand deciduoustrees, foraging tactics, and tor ! is most highly correlatedwith the major locomotoryfeeding movementswere quantifiedby primaries,8-5, and tail length (i.e. with length the methods of Robinson and Holmes (1982) and of wing and tail as determined through feather Holmes and Recher (1986). The major problems associatedwith thesetypes of data are biasesintroduced length; Table 1D). The two speciesdo not ap- by the interrelatednessof sequentialfeeding move- pear to separateon this axis; both have an al- ments,inadequate spreading of observationsover time, most equal distribution (Fig. 1D). Factor 2, a and small sample sizes (Petit et al. 1990). We reduced measure of width of wing and tail shape, is these problems by standardizingdata-gathering to highly positivelycorrelated with secondaries! 20-40 s per bird, monitoring eachindividual or feed- and 2, and strongly negatively correlatedwith ing party only once,sampling both the morning and tail fork depth (and with the vestigial primary evening feeding periods, and extending data collec- 10). The speciesdo not overlap along this axis. tion throughout the entire 4-5-week period. The Golden-crowned Kinglet has a relatively deeper tail fork and shorter secondariesthan RESULTS the Ruby-crownedKinglet. (E) Bodylong axis, skeletal (fused vertebral col- Morphology.--The analyseswere conducted umn [i.e. thoracic, lumbar, and sacralvertebrae], separatelyfor each of six functional units as length of cervicalvertebrae [total], skull length follows. without bill, bill length). Factor ! is defined by (A) Axisof leg,skeletal (femur, tibiotarsus,tar- the thoracic vertebrae, cervical vertebrae, and sometatarsus,and middle toe [digit 3] without skull (Table 1E). Bill length loads in the same claw). Femur, tibiotarsus, and tarsometatarsus directionbut to a lesserextent. Factor 2 is highly 882 KEASTAND SAUNDERS [Auk, Vol. 108

TABLE1. Correlations(loadings) of morphologicalvariables measured for Ruby-crownedand Golden-crowned kingletswith orthogonallyrotated multivariate factors. (A) Axisof leg, skeletal;(B) leg andfoot dimensions, museumskins; (C) wing axis,skeletal; (D) flight features,museum skins; (E) body long axis,skeletal; and (F) bill morphology,museum skins.

Correlations with Correlations with orthogonalfactors orthogonalfactors (F) (F) Variable F-1 F-2 Variable F-1 F-2 (A) Axis of leg, skeletal (D) Flight features,museum skins Femur length 0.923 0.234 Alula 0.076 -0.229 Tibiotarsuslength 0.958 0.112 Primary 10 0.369 -0.643 Tarsometatarsuslength 0.966 -0.018 Primary 9 0.274 0.045 Mid. toe (digit 3) length 0.101 0.993 Primary 8 0.804 -0.276 Primary 7 0.793 0.213 % prop. variance explained 72.1 27.9 Primary 6 0.841 0.123 Primary 5 0.844 0.102 Secondary1 0.303 0.906 Secondary 2 0.234 0.938 Tail length 0.740 -0.124 Tail fork depth 0.171 -0.918 % prop. variance explained 43.4 37.9

(B) Leg and foot dimensions,museum skins (E) Body long axis, skeletal Tarsometatarsuslength -0.851 -0.144 Bill length 0.29 0.950 Mid. toe (digit 3) length -0.046 -0.038 Skull length 0.927 0.166 Mid. toe (digit 3) claw length -0.284 -0.011 Cervical vertebrae 0.866 0.402 Hind toe (hallux) length 0.687 -0.230 Thoracic, lumbar, & sacral Hind toe (hallux) claw length 0.882 0.044 vertebrae 0.851 0.437 Foot span -0.455 0.738 % prop. variance explained 65.3 34.7 Foot span with claw 0.227 0.900 % prop. variance explained 39.9 24.7 (C) Wing axis, skeletal (F) Bill morphology, museum skins Humerus length 0.903 -0.290 Bill length 0.701 -0.637 Ulna length 0.858 -0.350 Bill width 0.949 -0.251 Carpometacarpuslength 0.691 0.654 Bill depth -0.270 0.949 Phalanx 1 (of major digit) 0.892 0.104 % prop. variance explained 51.7 48.3 Phalanx 1 + 2 (of major digit) 0.815 0.022 % prop. variance explained 84.4 15.6

correlated with bill length. Both factors sepa- Table2 and below) showthat the Ruby-crowned rate the species(Mann Whitney U-test,P < 0.05; Kinglet has a longer leg; the Golden-crowned Fig 1E).The Ruby-crownedKinglet hasa longer Kinglet hasrelatively longer hind toesand hind bill and longer body axis. claws.The Golden-crownedKinglet hasshorter (F) Bill morphology,museum skins (total length skeletal wing elements, particularly humerus to back of nares, width and depth at back of and ulna.The Golden-crownedKinglet hasrel- nares).Both bill width relative to length (factor atively shorter secondariesand a relatively 1) and depth relative to length (factor2) clearly deeper tail fork than the Ruby-crowned.The separatethe kinglets (see also Table 2). Bill bill is shorter,narrower, and deeper than that length and width load strongly and positively of the Ruby-crownedKinglet. Three of the ma- on factor 1. Bill depth loadsnegatively and more jor differencesbetween the kinglet speciesare weakly. Factor2 is correlatedstrongly and neg- in the distal members of linear series: the tar- atively with bill length, but positively with bill sometatarsusand toe within the leg, distalskel- depth (Table 1F). The Golden-crownedKinglet etal wing elements(carpometacarpus, phalanx has a shorter, narrower, and deeper bill. 1, phalanx 1+2) relative to the humerus and Factoranalyses (and univariatecomparisons; ulna, bill length along the main body axis. October1991] KingletEcomorphology 883

A) 2 5 [] 1.5. ß [] 1. ß [] ß t• .5- []

iD iDø ß GCKI •O 0. ß ß ß O -.5- iD [] ß [] RCKI < ß [] iD ß • -1. ß [] ß -1.5-

-2.

-2.•c -2 -1.5 -1 -.5 0 .5 1 1.5 -2 -1.5 -1 -.5 0 .5 1 1.5 FACTOR 1 FACTOR 1

c) 2.5

2 1.5 ' D1.51 t oø2• •2o oø ß ß GCKI < 0 ß g -.5 . m . ORCKI ß ß ß -1 -1.5 [] -1.5 Z -1t.' ' -3 -2.5 -2 -1.5 -1 -.5 FACTOR 1 FACTOR 1

E) 3 h 2.5

t• 1.5'

O 1' ß GCKI O .5' Oo [] RCKI ß ß ooøo[] [] ß ID -1.5 [] -1.5 -2 -1.5 -1 -.5 0 .5 1 1.5 -2.5-2-1.5-1 -.5 0 .5 1 1.5 2 2.5 FACTOR 1 FACTOR 1 Fig. 1. Factor1 vs. Factor2 from factoranalysis of groupsof morphologicalvariables measured in the Ruby-crowned(RCKI) and Golden-crowned(GCKI) kinglets (for detailssee text). (A) Axisof leg,skeletal; (B) Legand foot dimensions, museum skins; (C) Wingaxis, skeletal; (D) Flightfeatures, museum skins; (E) Body long axis,skeletal; and (F) Bill morphology,museum skins.

In the Ruby-crownedKinglet, bill length, bill the wing, and wing loading, appearsimilar in width, rictal bristle length, and tarsuslength the two species. are significantly greater (P < 0.0001), and the The speciesdiffer strikingly in the structure body is significantlyheavier (P = 0.01)than the of the underside of the toes and soles of the Golden-crowned(Table 2). The bill of the Ruby- feet. In the Ruby-crownedKinglet theseareas crowned Kinglet is significantly shallower (P are relatively smooth, but in the Golden- < 0.0001). The hallux length without claw of crowned Kinglet the papillae of the pads are the Golden-crownedKinglet is significantly exaggeratedto form protruding tuberclessep- longer (P = 0.0360).Wing chordand tail length aratedby deep furrows(Fig. 2). The scalesare are not significantly different. Aspect ratio of alsolarger in the Golden-crowned. 884 KEAStANt• SAtSNt•ERS [Auk,Vol. 108

TABLE2. Externalmorphological features of Ruby-crownedand Golden-crownedkinglets, based on museum skins (œ+ SD) and live birds (œ[range]). The measurementsare basedon seriesof 10 adult male Ruby- crownedKinglets and 20 adult male Golden-crownedKinglets. Significance (*) was determinedby two- tailed unpaired t-tests.

Measurement Ruby-crownedKinglet Golden-crownedKinglet P Museum skin Wing chord length (mm) 58.6 + 1.51 58.0 + 1.33 0.3599 Tail length (mm) 43.6 + 0.40 43.7 + 0.47 0.8735 Bill length (mm) 8.35 + 0.08 6.55 + 0.10 0.0001' Bill width (behind nostril) (mm) 3.55 + 0.06 3.17 + 0.06 0.0002* Bill depth (mm) 2.17 + 0.04 2.63 + 0.06 0.0001' Rictalbristle length (mm) 4.2 + 0.41 3.7 + 0.31 0.0001' Tarsuslength (mm) 17.52+ 0.45 15.35+ 0.38 0.0001' Hallux length without claw (mm) 5.86 + 0.33 6.17 + 0.28 0.0360* Bodyweight (g) 6.6 + 0.35 6.3 + 0.52 0.0078* Alula (mm) 11.2 + 0.42 11.65 + 0.83 0.2507 Live bird Wing aspectratio 1.71(1.51-1.87) 1.68(1.51-1.82) Wing loading(g.cm -2) 0.13(0.11-0.15) 0.13(0.12-0.15)

Ecology.--Twosets of ecologicaldata (choice gle of movementbetween foraging sitesas well of substrate and locomotory feeding move- asduring prey attacks.The two kinglets differed ments) were analyzed for differencesbetween little in the upwards angle of feeding. How- the kinglet species. ever, the Golden-crowned used a higher per- (1) Choiceof substrate.Golden-crowned King- centage of angled flights in foraging overall, lets showeda strongpreference for feeding in with a pronounced downward-angle compo- conifers.Twenty of twenty-five (80%)feeding nent. The Ruby-crowned Kinglet exhibited a groupsat Little CataraquiCreek and twenty- higher percentageof non-angled upward and two of twenty-five(88%) at PrinceEdward Point horizontal flights. Flight distancesduring feed- were in conifers. The remainder were in low ing of the two speciesdid not differ signifi- deciduousshrubs. For the Ruby-crownedKing- cantly (X2 = 11.21, df = 8, P > 0.05), although let at Little Cataraqui(65 observations),80% of the Golden-crowned Kinglet tended to make the feedinggroups were in deciduousand 20% more very short flights. in coniferous trees. At Prince Edward Point (48 Searchand prey-attack rates (hops per min- observations),55% were in deciduoustrees and ute, flights per minute, and prey-attacks per 45%in conifers.This suggeststhat the Golden- minute) appearhighly variablefor eachspecies crownedKinglet is a coniferspecialist and the (Table 3). However, the Golden-crownedKing- Ruby-crownedmore of a generalistin substrate let hopped significantlymore frequently than choice (see Bent 1949). the Ruby-crowned (t-test, P < 0.05). (2) Locomotoryfeeding movements. The species differedsignificantly in proportionsof different DISCUSSION foragingtactics used (X 2 = 4.9,df = 4, P < 0.0001; Fig.3). Thoughfoliage gleaning predominated The Golden-crowned and Ruby-crowned in both species(39% of the Ruby-crownedand kingletsdiffer in a seriesof morphologicalfea- 42% of the Golden-crownedprey attacks),the tures and most of these can be linked to differ- Golden-crownedKinglet fed much more by ences in feeding behavior and substrateuse. clinging to or hanging from the foliage.The The Ruby-crowned Kinglet has a longer leg, Ruby-crownedKinglet had a morepronounced and relatively longer tarsometatarsus.The tar- aerial feeding componentand fed more com- sustends to be longer in birds that utilize rigid monly by flutteringat or pluckingprey from perches(Grant 1971),and a longertarsus prob- deciduoustrees. Flight anglesduring feeding ably favorsexploitation of a wider range of re- alsodiffered significantly (X 2 = 13.56,df = 4, P sources (Grant 1971). On their wintering = 0.01; Fig. 4). The horizontal componentpre- grounds, Ruby-crowned Kinglets show great dominatedin both species,being the main an- variability in use of foraging substrates(leaf, October1991] KingletEcomorphology 885

GOLDEN-CROWNED AERIAL

FOLIAGE GLEAN ß GCKI FOLIAGE • RCKI HANG GL

FOLIAGE HOVER GL

BRANCH GLEAN

0 10 20 30 40 50 R % Fig. 3. Foragingtactics, Ruby-crowned (RCKI) and Golden-crowned (GCKI) kinglets, Little Cataraqui Creek and Prince Edward Point, Ontario, April to May, 1989. For explanation of categoriessee text.

specializedfoot structurein the Golden-crowned Kinglet, which foraged predominantly in conifers. Fig.2, Structureofthe feet and toe pads of Ruby- The Ruby-crownedKinglet has a relatively crownedand Golden-crowned kinglets viewed from longer humerus than the Golden-crowned. theside and underside; drawn and enlarged from live Leislerand Thaler (1982)relate the relatively birds by Keast, longer humerusof the Firecrestto longer and more sustained flights than the Goldcrest.The Ruby-crownedKinglet hasa longer wing than bark, flowers,ground) but are fairly restricted the Golden-crownedthough neither measure- in size of perchesused (Laurenzi et al. 1982). mentsof wing chordnor lengthsof the major The tarsustends to be shorterin birdsthat hang primariesindicate this. Lengtheningof the hu- from slender branches(Palmgren 1932).These merusand ulna concurrentwith lengthening conclusions are consistent with the observation of the secondariesimply an increasein overall thatRuby-crowned Kinglets are less specialized wing size and area. The longer secondariesof in choicesof perching substratesand with the the Ruby-crownedKinglet (the prime airfoils prominenceof clinging in the Golden-crowned providinglift in birds)may function along with Kinglet. However, the tarsus,more than any the longer humerusand ulna to allow longer univariate measurement, is often an indicator aerial bouts.We found that aerial foraging is of body size (Rising and Somers1989, Freeman more important in this species.On the winter- and Jackson1990). The tarsusof the Ruby- ing grounds,Ruby-crowned Kinglets fed by crowned may also reflect its larger overall size. hovering25-35% of the time, and by hawking The relatively longer hind toe and hind claws 5-10%of the time; the proportionsvaried only of the Golden-crownedKinglet are presumably slightly over the season(Laurenzi et al. 1982). an adaptationfor clinging, assuggested for the Wing loadingsand aspectratio are no different European Goldcrest (Leisler and Thaler 1982) betweenthe species.The need for maneuver- and for the Coal Tit (Parusater) relative to the ability within vegetationduring foraging may Blue Tit (P. caeruleus)(Partridge 1976). Differ- constrainboth speciesto similar low aspectra- encesin the solesof the feet appearto be exact tiosand wing loadings(see Norberg 1986,Nor- counterpartsof those found in the Goldcrest bergand Norberg 1986, Rayner 1988). The alula, and Firecrest.Leisler and Thaler (1982) point which produceshigh lift at very high anglesof out that the specializedstructure of the Gold- attack (Nachtigall and Kempf 1971) is not lon- crest'sfeet permitsit to graspand cling from ger in the Ruby-crownedKinglet than in the pine needles.Presumably this alsoexplains the Golden-crownedKinglet (Table 2). The Euro- 886 Kr•.sz AND SAUNDERS [Auk, Vol. 108

A) 30 GOLDEN-CROWNED Generally, there is a positive correlation be- KINGLET tween size of bill and size of prey consumed (Hespenheide 1973, Lederer 1975, Smith and 20 Zach 1979), and a larger bill favors taking prey % 9 n=156• of a wider range of sizes (Grant 1968, Herrera 3 47 1978).However, Hespenheide(1971) found that, 10 in the Tyrannidae, body size predictsprey size better than bill size does, and Wiens and Re- tenberry (1980) failed to find a correlation be- 1.0 1.8 2.7 3.6 tween bill and prey sizein grasslandbirds. There •=0.8• 0.8 DISTANCE (m) is a direct relationship between bill length and speed at which the mandibles can be closed (Beechef 1962); longer-billed birds should capture fast-movingprey more readily. The greater

B) 30 - RUBY-CROWNED KINGLET force exerted by the longer bill may also im- prove handling of larger prey (Beck 1964, Led- erer 1975). Most of these correlations remain untestedfunctionally. Thaler (1988,pets. comm.) % =200 •, found no difference in sizesof prey consumed by the Ruby-crowned and Golden-crowned 10' II 3'• 3 54 kinglets, but the Ruby-crowned prefers flying insects and the Golden-crowned prefers soft-

0 bodied prey from the substrate. 0.3 1.0 1.8 2.7 3.6 >5 Aerial-feeding birds characteristicallyhave

X=l.0ñ 0.9 broadenedbills and long rictal bristles.Leisler DISTANCE (m) and Winklet (1985) showed positive correla- Fig. 4. Flight distancesand angles when feeding tions of bill width and bill length with rictal in Ruby-crownedand Golden-crownedkinglets, April bristle length. Rictal bristleshave been thought to May, 1989. to function as a funnel that gives birds larger and more effective gapes (Lederer 1972, Stet- tenhelm 1974), as protection for the eyes from pean Regulusdiffer in this feature (Leisler and food items the bird is trying to capture (Con- Thaler 1982). over and Miller 1980) or as a sensoryfunction The tails of small passefine birds are used (Stettenheim 1974). Lederer (1972) was unable mainly for steering,and longer outer tail feath- to establishthat they aid in prey capture in ers are particularly important. Phylloscopusspe- tyrannid flycatchersthat capturethe prey with cies,especially those that take prey on the wing, the tip of the bill, and Conover and Miller (1980) tend to have short central feathers and longer found that taping the bristles down did not outer ones (Gaston 1964). Tail length is not sig- impede prey capture. Nevertheless, we suggest nificantly different betweenthe kinglet species that the longer and broader bill and longer ric- (Table 2), but the tail of the Golden-crowned tal bristles of the Ruby-crowned Kinglet are Kinglet is more forked. This seemsanomalous linked functionally with the demonstrated relative to our findings that the Ruby-crowned greater emphasison aerial feeding in this spe- Kinglet is the predominantaerial feeder. How- cies. ever, forking of the Golden-crownedKinglet's In many featuresthe Ruby-crowned Kinglet tail may impart greater maneuverability, re- and Firecrest,and Golden-crownedKinglet and flected in the higher percentageof angled for- Goldcrest,are counterparts.Presumably the uti- aging flights in this species.In Europe, the lization of comparableecological opportunities Goldcresthas the more forked tail, which may involved the humerus, rictal bristles, bill width, be related to different modes of hovering (Leis- structure of the feet, claws, and soles of the feet let and Thaler 1982). Goldcrests make short in similar ways. In contrast,other features have flights to hover in front of twig tips, whereas evolved differently and the Ruby-crowned Firecrestsswitch from one twig to another, cov- Kinglet has a significantlylonger bill and lon- ering greater distances. ger tarsus than its North American congener. October1991] KingletEcomorphology 887

TABLE3. Searchand prey-attack rates (œ + SD)of Ruby-and Golden-crowned kinglets. The numberof feeding sequences(n) andthe total numberof minutesduring which feedingwas observed are given; * = significant t-test (a -< 0.05).

Species n Total min Hops/min Flights/min Preyattacks/min Ruby-crownedKinglet 49 16.3 25.1 + 14.5' 9.3 + 6.6 9.3 +_8.2 Golden-crownedKinglet 30 10.0 32.5 + 11.8 10.3+ 4.7 9.4 + 3.8

We cannot determine if the two North Amer- BOOKSTEIN,F. B., B. CHERNOFF,R. ENDLER,J. HUMPHRIm, ican Regulusare representativeof the two Eur- G. SMITH,,e,r R. STRAUSS.1985. Morphometrics asianstocks, or if they were derived by double in evolutionary biology. Acad. Nat. Sci. Phila- invasion of a single one. When comparedwith delphia, Spec.Publ. 15. CONOvER, M. R., & D. E. MILLER. 1980. Rictal bristle the data of Leisler and Thaler (1982), each North function in the Willow Flycatcher. Condor 82: American speciesappears to be a partial eco- 469-471. morphologicalequivalent of one of the Eur- FREEMAN,S., & W. M. JACKSON.1990. Univariate met- asian ones. Mayr (1956) pointed out the dis- rics are not adequateto measureavian body size. tinctivenessof the Ruby-crowned Kinglet and Auk 107: 69-74. its strong differencesfrom both Eurasianspe- GASTON,A. J. 1964. Adaptation in the genus Phyl- ciesin many features,as well as from the Gold- loscopus.Ibis 116: 432-450. en-crownedKinglet. We presumethis indicates GRANT,P.R. 1968. Bill size, body size, and the eco- a long historyin North America.Thaler (1990a, logical adaptationsof bird speciesto competitive b; pets. comm.) has found that in the sum of situationson islands.Syst. Zool. 17: 319-333. 1971. Variation in the tarsal length of birds behavioral patterns, postembryonic develop- in island and mainland regions. Evolution 25: ment, and postembryonicvocalizations, both 599-614. Americankinglets are more closelyrelated to 1986. Ecologyand evolution of Darwin's the EuropeanFirecrest than the Goldcrest.This Finches.Princeton, New Jersey,Princeton Univ. couldindicate that originallyboth were derived Press. from the ancestors of modern Firecrest stocks. HERRERA,C. M. 1978. Individual dietary differences associatedwith morphologicalvariation in rob-

ACKNOWLEDGMENTS ins (Erithacusrubecula). Ibis 120: 542-545. HESPENHEIDE,H.A. 1971. Food preference and the This researchwas funded by the Natural Sciences extent of overlap in some insectivorousbirds, and EngineeringResearch Council of Canada,and we with specialreference to the Tyrannidae.Ibis 113: thank this body for its support.Queen's University 59-72. BiologicalStation, Chaffey's Lock, Ontario, provided 1973. Ecologicalinferences from morpho- accommodation and research facilities. Skeletal ma- logical data. Annu. Rev. Ecol. Syst. 4: 213-229. terial and museum skins were from the collections of HOLMES, R. T., & H. F. RECHER. 1986. Search tactics the Royal Ontario Museum, Toronto, and National of insectivorousbirds foraging in an Australian Museum of Canada, Ottawa. We thank these insti- eucalypt forest. Auk 103: 515-530. tutions for encouragementand trusting us with their LAURENZI, A. W., B. W. ANDERSON,& R. W. OHMART. material.We acknowledgethe help of Tim Ehlinger 1982. Wintering biology of Ruby-crownedKing- in the developmentof the multivariateanalyses. We lets in the Lower ColoradoRiver Valley. Condor thank CharlesFrancis, who let us handle birdsduring 84: 385-398. his banding work of Prince Edward Point, Ontario. LEDERER,R.J. 1972. The role of avian rictal bristles. Finallywe thankB. Leisler and E.Thaler for agreeing Wilson Bull. 84: 193-197. to read this manuscript. 1975. Bill size, food size, and jaw forcesof insectivorous birds. Auk 92: 385-387.

LITERATURE CITED LEISLER,B., & E. THALER. 1982. Differences in morphologyand foragingbehaviour in the Goldcrest BEECI-IER,W. J. 1962. The biomechanics of the bird Regulusregulus and FirecrestR. ignicapillus.Ann. skull. Bull. ChicagoAcad. Sci. 11: 10-33. Zool. Fennici 19: 277-284. BENT, A. C. 1949. Life histories of North American --., & H. WINKLER. 1985. Ecomorphology.Curr. thrushes,kinglets, and their allies.U.S. Natl. Mus. Ornithol. 2: 155-186. Bull. MAYR,E. 1946. History of the North Americanbird BOCK,W.J. 1964. Kinetics of the avian skull. J. Mor- fauna. Wilson Bull. 58: 3-41. phology 144: 1-42. MOSIMANN,J. E., & F. C. JAMES.1979. New statistical 888 KE•STAND SAUNDERS [Auk,Vol. 108

methodsfor allometry with applicationto Florida ficationof avianforaging behaviour. Pp. 193-198 Red-winged Blackbirds. Evolution 33: 444-459. in Avian foraging:theory, methodology, and ap- NACHTIGALL,W., & B. KEMPF.1971. Vergeleichende plications(M. L. Morrison,C. J.Ralph, J. Verner, Untersuchungenzur fliigbiologischenFunktion and J. R. Jehl Jr., Eds.). Cooper Ornithol. Soc., des Daumenfittichs (Alula spuria) bei V•Sgeln. Stud. Avian Biol. No. 13. Zeitschr. Vergl. Physiol. 71: 326-341. RAYNER,J. M. V. 1988. Form and function in avian NORBERG,U.M. 1986. Evolutionary convergencein flight. Curr. Ornithol. 5: 1-65. foraging niche and flight morphologyin insec- RISING, J. D., & K. M. SOMERS. 1989. The measure- tivorous aerial-hawking birds and bats. Ornis ment of overall body sizein birds.Auk 106:666- Scandinavica 17: 253-260. 674. , & R. A. NORBERG.1986. Ecomorphology of RoBitasota,S. K., & R. T. HOLMES.1982. Foraging flight and tree-trunkclimbing in birds.Pp. 2271- behaviourof forestbirds: the relationships among 2282 in Proc. 19th Int. Ornithol. Congress(H. A. searchtactics, diet, and habitat structure. Ecology Ouellet, Ed.). Ottawa, Univ. Ottawa Press. 63: 1918-1931. OXNARV,C. E. 1973. Form and pattern in human SMItM,J. N.M., & R. ZACn. 1979. Heritabilityof evolution. Some mathematical,physical and en- somemorphological characters in a SongSpar- gineering approaches.Chicago, Univ. Chicago row population. Evolution 33: 460-467. Press. StETrmWI-IEIM,P. 1974. The bristlesof birds.Living 1975. Uniquenessand diversity in human Bird 12: 201-234. evolution: morphometric studies of Australo- THALER,E. 1990a. EthodSkologischeDifferenzierun- pithecines.Chicago, Univ. ChicagoPress. genbei amerikanischengoldenh•ihnchen: Regu- PALMGREN,P. 1932. Zur Biologievon Regulusr. reg- lussatrapa, Regulus calendula. Proc. Int. 100 DO-G ulus(L) und Parusatricapillus borealis Selys. Eine meeting. Curr. Topics Avian Biol., Bonn 1988: vergleichend-okologischeUntersuchung. Acta 373-382. Zool. Fennici 14: 1-113. 1990b. Eine spannendeErstzucht: das amer- PARTRIDGE,L. 1976. Someaspects of the morphology ikanische"Robinkr•Snchen" Regulus calendula. Gef. of Blue Tits (Paruscaeruleus) and Coal Tits (Parus Welt. 114 (May): 134-139. ater) in relation to their behaviour. J. Zool. 179: WIFaWs,J. A., & J. T. ROTm'WBERR¾.1980. Patterns of 121-133. morphologyand ecology in grasslandand shrub- PETI?, L. J., D. R. PE•?, & K. G. SMIT•. 1990. Preci- steppebird populations.Ecol. Monogr. 50: 287- sion,confidence, and samplesize in the quanti- 308.