DISTINGUISHING MORPHS OF THE WHITE-THROATED SPARROW IN BASIC PLUMAGE

j. Field Ornithl., 6(1):73-83 DISTINGUISHING MORPHS OF THE WHITE-THROATED SPARROW IN BASIC PLUMAGE WALTER H. PIPER AND R. HAVEN WILEY Department f Bilgy University f Nrth Carlina Chapel Hill, Nrth Carlina 2799-328 USA Abstract.--The White-thrated Sparrw (Zntrichi albicllis) has tw genetic mrphs, which are clearly distinguishable by clr in alternate plumage, but less distinguishable in basic plumage. T determine the accuracy with which the mrphs can be identified in basic plumage in a wintering ppulatin in Nrth Carlina, we measured the brightness (extent f pure white r black as ppsed t tan r brwn) in the median, lateral, and superciliary stripes f 279 birds f knwn age and sex and determined each bird's mrph after the prealternate mlt in the spring. Age, sex, and plumage all influenced the clr f crwn stripes in bth mrphs. Hwever, white mrphs were significantly brighter than tan mrphs in all age- and sex-classes in basic plumage. A simple frmula based n a bird's age, sex, and brightness f its median and lateral stripes identified tan and white mrphs in basic plumage with a success rate f 89%. In this wintering ppulatin, an estimated 61% f males and 49% f females were white mrphs. LA IDENTIFICACI N DE FENOTIPOS PARTICULARES EN INDIVIDUOS DE ZONOTRICHI ALBICOLLIS EN PLUMAJE B/i, SICO Resumen.--En el grri6n Zntrichialbicllise ban descrit ds fentips (mrphs) particulares que pueden distinguirse fftcilmente pr su clr cuand las aves ban mudad a su plumaje altern. N bstantes dificil diferenciar entre las ds frmas, cuand las aves est/tn cn su plumaje b tsic. Para determinar cn que exactitud pdlams identificar las ds frmas en su plumaje b tsic, estudiams pr afts una pblaci6n de grrines que pasan el invicrn en Carlina del Nrte. De 279 individus de edad y sex cncids, medims la brillantez (l extensiv del blanc negr en cntraste cn el acanelad pard) de las lineas medias, laterales y superciliares durante la primavera cuand las aves ya hablan adquirid su plumaje prealtern. Se encntr6 que la edad, sex y el plumaje influyen en el clr de las bandas de la crna en ambs fentips. N bstanten aves cn plumaje b tsic, las frmas blancas resultarn m ts brillantes que las frmas acaneladas, sin imprtar la edad sex de ls grrines. Tmand en cnsideraci6n la edad de las aves, sex y brill de las bandas medias y laterales pudims identificar en grrines cn plumaje b/tsic el 89% de ls fentips acanelads blancs. En la pblaci6n estudiad aprximadamentel 61% de ls machs y el 49% de las hembras resultarn ser frmas blancas. Lwther (1961) first described tan- and white-striped mrphs in bth sexes f the White-thrated Sparrw (Zntrichialbicllis) in alternate plumage. He distinguished the tw types, which fr brevity we shall call tan and white mrphs, by the whiteness f the median crwn stripe and als nted that they differed in the clratin f the lateral and superciliary crwn stripes and in the amunt f streaking n the breast. Thrneycrft (1966, 197) fund a chrmsmal basis fr this difference in that white birds have a distinctive autsme absent in tan birds. The tw mrphs have subsequently been reprted t differ in a number f behaviral and Current address: Department f Bilgical Sciences, Purdue University, West Lafayette, Indiana,1797 USA. 73

74] w. H. Piper and R. H. Wiley j. Field Ornithl. Winter 1989 eclgical aspects, bth during the breeding seasn and during the winter (Ficken et al. 1978; Knaptn and Falls 1982, 1983; Lwther 1961; Lwther and Falls 1968; Watt et al. 1984). The difficulty f distinguishing white and tan mrphs in basic plumage was first dcumented by Vardy (1971). Mre recently, Atkinsn and Ralph (198) and Watt (1986) measured plumage clratin in grups f White-thrated Sparrws held in aviaries during winter and spring. They shwed that mst measures f clratin are bimdally distributed amng birds in alternate plumage, but unimdally distributed in basic plumage. These results suggested that the mrphs are nt separable in basic plumage. The present study, part f a -yr prject t examine the behavir f a wintering ppulatin f individually banded White-thrated Sparrws, examined differences in plumage clratin in a large sample f resident, free-living birds f knwn age, sex, and mrph. Our results indicate that the mrphs can in fact be identified in basic plumage with nearly 9% accuracy. METHODS Basic prcedures.--white-thrated Sparrws were studied at the Masn Farm Bilgical Reserve in Chapel Hill, Nrth Carlina, where they are abundant frm Nvember thrugh April. During the winters f 1983-1984 thrugh 198-1986, we trapped and clr-banded 3-4 Whitethrated Sparrws a year in 4 ha f thickets and hedgerws at this site. When each bird was first captured, we checked its skull fr ssificatin, sexed it by lapartmy, and measured its unflattened wing chrd. Previusly unbanded birds captured n r befre 1 Dec. with fully ssified skulls were cnsidered t be mre than ne year ld (AHY r afterhatching-year birds), and thse with incmpletely ssified skulls were cnsidered t be in their first year (HY r hatching-year birds). Birds first captured after 1 Dec. were excluded frm this analysis. Scring f crwn stripes.--inspectin f birds at the utset f the study revealed that the median and superciliary stripes in many birds changed frm pure white near the beak t tan r gray psterirly, whereas the lateral stripes in many birds began as pure black near the beak and became flecked with brwn psterirly. On the brightest birds, these transitin pints ccurred well back n the crwn, whereas the stripes f dull birds were ften tan, gray, r brwn-flecked at their anterir ends. Hence all birds had tan and brwn in their crwn stripes, but varied in the extent f pure white r black. Based n these bservatins, we scred each bird fr the extent f pure black r white in its median, lateral, and superciliary crwn stripe with a handmade scale ( Fig. 1) that measuredistance alng the midline f the skull frm the anterir edge f the nares (scre = ) t a pint halfway dwn the back f the skull (scre = 8). All scres were truncated t integer values. Birds with high scres fr each stripe had extensive white

Vl. 6, N. I Plymrphism in White-thrated Sparrws [7 ¾ FIGURE 1. Scale used t measure brightness f stripes n the crwns f White-thrated Sparrws. The zer pint ("N") is aligned with the anterir edge f the naris. in the median and superciliary stripes and extensive black in the lateral stripes. Stripes were scred independently by tw bservers fr every bird. Inter-bservereliability was examined fr each stripe in a sample f 389 birds measured by the tw authrs (fr ther birds an assistant made ne f the tw measurements), by determining the prprtin f birds in which the scres differed by n mre than ne unit. The reliability f scres fr the median and lateral stripes exceeded 91%, fr the superciliary stripe 94%. Determinatin f mrph in alternate plumage.--many birds finished the prealternate mlt f the crwn befre leaving the study area in April and May. The earliest date fr the nset f mlt n the crwn (ne grwing feather) was 17 Mar.; the mean dates f nset were 26 Mar. (SD = 3.9 d, n = 2) fr males and 29 Mar. (SD = 4.4 d, n = 69) fr females. On each day frm 7 t 2 Apr., mre than half f the birds trapped were in heavy mlt (at least 2 grwing crwn feathers and usually many mre). Mean dates fr the final stage f mlt (fewer than 2 grwing feathers) were 21 Apr. (SD = 3.2, n = 26)and 24 Apr. (SD = 3.2, n = 36) fr males and females respectively. Frm 1984 thrugh 1986, we determined the mrph f birds captured between 17 Apr. and May, during r after the final stage f mlt, when mst birds were in fresh alternate plumage. Individuals were classified as white mrphs if the median and superciliary stripes were mstly pure

76] w. H. Piper and R. H. Wiley J. Field Ornithl. Winter 1989 MEDIAN LATERAL SUPERCILIARY I i i i i i i i 6O FEMALES (N = 132) 4O 2O 4 8 12 16 4 8 12 16 4 8 12 16 uj 6 m MALES (N= 8) Z 4O 2O 4 8 12 16 4 8 12 16 4 8 12 16 FIGURE 2. Distributins f scres fr median, lateral and superciliary stripes in alternate plumage. white and their lateral stripes mstly black with few r n brwn flecks. They were classified as tan mrphs if their median and superciliary stripes were mstly tan r gray and their median stripes had many brwn flecks (see Knaptn and Falls 1982, 1983; Lwther 1961; Lwther and Falls 1968; Thrneycrft 1966, 197). Of the 134 birds captured during the study, 23 (24%) were recaptured after the prealternate mlt. Birds that were captured befre and after prealternate mlt in mre than ne winter (n = 26) were treated as separate bservatins fr each winter, s the ttal sample size was 279. Inclusin f these 26 bservatins did nt alter the results. Predicting mrph frm basic plumage.--having classified a sample f birds as either white r tan mrph in alternate plumage, we entered the scres fr each f these birds' crwn stripes in basic plumage and its sex, age, and wing chrd int a stepwise discriminant analysis (SYSTAT, Inc., Evanstn, Illinis; Kleinbaum and Kupper 1978), in rder t determine which cmbinatin f these measures was the best predictr f mrph. RESULTS Identificatin f mrphs.--of the sample f 279 birds captured after prealternate mlt, we culd easily identify the mrph f 269 (96%) by their plumage. Bth sexes had strngly bimdal distributins f scres fr the median stripe, as described by Lwther (1961), whereas scres frm the superciliary stripe were strngly bimdal amng males nly (Fig. 2). The scres fr the lateral stripe had a unimdal distributin fr females and nearly s fr males (Fig. 2). Ten birds (4%) intermediate in

Vl. 6, N. 1 Plymrphism in White-thrated Sparrws [77 MEDIAN LATERAL SUPERCILIARY I I I i i i, i [ 1 f, 4 8 12 16 4 8 12 16 4 8 12 16 2 HY MALES (N=282) O3 1 4 8 12 16 4 8 12 16 4 8 12 16 2 AHY FEMALES (N=234) - 4 8 12 16 4 8 12 16 4 8 12 16 2f AHY MALES (N =248) 4 8 12 16 4 8 12 16 4 8 12 16 FIGURE 3. Distributins f scres fr median, lateral and superciliary stripes in basic plumage. clratin culd nt be assigned t a mrph, in sme cases because mlt was incmplete at the time f examinatin; these birds were excluded frm the discriminant analysis. In basic plumage, scres fr the crwn stripes were in general distributed unimdally (Fig. 3), althugh amng males scres fr the median stripe had slight bimdality. Since sample sizes were large and the distributins f scres in many cases were clse t nrmal, tw-tailed t-tests are used in the fllwing sectins t cmpare differences in scres fr each mrph in each categry f age, sex, and plumage. Because f the large number f tests, we chse a =.1 as ur criterin fr statistical significance. See Figure 4 fr the means, standard deviatins, and sample sizes fr all tests. Fr cnvenience we use "brightness" t refer t the extent f pure white r black in the crwn stripes.

78] W. H. Piper and R. H. Wiley J. Field Ornithl. Winter 1989 HY FEMALES BASIC HY MALES BASIC AHY AHY ALL ALL FEMALES MALES FEMALES MALES BASIC BASIC ALTERNATE ALTERNATE N = 31 38 11 31 24 3 21 7 62 7 2 6 16 ß TAN MORPH 12 [] WHITE MORPH MEDIAN STRIPE O m 16 I 12 LATERAL STRIPE, w 8 4 16 SUPERCILIARY STRIPE 12 8 4 FIGURE 4. Means and standardeviatins f scres fr crwn stripes fr each age-sex class in basic and alternate plumages. Brightness in basic plumage in relatin t mrph.--within age- and sexclasses in basic plumage, white mrphs averaged brighter than tan mrphs. Their median and superciliary stripes were significantly whiter and their lateral stripes significantly blacker in all fur age- and sex-classes (12 tests in all). Brightness in relatin t age.--in basic plumage, amng white mrphs f bth sexes, AHY birds had all three stripesignificantly brighter than HY birds (6 tests in all). Amng tan mrphs, althugh AHY birds had brighter stripes n average than HY birds, the differences were significant nly fr the lateral stripe in males (6 tests). In alternate plumage, there were n significant differences in brightness between age-classes fr either sex r mrph (12 tests). Cnsequently, we cmbined age-classes f birds in alternate plumage fr the remainder f the analysis. Brightness in relatin t sex.--in basic plumage, amng white mrphs,

Vl. 6, N. 1 Plymrphism in White-thrated Sparrws HY FEMALES (N=69) [79 1 1 I ß TAN MORPH [] WHITE MORPH 2. 2. 1. 1.. -. -1. HY MALES (N-42) 1 lo 2. 2. 1. lo.o. -. -lo.o AHY FEMALES (N =81) I 1 z lo 2. 2. 1. 1.. -. -1. 1 AHY MALES (N=77) 1 2. 2. 1. 1.. -. -1. FIGURE. Distributins f d frm the discriminant functin fr birds f knwn mrph separated by age-sex class. The dashed line indicates the value fr best discriminatin between mrphs. males f bth age-classes were brighter n average than crrespnding females (6 tests). Amng tan mrphs, althugh males were at least slightly brighter than females n average fr all three stripes in bth age-classes, males differed significantly frm females nly fr the lateral stripe amng AHY birds (6 tests). In alternate plumage, with age-classes pled, white males had all three stripesignificantly brighter than white females (3 tests). Tan males were significantly brighter than tan females in the median and lateral stripes, but nt in the superciliary stripe (3 tests).

8] W. H. Piper and R. H. Wiley j. Field Ornithl. Winter 1989 1 1 HY FEMALES (N=69),, ',, ß TAN MORPH [] WHITE MORPH 1. 2. 3. 4.. 6. Z.O 1 HY MALES (N=42) lo 1. 2. 3. 4.. 6. 7. c uj 1 AHY FEMALES I (N-81) lo 1. 2. 3. 4.. 6. 7. AHY MALES (N=7'7) lo 1. 2. 3. 4.. 6. 7. FIGURE 6. Distributins f X frm Frmula (1) fr birds f knwn mrph separated by age-sex class. The dashed line indicates the value fr best discriminatin between mrphs. Brightness in relatin t plumage.--white mrphs in all age- and sexclasses had all three stripes significantly brighter in alternate plumage than in basic plumage (12 tests in all). Althugh tan mrphs f all ageand sex-classes averaged brighter stripes in alternate plumage than in basic plumage, the differences were significant nly fr the superciliary stripe f HY females and the lateral and superciliary stripes f HY males (12 tests in all). Predictin f mrph frm measurements f basic plumage.--discriminant analysis shwed that brightness f the lateral stripe, brightness f the median stripe, age, and sex were the fur best predictrs f mrph during the winter. Scres fr the superciliary stripe did nt cntribute t the

Vl. 6, N.! Plymrphism in White-thrated Sparrws [81 discriminant functin. With r 2 =.36 and n = 269, the discriminant functin was: d = 16.9-1.2M - 1.41L + 1.87A + 1.8S +.18ML -.228MA where M is the scre fr the median stripe, L is the scre fr the lateral stripe, A is a scre indicating age (1 fr HY, 2 fr AHY), and S is a scre indicating sex (1 fr female, 2 fr male). The gal f discriminant analysis t select a value fr d that permits classificatin f bservatins with the fewest errrs (Kleinbaum and Kupper 1978). In ur case, d -- 1 prduced the fewest errrs in classifying mrphs. Althugh ur verall success in predicting mrphs was then 87%, the success with which this functin predicted mrph varied with ageand sex-class. The functin crrectly assigned the mrph f HY females in 77% f cases, but classed HY males, AHY females and AHY males with success rates f 93, 91, and 93% respectively (Fig. ). Use f different discriminant functins fr each age- and sex-class did nt imprve success. Several simpler frmulas, invlving nnlinear cmbinatins f ur measures fund by trial and errr, distinguished mrphs as well as r better than the discriminant functin. Fr instance, the fllwing frmula classified 89% f the birds in ur sample crrectly: X = (M + L)/(A + S). (1) Birds with X < 2. were classified as tan mrphs, thse with higher scres as white mrphs. As in the case f the discriminant functin, the success f this frmula in predicting mrph varied with age- and sexclass. HY females, HY males, AHY females, and AHY males were crrectly classified in 78, 93, 92, and 94% f cases respectively (Fig. 6). If data n sex are nt available, then a mdified frmula can be used t predict mrph with an verall accuracy f 88% when applied t ur sample: Y= (M + L)/A. (2) Birds with Y < 1. were classified as tan mrphs, thse with higher scres as white mrphs. The success rates f this frmula fr HY females, HY males, AHY females, and AHY males, respectively, were 77, 9, 94 and 92%. A third frmula can be used when bth age and sex are unknwn: Z = M + L. (3) Birds with Z <. were classified as tan mrphs, thse with higher scres as white mrphs. This frmula predicted mrph in ur sample with an verall accuracy f 8% and assigned HY females, HY males, AHY females, and AHY males t mrphs with 68, 9, 91, and 88% SUCCESS. DISCUSSION A number f different techniques have been used t measure plumage variatin in White-thrated Sparrws. Lwther (1961) classified mrphs

82] w. H. Piper and R. H. Wiley J. Field Ornithl. Winter 1989 by cmparing birds' median stripes with a graded series f museum specimens. He used a similar prcedure t scre variatin in ther features f plumage as well. Vardy (1971) als used a series f museum specimens but based her classificatin f mrphs n bth median and lateral stripes. Atkinsn and Ralph (198) and Watt (1986) investigated variatin in a number f individual measures and cmpsite indices f plumage clratin. In the present study, fr simplicity and repeatability we fcused n the crwn stripes in classifying mrphs, althugh we recgnize that the clratin f the lres, thrat, and chest als varies. Our results in general cnfirm previus reprts that crwn clratin is bimdally distributed in alternate plumage and unimdally distributed in basic plumage. Results frm ur study demnstrate that the shift in the distributin ccurs because white mrphs--especially first-year birds and lder females--tend t increase markedly in brightness fllwing the prealternate mlt. Previus authrs have cncluded that the white mrph is nt phentypically expressed in basic plumage by immature and sme adult birds (Atkinsn and Ralph 198, Thrneycrft 197). Hwever, ur data shw that HY males differ enugh in brightness that they can be assigned t mrph with almst the same accuracy as can AHY males and females (see Figs. and 6). Even HY white females are bright enugh in relatin t their tan cunterparts t permit recgnitin f mrph in 78% f cases. Our results thus shw that mrphs f bth sexes and age-classes can be identified with cnsiderable accuracy in basic plumage. Fr males and adult females the accuracy exceeds 9%. Even withut infrmatin abut age and sex, we fund an verall accuracy f 8% in identifying mrphs in basic plumage. Thus a White-thrated Sparrw encuntering a strange cnspecific during the winter culd cnceivably determine the stranger's mrph in abut ut f 6 cases by its crwn stripes alne. If previus experience with a cnspecific prvided sme infrmatin abut its prbable sex r age, r if birds culd discriminate ther differences between mrphs, even greater success wuld be pssible. By using frmula (1), (2) r (3) and scres fr plumage stripes, it is pssible t estimate the prprtin f each mrph wintering in ur study area. In ding s it is necessary t take int accunt bth crrect and incrrect classificatins f the frmulae. Frmula (1), fr example, misclassifies 11% f the birds and is biased slightly twards the tan mrph (see Fig. 6). This bias must be crrected within each age- and sex-class. Fr example, changing 3.6% f the HY females (2. birds fr the sample shwn in Fig. 6) frm tan t white mrphs crrects the bias in this class. The prprtins f white mrphs amng winter residents (birds with a minimum interval f 6 d between their first and last captures in a winter) at ur study site, as estimated by this methd, are.61 (237/ 391) amng males and.49 (16/336) amng females. Lwther (1961), based n a survey f museum specimens frm breeding ppulatins, cncluded that the prprtins f white mrphs amng males and females varied inversely acrss Canada, with white males and tan females less

Vl. s, N. 1 Plymrphism in White-thrated Sparrws [83 frequent in the east. At ur study site during winter, the prprtin fr males crrespnds t Lwther's value fr breeding males in central Canada, while the prprtin fr females crrespnds t his value fr breeding females in Newfundland. Additinal research is needed t check Lwther's figures, especially since the mrphs are nw knwn t differ in breeding habitat (Knaptn and Falls 1982). In additin, the frmulas presented here prvide a means fr banders t cllect infrmatin abut the gegraphical distributin f mrphs during winter. ACKNOWLEDGMENTS We thank D. Westneat, M. Archawarann, B. Bernier, J. Perry, L. Dve, A. Ayers, and M. Wiley fr help in the field. H. Mueller, D. Westneat, C. Jhn Ralph and Dris Watt prvided helpful suggestins the manuscript. This research was supprted in part by a grant frm the Frank M. Chapman Fund f the American Museum f Natural Histry and is a cntributin frm the Behaviral Research Statin in the Masn Farm Bilgical Reserve f the Nrth Carlina Btanical Garden. LITERATURE CITED ATKINSON, C. T., AND C. J. RALPH. 198. Acquisitin f plumage plymrphism in White-thrated Sparrws. Auk 97:24-22. FICKEN, R. W., M. S. FICKEN, AND J.P. HAILMAN. 1978. Differential aggressin in genetically different mrphs f the White-thrated Sparrw Zntrichia albicllis. Z. Tierpsychl. 46:43-7. KLEINBAUM, D. G., AND C. C. KUPPER. 1978. Applied regressin analysis and ther multivariate methds. Duxbury Press, Nrth Scituate, Massachusetts. KNAPTON, R. W., AND J. B. FALLS. 1982. Plymrphism in the White-thrated Sparrw: habitat ccupancy and nest-site selectin. Can. J. Zl. 6:42-49. --, AND. 1983. Differences in parental cntributin amng pair types in the plymrphic White-thrated Sparrw. Can. J. Zl. 61:1288-1292. LOWTHER, J. K. 1961. Plymrphism in the White-thrated Sparrw, Zntrichialbicllis. Can. J. Zl. 39:281-292., AND J. B. FALLS. 1968. White-thrated Sparrw. Pp. 1364-1392, in A. C. Bent and O. L. Austin, Jr., eds. Life histries f Nrth American cardinals, grsbeaks, buntings, twhees, finches, sparrws and allies. U.S. Nat. Mus. Bull. 237. THORNEYCROFT, H.B. 1966. Chrmsmal plymrphism in the White-thrated Sparrw, Zntrichia albicllis. Science 14:171-172. 197. A cytgenetic study f the White-thrated Sparrw, Zntrichialbicllis. EvOlutin 29:611-621. V^}tDY, L.E. 1971. Clr variatin in the crwn f the White-thrated Sparrw, Zntrichia albicllis. Cndr 73:41-414. WATT, D. J. 1986. Plumage brightness index fr White-thrated Sparrws. J. Field Ornithl. 7:1-113. --, C. J. RALPH, AND C. T. ATKINSON. 1984. The rle f plumage plymrphism in dminance relatinships f the White-thrated Sparrw. Auk 11:11-12. Received 13 Jan. 1988; accepted 28 Jul. 1988.