sign in sign up
<<
Home , Bucco, Callipepla, Flammulated flycatcher, Hypnelus, Notharchus, Old World quail

24 PAGEET AL. [Auk, Vol. 100

Bird Observatory. Ralph Widrig, Steven Cellman, tain Plover social systems.Living 12: 69- Barbara Swarth, Susan Peaslee, Erica Buhrman, Don- 94. na Vaiano, Paul Neal, Sharon Goldwasser, Dennis HAYS, H., & M. LECI•o¾. 1971. Field criteria for Parker, John Harris, David Shuford, Stuart Johnson, determining incubation stage of the Common John and Ricky Warriner, Stephanie Zeiler, Elliot . Wilson Bull. 83: 425-429. Burch, Brett Engstrom,Leslie Smith, BernadetteAl- HOLMES,R.T. 1966. Breeding ecologyand annual len, Greg Calllet, Tom Pogson,Tom Cassidy, Geoff cycle adaptationsof the Red-backedSandpiper Geupel, Doug Burgess,Judy Wagner, Kim Sullivan, (Calidrisalpina) in northern Alaska. Condor 68: JanetKjelmyr, and Tad Theimerhelped with the field 3-46. work and Philip Henderson, Carolyn Frederiksen, 1972. Ecologicalfactors influencing the and Frances Bidstrup provided valuable observa- breeding season schedule of Western Sandpi- tions. John and Ricky Warriner generouslyshared pers (Calidrisrnauri) in subarcticAlaska. Amer. with us their knowledge of Snowy Plovers at Mon- Midi. Natur. 87: 472-491. terey Bay, and CharlesSimis provided weather data. MAYFIELr•,H. F. 1961. Nesting successcalculated Jan Simis allowed us to camp on her land and ex- from exposure.Wilson Bull. 73: 255-261. tended to us very warm hospitality while we were --. 1975. Suggestionsfor calculatingnest suc- there. David Ainley, Peter Connors,David DeSante, cess. Wilson Bull. 87: 456-466. Marshall Howe, Harriet Huber, Frank Pitelka, and OI•INO, L. W., & M. L. KNOrrSON. 1972. Monogamy Chris Ribic provided helpful comments on the and polyandry in the SpottedSandpiper. Living manuscript.We thank theseorganizations and peo- Bird 11: 59-72. ple for their generosity.This is ContributionNo. 189 PITELKA, F. A., R. T. HOLMES, & S. F. MACLEAN. of Point ReyesBird Observatory. 1974. Ecologyand of social organi- zation in arctic sandpipers.Amer. Zool. 14: 185- LITERATURE CITED 204. RITIINGHAUS,H. 1975. Seeregenpfeifer,Charadrius ANDEI•SSOtW,M., & C. G. WI•

The American Ornithologists'Union offers several Marcia Brady Tucker Travel Awards to help defray expensesof outstanding students wishing to present researchpapers at the society'sCentennial Meeting in New York (26 September to 1 October, 1983). To apply, send the following materials to Douglas Mock, Departmentof Zoology,University of Oklahoma,Norman, Oklahoma73019 by 1 May 1983:(i) ex- panded abstractof the paper (max. 3 pp, typed, double-spaced;to include methods,major results,and scientificsignificance), (ii) curriculumvita, (iii) your anticipatedbudget of expenses,and (iv) a letter of support from the academicadvisor supervisingyour research(which should be mailed by itself). Please note that you must provide seven(7) copiesof everything,including the letter of support,because each application is reviewed by the AOU Student Awards Committee's seven members. Also, be advised that applying for a MBT Travel Award does not guaranteea place on the Scientific Program--each student must make arrangementsfor that as per the instructions given with the Call for Papers announcement. INTRASEASONAL REPRODUCTIVE COSTS FOR THE HOUSE SPARROW (PASSER DOMESTICUS)

W. BRUCE McGILLIVRAY • Museumof NaturalHistory, University of Kansas,Lawrence, Kansas 66045 USA

ABSTP,•CT.--HouseSparrows (Passer domesticus) near Calgary, Alberta begin breeding in earlyspring and continuethrough to late summer.High productivityfrom previousbroods is negativelycorrelated with fledglingproduction from secondand third broods.Although fat reservesmay limit the ability of femalesto raiseyoung, there is no concomitantdrop in clutchsize or in the probabilityof renesting.Pairs that fledgemany young in a yearspace fledglingproduction evenly over the breeding seasonbut are most productivein mid- season.The interval between fledging and the initiation of the next clutchincreases with the numberfledged. This delay,an indicationof the physiologicalstrain involved in rearing young,is greaterfor laterbroods and for femalesnesting in trees.Measures of reproductive effort(dutch size, numberfledged, length of the nestlingperiod) vary seasonallybut give no indicationof peakingfor last broods.Thus, reproductiveeffort is not adjustedto parallel changesin the probability of surviving to the next breeding attempt. Received3 November 1981, accepted24 April 1982.

A COMMONassumption underlying theoret- ship of double-broodedfemale House Martins ical discussionsof life-history phenomena is (Delichonurbica), and male Pied Flycatchers that trade-offsin life-history characteristicsoc- (Ficedulahypoleuca) feeding many young were cur (Williams 1966, Chamov and Krebs 1974, lesslikely to return to the study area than those Steams 1976, Snell and King 1977). For exam- feeding fewer young (Askenmo 1979). Lack ple, Pianka and Parker (1975)partition the re- (1966) and Kluyver (1970) present evidence productive value of an individual of age X in suggesting an inverse relationship between a stablenongrowing population into fecundity brood size and parental survivorship. De Ste- at age X and expected fecundity conditioned ven (1980), however, did not find a reduction by survival to ages X + 1, X + 2.... , X + in the survivorship of female Tree Swallows N. Clearly if survivorshipis lowered by current (Iridoprocnebicolor) as a result of brood en- reproduction,future fecundity is reduced. largement. Smith (1981), after demonstratinga Many researchersof life-historyphenomena positive associationbetween reproductive ef- in birds have investigated whether or not the fort and survivorship, concludedthat a trade- modal clutch size is the most productive, as off between fecundity and survivorshipshould predicted by Lack (1947, 1966). If productivity not be assumedfor passerinebirds. is measured as the number of young fledged The longevity and mobility of birds usually from a brood, then clutch sizes larger than prevents a precise determination of mortality modal appear to be optimal (Klomp 1970, von rates for individuals with a known reproduc- Haartman 1971,Jones and Ward 1976, Murphy tive history. If dispersers are a nonrandom 1978). If parental survivorship is reduced by sampleof the population (Lowther 1979),then the effort of raising a large brood, however, estimatesof survivorshipbased on individuals then the optimal clutch size might be lower staying at or returning to the study area will than the most productive (Fisher 1930, Char- be biased. As a consequence,the effect of re- nov and Krebs 1974). Studies testing the rela- production on survivorship is often estimated tionship between brood size and parental sur- from physiologicalevidence. Weight lossesfor vivorshipshow equivocal results. Bryant (1979) femalesthrough the breeding seasonsuggest a found a reduction in the overwinter survivor- physiologicalstrain that may reduce survivor- ship (Newton 1966,Hussell 1972, Bryant 1979). For femaleHouse Sparrows(Passer domesticus), Pinowska (1979) has shown that levels of fat • Present address: Water ResourcesBranch, On- influence clutchsize, while lean dry weight (an tario Ministry of the Environment 135 St. Clair Av- indicator of protein level) may determine the enue W., Toronto, Ontario M4V 1P5, Canada. number of broods raised in a season.Norberg 25 The Auk 100: 25-32. January 1983 26 w. BRUCEMcGxLLXVRA¾ [Auk, Vol. 100

(1981) presented the hypothesis that adult countedand numbered. Nestlings 5•5 days old were weight lossesduring breedingmay be adap- banded with a U.S. Fish and Wildlife Service alu- tive. This would be true if the reduced cost of minum leg band and color marked with leg bands collecting food for the nestlings increases after reachinga 20-g weight. Data analysis.--Becausethe nestswere not checked fledgling numbersmore than it reducessur- every day, exact clutch-initiation data and nestling vivorship and future fecundity. ages were not known for all nests. For these cases, Birds that raise more than one brood in a estimateswere used. Nestling age when first found seasonmay be expectedto show differential was estimatedby comparingthe weight of the largest reproductiveeffort and productivitythrough nestling in a brood to averagevalues obtained from the breedingseason. There is a higher proba- broods of known age. Clutch-initiation date was es- bility of survivingfrom one broodto the next timated by assuming an average incubation period ct-uringthe breedingseason than from the last of 11 days and a laying rate of one per day brood of one seasonto the following spring. (McGillivray 1978). This estimate was needed only An increasein the level of reproductiveeffort for clutchesbegun before my arrivalat the studyarea. would be expectedas the probability of sur- In this study I investigatedthe productivity from multiple broodsof a pair; hence,I made the general viving to the next breeding attempt decreases. assumptionthat broodsraised at a singlenest through There is someevidence that high levelsof re- the breeding seasonwere raised by the same pair. productiveeffort lower the ability of females House Sparrowsusually remain at the same site un- to invest in a subsequentbrood. For Great Tits less one of the pair dies (Summers-Smith 1963). To (Parusmajor), Kluyver (1963) found a lesser reducethe importanceof this potentialsource of error, probabilityof a secondbrood if the first was I imposed some residency conditions. If a tree nest large. Pinkowski(1977) noted a drop in the was extensively modified following a nesting at- clutch size of second broods for Eastern Blue- tempt, ! assumedthat a new pair was nesting. The birds(Sialia sialis) that raised a largefirst brood, average interval between successiveclutch initia- and Smith and Roff (1980) found that an in- tions for first through third clutchesfor each brood crease in the interbrood interval was associate0 size was determined. Ninety five percent prediction intervals were calculated for each clutch-interval, with a largefirst brood for SongSparrows (Me- brood-size combination. Nests where the interclutch lospizamelodia ). interval fell outside these limits were excluded from In this study I look at the productivityof further analyses. multibrooded House Sparrows through the The interval between the departureof the last nest- breeding seasonand address two questions. ling and the initiation of the next clutchis called the First, is there variation in the productivity of interbrood interval. The date of fledgling departure successivebroods within the breeding season was considered to be the midpoint of the interval associatedwith brood size in earlier broods (is between the last observationof the nestlingsand the there a demonstrable within-season cost as- first nest checkafter departure. Fledging is a gradual process,so a departure date is only an approximate sociatedwith reproduction)?Second, do adults measure. Paired comparison t-tests were used to modify their reproductiveeffort in accordance compare the mean interbrood interval of successive with seasonalchanges in the probabilityof re- broods and different nest types. nesting and surviving? For the investigationof the effectsof one brood on the next breeding attempt, data for the two years were pooled.Pooling requires trends to be consistent METHOVS ANt) MATERIALS over the two years in order for the trends to be de- Studysite.--In 1977,the studyareas were six farms tected, and productivity for both years was similar near Conrich, Alberta, 5 km eastof Calgary. For 1978, except for the first broods. The strength of trends all data were collected at one of these farms, 8 km associatedwith brood size was measured by Pear- east of Calgary. At each of the farms, about 20 nest son's product-momentcorrelation coefficient, r. Al- boxeswere erectedin fall 1974 (seeMurphy 1978 for though significancelevels were determined for r, details). As well as nesting in boxes, the sparrows paired variableswere not testedfor bivariate normal- used buildings, machinery, and trees. At the farm ity. Nonetheless, r is a useful indicator of trend. studied both years, nests built in trees were moni- This analysis assumesthat fledglings of equal tored in addition to those in nest boxes. weight contribute equally to parental fitness. This Data collection.--Nestswere inspectedat 3-5 day assumptionmay not be valid for a specieswith a intervalsfrom 2 May to 15 August 1977and from 20 long breeding season.Dyer et al. (1977) published April to 13 August 1978. If weather conditionsper- the following survivorship equation for juvenile mitted, a 3-day interval was maintained. Eggswere House Sparrows (from Summers-Smith 1963); Y = January1983] HouseSparrow Reproductive Costs 27

TABLE2. Interbroodinterval (days).

Number of 1977-box 1977-tree fledglings 1978 nestsa nests•

First to second 0 11.4 6.0** 8.1' i 6.4 6.3 8.0 2 7.4 7.3 8.3 3 10.8 8.0 9.4 4 9.5 9.1 11.0 5 11.0 8.3 --

Second to third 0 10.7 6.9** 7.2 1 10.0 7.4 9.2 2 10.9 7.9 9.4 3 11.7 8.7 10.0 4 10.2 8.1 9.2 5 12.0 10.6 9.5

• * - correlationwith numberof fledglingsis significantat P < 0.05. ß* = correlationwith numberof fledglingsis significantat P < 0.01. Paired comparisont-test for differencesbetween first to second and secondto third intervals: d = 0.694, t - 1.92, 16 df, 0.05 < P < 0.10. Paired comparisont-test for differencesbetween tree and box nests for 1977: d = 1.18, t - 4.32, 10 df, P < 0.01.

102.9X0.•;s,, where Y = percentageof juvenilessur- viving and X = month after fledging. If this equa- tion accuratelydescribes juvenile survivorship, then late-fledgingyoung are more likely to be recruited into the population, and reproductive effort should be concentratedtoward thff end of the breedingsea- son. In a seasonal environment such as that at Cal- gary, however, late-fledgingyoung experiencecold temperatures, snowcover, and food shortagesat a younger age than do early-fledging young. Thus, a late-fledgingcohort might suffer greateroverwinter mortality(McGillivray 1981a). Predation on and nestlingswas rare at this site. Few entire clutchesor broodsdisappeared, and mostnesting mortality canbe attributed to starvation.

RESULTS

For speciesthat raise a single brood each season,the breedingseason would be that pe- riod of the year during which parentshave the highest probability of successfullyfledging young. The same is true for a multibrooded ,but conditionsare unlikely to be con- stant through the season.The averageHouse Sparrow productivity for clutch initiations 1 through4 is displayedin Table 1. If one con- siders only the averagenumber fledged, the secondbrood is the most productive,the third slightly less, the fourth relatively unproduc- tive, and the first quite differentbetween years. 28 w. BRUCEMcGILLIVRAy [Auk, Vol. 100

TABLE3. Relationship between fledgling production and measuresof reproductive effort (and standard error) for each brood.

Number Nestling period Breeding attempt fledged Clutch sizea Fledglingweight length(days) • length(days) a'b Brood 1 0 4.68** (0.06) -- 2.82* (0.25) 26.32** (0.60) 1 4.81 (0.17) 24.52 (0.48) 13.03 (0.50) 33.32 (0.55) 2 4.50 (0.13) 25.47 (0.35) 13.56 (0.41) 35.73 (0.31) 3 4.87 (0.10) 25.21 (0.36) 13.58 (0.28) 36.65 (0.52) 4 5.29 (0.11) 24.30 (0.39) 13.77 (0.36) 37.23 (0.55) 5 5.58 (0.15) 23.74 (0.45) 13.91 (0.72) 37.80 (1.08) 6 6.00 (0.00) 27.84 (0.00) 15.00 (0.00) 39.00 (0.00) Brood 2 0 4.95** (0.08) -- 2.28* (0.27) 24.00** (0.88) 1 4.96 (0.19) 25.78 (0.64) 12.41 (0.55) 33.94 (0.84) 2 5.13 (0.11) 26.57 (0.36) 12.31 (0.30) 35.72 (0.66) 3 5.20 (0.09) 26.17 (0.35) 12.90 (0.26) 37.46 (0.56) 4 5.37 (0.10) 25.07 (0.38) 12.76 (0.52) 36.03 (0.51) 5 5.45 (0.11) 24.96 (0.63) 13.23 (0.50) 39.03 (0.98) 6 6.00 (0.00) 25.75 (3.50) 14.50 (2.50) 43.00 (0.00) Brood 3 0 4.59* (0.10) -- 2.06 (0.27) 23.44 (1.03) 1 5.00 (0.25) 26.20 (0.83) 12.08 (0.57) 37.00 (3.50) 2 4.50 (0.19) 26.64 (0.52) 11.74 (0.51) --• 3 4.89 (0.14) 26.80 (0.42) 11.97 (0.32) 37.25 (1.70) 4 5.19 (0.09) 26.77 (0.41) 12.67 (0.32) 35.80 (1.96) 5 5.25 (0.11) 26.05 (0.58) 12.57 (0.80) 33.60 (0.68) 6 6.00 (0.00) 25.42 (0.00) 13.00 (0.00) __c Brood 4 0 4.22 (0.17) -- 1.87 (0.50) __c 1 4.67 (0.42) 26.65 (0.83) 10.67 (1.10) -- 2 4.20 (0.26) 24.76 (1.74) 10.50 (0.87) -- 3 4.80 (0.37) 27.60 (0.72) 11.80 (0.58) -- 4 5.00 (0.00) 24.31 (1.31) 13.50 (1.50) --

• * - correlationwith the number of fledglings significant at P < 0.05. ß* - correlationwith the number of fledglings significant at P < 0.01. • Interval between successive clutch initiations. "No subsequentclutch initiated.

High fledgling output from first broodsis un- an increase over the second to third interval is usualfor House Sparrowsin this area(Murphy observed.The longer interbrood interval for 1977), because cold weather, rain and snow are pairs nesting in trees suggeststhat daily en- commonthroughout May (McGillivray 1981b). ergy requirementsare lower for box-nesting That the interval between the fledging of a females. brood and the initiation of the next clutch (in- Why the physiologicalcost of nestingmight terbrood interval) increases as the number increasewith the numberof youngfledged from fledging increasesis shown in Table 2. The a brood is shown in Table 3. On the average, longer interval between second and third femalesmust expendenergy laying additional broods comparedto the interval between first eggs,and the pair must spenda longer time and secondbroods implies that the physiolog- feeding the young in large broods.As brood ical costsof breeding are cumulative. Sample size increases, adult House Sparrows increase sizes for the third to fourth brood interval are the rate of feedingtrips to the nestlings(Sum- too small to consider(<5), exceptfor box nests mers-Smith1963, Seel 1969,Sappington 1975, in 1977that fledgedno youngin the third brood McGillivray 1981a), and, at least for these two (interbroodinterval = 7.6 days,n = 17), where years,average fledging weight doesnot change January1983] HouseSparrow Reproductive Costs 29

TABLE4. Relationship between the number of TABLE5. Modal number fledgedin each brood for fledglingsproduced and productivityin the sub- pairs groupedby seasonaltotal fledgling produc- sequentbreeding attempt. tion.

Proba- Seasonaltotalfledglingproduction bility of starting Brood 1-2 3•4 •6 7-8 %10 >10 Number Clutch Number next I 2.0 3.0 3.0 3.5 3.0 3.0 fledged n size fledged clutch 2 2.0 3.0 3.0 4.0 5.0 4.5 Brood 1 Brood 2 3 0.0 0.0 0.0 0.0 4.5 4.0 4 0.0 0.0 0.0 0.0 1.5 1.5 0 234 5.10 1.69 0.70 1 38 5.48 2.36 0.82 2 61 5.21 2.23 0.77 3 81 5.06 2.07 0.84 4 56 5.10 1.88 0.73 broods, I compared third and fourth brood 5 14 5.40 1.60 0.71 productivitywith summedfledgling totals from 6 1 4.00 0.00 1.00 all previous broods. For pairs fiedging at least Brood 1 and Brood 2 Brood 3 one young, the summed fledgling production 0 41 4.63 1.22 0.60 from broods i and 2 is negatively correlated 1 13 5.08 1.69 0.81 2-3 75 4.84 1.83 0.75 with the number fiedging in brood 3 (Table 4, 4-5 65 4.83 1.54 0.69 r = -0.81, 4 df, 0.05 < P < 0.10). Again, the 6-7 37 4.51 1.46 0.51 probability of renesting and the average clutch 8-9 12 5.00 0.50 0.75 size are not related to productivity earlier in >9 1 5.00 0.00 1.00 the season. Broods 1, 2, and 3 Brood 4 There is a nonsignificant, negative relation- 0 7 4.00 1.14 0.22 ship between total productivity in broods 1, 2, 1-3 18 4.78 0.50 0.41 4-6 20 4.20 1.00 0.22 and 3 and the averagenumber fiedging in brood 7-9 8 4.25 0.62 0.15 4 (Table4). The probability of initiating a sub- 10-12 2 4.00 0.50 0.15 sequent brood has dropped from the previous two comparisonsand is lowest for pairs with the highest fledgling numbers from the pre- vious three broods. with brood size. Although thermoregulatory When the fledgling data from the three com- costs should be lower for nestlings in larger parisons (broods I vs. 2; i and 2 vs. 3; and 1, broods, these are unlikely to offset the extra 2, and 3 vs. 4) are combined and standardized feeding demands on parents created by more by transformation to Z-scores, previous pro- nestlings. ductivity is shown to be negativelycorrelated For the combined breeding seasons1977 and with future productivity(r = -0.63, 14 df, P < 1978, pairs that showed higher productivity 0.05). from the first clutch tended to have lower out- The strongnegative relationship between first put from the secondclutch (Table 4). For pairs and secondbrood productivity is reflected in that fledged at least one young from the first the distribution of fledgling production over brood, there is a negative correlationbetween the breeding season.Table 5 demonstratesthat the number fledged in brood 1 and the number the highest fledgling totals were achieved by fledged in brood 2 (r = -0.85, 4 df, P < 0.05). pairs with high post-first-broodproductivity. The classcontaining pairs fiedging no young A prediction from life-history theory is that, was excludedfrom analysesbecause of its het- unlessthe raising of last broods lowers adult erogeneity. It includes a range from pairs rais- survivorship, reproductive effort should be ing young to near fiedging to others losing higher for that brood than for broodsearlier in clutches before hatching. The probability of the season. Last broods, as defined here, in- startinga secondclutch and the size of second clude fourth broods and third broods after clutches are uncorrelated with the number of which no fourth clutch was initiated. Because fledglings from brood one. no physiologicalmeasures were made, only in- Becausethe interbrood interval data sug- direct estimates of effort can be used. Table 6 gestedthat physiologicalstrain is additive over demonstratesthat, for four possible indicators 30 w. BRUCEMcGILLIVRAY [Auk, Vol. 100

TABLE6. Measuresof reproductiveeffort (and standarderror) by pairs for broodsafter which fourth clutches were initiated, third broods that were last broods and fourth broods.

Third broods

Variable Not last Last Fourth broods

Clutch size 4.83 (0.10) 4.74 (0.07) 4.36 (0.13) Number fledged 2.93 (0.18) 2.78 (0.13) 2.11 (0.23) Fledglingweight (g) 26.54 (0.40) 26.48 (0.25) 26.01 (0.65) Nestling period (days) 12.36 (0.37) 11.70 (0.25) 11.22 (0.48)

of reproductiveeffort, averagevalues for third secondbroods indicatesa cumulativephysio- broods that are also last broods are not signif- logicaleffect of the first two broods.This effect icantly different from those of third broods af- is suggestedby Pinowska's observations(1979) ter which a fourth clutch was laid. There is also of protein levels and by the reduced survivor- no evidence of enhanced effort for fourth ship of double-brooded House Martins found broods. by Bryant (1979).

DISCUSSION The longer interbrood interval for pairs us- ing tree nests supports other observations These data suggestthat the trade-offs and (Murphy 1978,McGillivray 1981b)showing that the theoryrequired to accountfor them are dif- box nestsare superior nest sites. A warmer mi- ferent for single and multibrooded species.An croclimate (Mertens 1977) would reduce nest- inverse relationship between present and fu- ling thermoregulatorycosts, and a reduction in ture fecundity is suggestedbut not as a func- the need for nest maintenance would facilitate tion of reduced survivorship. Rather, high pro- renesting in boxes. ductivity delaysthe initiation of the following While the tree-versus box-nest comparisons clutch and is negatively correlatedwith pro- provide appealing evidence for physiological ductivity from future clutches. correlatesof reproductive costs,they point out Two studies (Schifferli 1976, Pinowska 1979) a problem in the evaluation of reproductive ef- have monitored the physiologicalcondition of fort. If nest-site selection can modify the cost adult House Sparrows through the breeding of raising young, then other factors such as season.For females, fat level drops during egg weather, distance to food sources, time of clutch laying, is partially replacedduring incubation, initiation, and parental age and size are also drops again during the nestling period, and likely to modify the reproductiveeffort asso- presumably is restored in the interbrood in- ciated with raising a brood of fixed size. The terval. Protein level drops gradually through importanceof factorsother than productivity the breeding season. Pinowska (1979) found may provide a partial explanationfor the con- that clutch size was correlated with fat level tradictoryresults obtained in previous studies and the number of clutches initiated in a sea- of the trade-offbetween reproductiveeffort and son was limited by protein level. For males,fat survivorship in (Bryant 1979, As- level drops during the nestling period and is kenmo 1979, De Steven 1980, Smith 1981). replenished in the interbrood interval. There is no simple physiologicalhypothesis These two studies suggestthat, after com- to accountfor reduced productivity from pairs pleting a brood, both sexesof a pair fiedging fledging many young. Becauseclutch size and many young would have lower fat levelsthan the probability of renesting are generally un- those of a pair fiedging few young. The lower related to previous productivity, the cost of re- fat level should result in either a smaller clutch production is felt by parents during the incu- size or a longer delay in the initiation of the bation and nestling period. While the nestling subsequentclutch. My observationsshow that period is consideredhighly demanding(Rick- clutch size is not affectedby previousproduc- lefs 1973), much of the evidence (particularly tivity but subsequentclutch initiations are de- weight lossesfor adults) is now questionable layed. The longerinterval betweensecond and (Norberg 1981). third broodscompared to that betweenfirst and There is no evidence from the data that re- January1983] HouseSparrow Reproductive Costs 31 productive effort increasesas the probability BRYANT,D. 1979. Reproductivecosts in the House of surviving to the next breeding attempt de- Martin ( urbica). J. Anim. Ecol. 48: 655- creases.As mentioned earlier, however, only 675. partial estimatesof reproductiveeffort are pos- CHARNOV, E. L., & J. R. KREBS. 1974. On clutch sible. Pinowska (1979) felt that breeding in size and fitness. Ibis 116: 217-219. House Sparrowsfollowed a cyclicalpattern in DE STEVEN,D. 1980. Clutch size, breeding success and parental survival in the Tree Swallow (Iri- which a poolof resourcesis alternativelytapped doprocnebicolor). Evolution 34: 278-291. and partially replenisheduntil depletedbelow DYER, M. I., J. PINOWSKI, & B. PINOWSKA. 1977. a critical point and breeding ceases.Hence, Populationdynamics. Pp 55-103 in Granivorous second, third, and fourth broods are initiated birds in ecosystems(J. Pinowski and S. C. Ken- only by females that possess sufficient re- deigh, Eds.). Cambridge, England, Cambridge sources. Last broods, which should show en- Univ. Press. hanced reproductiveeffort accordingto life- FISHER,R. A. 1930. The geneticaltheory of natural history theory, might, in fact, show evidence selection. Oxford, Clarendon. of physiologicalstrain. Seasonaldifferences VON HAARTMAN,L. 1971. Populationdynamics. Pp. make comparisonsamong broods difficult to 392-459 in Avian biology (D. S. Famer and J. R. King, Eds.). New York, AcademicPress. interpret, but the comparisonbetween third HUSSELL,D. J.T. 1972. Factorsaffecting clutch size broods that were last broods and third broods in arctic passerines. Ecol. Monogr. 42: 317-364. after which a fourth clutch was initiated is re- JONES,P. J., & P. WARD. 1976. The level of reserve vealing. Table 6 indicatesthat, while none of protein as the proximate factor controlling the the chosen measures of effort is different be- timing of breeding and clutch size in the Red- tween groups, the direction of difference al- billed Quelea, Queleaquelea. Ibis 118: 547-573. ways suggestslower resourcelevels for pairs KLOMP, H. 1970. The determination of clutch size raising last broods. in birds, a review. Ardea 58: 1-124. A reviewer pointed out that life-history KLUYVER,H. N. 1963. The determination of repro- ductive rates in Paridae. Proc. 13th Intern. Or- theorists consider ultimate causes, but field nithol. Congr.: 706-716. workers are forced to explain proximate rea- 1970. Regulation of numbers in popula- sonsfor phenomena such as clutch size, num- tions of Great Tits, Parusm. major. Pp. 507-523 ber of breedingattempts per season,and sur- in Dynamics of numbers in populations (P. J. vivorship. The lack of concordance between den Boer and G. R. Gradwell, Eds.). Wagenigen, theory and field observations(Smith 1981, this The Netherlands. Proc. Adv. Study Inst. Oos- study), while perhapsan indictmentof the the- terbeck, The Netherlands, PUDOC. ory, is equally likely to be the result of failure LACK,D. 1947. The significanceof clutchsize. Ibis to controland accountfor proximatevariation. 89: 302-352. 1966. Population studiesof birds. Oxford, ACKNOWLEDGMENTS Clarendon Press. LOWTHER,P. E. 1979. Growth and dispersal of I firstwould like to thankEdward C. Murphy and nestlingHouse Sparrows: sexual differences. In- Peter E. Lowther, whose ideas generatedmy study land Bird Banding 5: 23-29. and focusedmy attention. RichardF. Johnstonpro- McGirriVR•Y, W.B. 1978. The effectsof nest po- vided constantencouragement and financialsupport sition on reproductiveperformance of the House and gave freely of his extensiveknowledge of Passer Sparrow. Unpublished M.A. thesis, Lawrence, biology. S.C. McGillivray, R. C. Fleischer,G. Pitt- Kansas, Univ. Kansas. man, and J. T. Paul, Jr. aided with the fieldwork. I 1981a. Breeding ecology of House Spar- appreciatethe time given by H. L. Levenson,P. E. rows. Unpublished Ph.D. dissertation, Law- Lowther, R. C. Fleischer,M. T. Murphy, and G. E. rence, Kansas, Univ. Kansas. Gurri-Glass in considerationof my ideas. R. F. John- 1981b. Climatic influences on productivity ston, M. S. Gaines, N. A. Slade, D. C. Seel, and an in the House Sparrow. Wilson Bull. 93: 196-206. anonymousreviewer all read and improveddrafts of MERTENS,J. A. L. 1977. Thermal conditions for this manuscriptwith their comments.Financial sup- successfulbreeding in Great Tits (Parusmajor port was supplied by National ScienceFoundation L.). II. Thermal propertiesof nestsand nestbox- grants DEB-02374and DBS-7912412to R. F. Johnston. es and their implications for the range of tem- peraturetolerance of Great Tit broods.Oecologia LITERATURE CITED 28: 31-56. ASKENMO,C. 1979. Reproductive effort and return MORVHY,E. C. 1977. Breeding ecologyof House rate of male Pied Flycatchers.Amer. Natur. 114: Sparrows. Unpublished Ph.D. dissertation, 748-752. Lawrence, Kansas, Univ. Kansas. 32 w. BRUCEMcGILLIVRAy [Auk, Vol. 100

1978. Seasonal variation in reproductive SCHIFFERLI,L. 1976. Weight conditions in the output of House Sparrows:the determinationof HouseSparrow particularly when breeding.Un- clutchsize. Ecology59: 1189-1199. published Ph.D. dissertation, Oxford, Oxford NEWTON,I. 1966. Fluctuationsin the weight of Univ. bullfinches. Brit. Birds 19: 89-100. SEEr, D.C. 1969. Food, feeding rates and body NORBERG,R. •. 1981. Temporaryweight decrease temperature in the nestling House Sparrow in breeding birds may result in more fledged (Passerdomesticus) at Oxford. Ibis 111: 36-47. young. Amer. Natur. 118: 838450. SMITH,J. N.M. 1981. Does high fecundityreduce PIANKA,E. R., & W. S. PARKER.1975. Age specific survival in Song Sparrows?Evolution 35: 1142- reproductivetactics. Amer. Natur. 109: 453-464. 1148. PINKOWSK•,B.C. 1977. Breeding adaptationsin , & D. A. RUFF. 1980. Temporal spacingof the Eastern . Condor 79: 289-302. broods, brood size and parental care in song PINOWSKA,J. 1979. The effect of energy and build- sparrows (Melospiza melodia). Can J. Zool. 58: ing resourcesof females on the production of 1007-1015. House Sparrows Passerdomesticus L. popula- SNELL,T. W., & C. E. KINe. 1977. Lifespan and tions. Ecol. Polska 27: 363-396. fecunditypatterns in rotifers:the costof repro- R•CKLEES,R. E. 1974. Energeticsof reproduction duction. Ecology31: 882490. in birds. Pp. 152-292 in Avian energetics(R. A. STEARNS,S.C. 1976. Life-history tactics:a review Paynter, Jr., Ed.). Publ. Nuttall Ornithol. Club of the ideas. Quart. Rev. Biol. 51: 3-47. No. 15. SUMMERS-SMITH,D. 1963. The House Sparrow. SAPPINGTON,J. N. 1975. Cooperative breeding in London, Collins. the House Sparrow (Passerdomesticus). Unpub- WILL•AMS, G.C. 1966. Natural selection, the costs lished Ph.D. dissertation, Mississippi State, of reproductionand a refinementof Lack'sprin- Mississippi, Mississippi State Univ. ciple. Amer. Natur. 100: 687-690.

Now AVAILABLE

WorldInventory of Avian Skeletal Specimens. 1982. D. ScottWood, RichardL. Zusi, and Marion Anne Jenkinson. WorldInventory of Avian SpiritSpecimens. 1982. D. ScottWood, Richard L. Zusi, and Marion Anne Jenkinson. Published by The American Ornithologists'Union and The Oklahoma BiologicalSurvey, the cost of each inventory is $25.00, including surface mail postage.For air mail, add $5.00 (U.S.), $5.50 (Canada, Alaska, Hawaii), $9.00 (Mexico, Central America), $15.50 (Europe, ), or $22.00 (Asia, Australia, New Zealand, Africa). Make checkor money order, in American dollars, payable to University of Oklahoma. Order from Dr. Gary D. Schnell, Oklahoma BiologicalSurvey, Sutton Hall, University of Oklahoma, Norman, Oklahoma 73019, U.S.A.

The First Conferenceon Birds Wintering in the MediterraneanRegion will be held on 23-25 February1984 at Aulla, Italy. Emphasiswill be on the ecology,ethology, distribution, and migration of birds wintering in this region. For further information and preregistrationmaterials contact Dr. Almo Farina, Museum of Natural History of Lunigiana, Fortezzadella Brunella, 54011Aulla, Italy.

The Raptor ManagementInformation System(RMIS) is a collectionof published and unpublishedpapers, reports, and other works on raptor management and human impacts on raptors and their habitats. It currentlyconsists of nearly 2,500 original papers, 160 keyworded notecarddecks comprised of 15,000key paragraphsfrom the original papers,and a computerprogram to retrieve partially annotatedbibliographies by species,by keyword, or by any combinationof keywordsand/or species. A geographicalindex is under development, and new papers are added as they are received. Originally designed to facilitate land-use planning and decisionmakingby governmentagencies and industry, the RMIS has since grown into a powerful researchand environmentalassessment tool for scholars,students, consultants, as well as land managersand their staff biologists. For more information write Dr. Richard R. Olendorff, U.S. Bureau of Land Management,2800 Cottage Way, Sacramento,California 95825 U.S.A., or phonecommercial (916) 484-4701or throughthe FederalTelephone System 468-4701. GENIC VARIATION, SYSTEMATIC, AND BIOGEOGRAPHIC RELATIONSHIPS OF SOME GALLIFORM BIRDS

R. J. GUTIERREZ,1 ROBERTM. ZINK, AND SUH Y. YANG2 Museumof VertebrateZoology, University of California,Berkeley, California 94720 USA

ABSTRACT.--Starchgel electrophoresiswas used to evaluate levels and patterns of genic differentiationamong 10 speciesof galliform birds in the (9) and Tetraonidae (1). The phasianidsincluded an Old World , a , a ,and six species of . Measuresof within-speciesgenetic variation includedheterozygosity, percentagepolymorphic loci, and number of allelesper polymorphiclocus. These values were similar to but lower than those reported for other birds. Genetic distancesamong conspecificpopulations and among congenericspecies were low comparedto other avian results. Genetic distancesamong noncongenersboth within and between families were considerablyhigher, however, than those reported for passerinebirds. Thus, more studies of levelsof genic differentiationamong nonpasserines are required to complementthe lit- erature on genic divergenceamong passerinesand to enable us to make general statements about genic evolution in birds. Phenogramsand phylogenetictrees suggestedthat Phasianuscolchicus, pal- lidicinctus, coturnix, chukar, and the New World quail (Odontophorinae) are genically distinct taxa. The branching sequenceamong the non-Odontophorine taxa is unresolvedby our data. The branching order among taxa in the Odontophorinaefrom a commonancestor is: Cyrtonyxmontezumae, Oreortyx pictus, virginianus, squamata,Lophortyx gambelii, and L. californicus.The generaCyrtonyx, Oreortyx, and Colinus are clearly distinct from Callipeplaand Lophortyx,which are quite similar to each other genically. We use a fossil speciesfrom the mid-Miocene of Nebraskato calibrateour geneticdis- tances.We estimatedates of divergenceof taxain the Odontophorinaeand offer a hypothesis on their historicalbiogeography. Our analysissuggests that three east-westrange disjunc- tions could accountfor the origin of Oreortyx (12.6 MYBP), Colinus(7.0 MYBP), and Calli- pepla-Lophortyx(2.8 MYBP). We suggestthat L. californicusand L. gambeliishould be con- sidereddistinct speciesbecause of an apparentlack of panmixia in zonesof sympatry,even though the D between them is typical of that found between subspeciesof other birds. Oreortyxand Colinusshould remain as distinct genera,while our data are equivocalon the statusof Callipeplaand Lophortyx?Received 9 March 1982, accepted5 July 1982.

ALLOZYMEelectrophoresis has been used less examinedavian intrafamilial relationships(e.g. frequently to examinegenetic variation within Barrowclough and Corbin 1978, Avise et al. and among groups of birds than in other ver- 1980a-c).They have found that passerinebirds tebrates (see review in Nevo 1978). Some possess considerably lower levels of genic workers have examined patterns of intraspe- (= allozymic) differentiation than other verte- cificgenic variation in passerines(e.g. Barrow- brate taxa, at comparable taxonomic levels. clough 1980,Johnson and Brown 1980, Corbin Several workers have compared the level of 1981, and references therein), and a few have genic divergence and taxonomicrank for var- ious vertebrate and invertebrate taxa (e.g. Ay- • Present address: Department of Wildlife Man- ala 1975, Avise et al. 1980b). Barrowcloughet agement, Humboldt State University, Arcata, Cali- fornia 95521 USA. al. (1981)present similar data for birds but dis- 2 Present address:Department of Biology, Inha cussreasons why comparisonsacross different University, Inchon, Korea. groups of organisms may be inappropriate; :3Lophortyx has been merged with Callipeplain the these include taxonomic artifacts [e.g. avian, Thirty-fourth Supplement to the American Orni- mammalian, and reptilian genera may not be thologists'Union check-listof North American Birds comparablebecause of the way in which tax- [Auk 99 (3, Suppl.): 5CC]. onomistspartition variation (Sibley and Ahl- 33 The Auk 100:33-47. January1983 34 GOX•gRREZ,ZINK, ANDYANG [Auk, Vol. 100

quist 1982)] and differences in rates of evolu- zygosity(/•) was definedas the numberof hetero- tion, mating systems,effective population sizes, zygous genotypesrecorded in a sample divided by recencyof common ancestry,and dispersalpa- the product of the number of loci and the number of rameters. Whether or not low levels of genic individualsassayed (see Corbin 1981for a discussion divergencetypify birds as a group is unclear, of calculating/z/).Estimates of percentagepolymor- becauseonly Barrowcloughet al. (1981)studied phismwere based on thenumber of locihaving more than one allele divided by the total number of loci a nonpasserinetaxon. They found higher levels (27) examined. of differentiation among some procellariiform The measuresof Nei (1978)and Rogers (1972) were taxa than those usually found among passer- used to estimate genetic distances between taxa. ines. Becausethe levelsof geneticdifferentia- Clusteranalyses, summarizing the matrixof Rogers' tion are sometimes used to make inferences D-values,were performed with boththe unweighted about evolutionaryprocesses (Avise et al. 1980c, and weighted pair-group methods,using arithmetic Templeton 1980), we clearly need additional means(UPGMA and WPGMA, respectively).The co- studies of nonpasserinesbefore we can make phenetic correlation coefficient, rcc,was used to eval- general statementsabout genic evolution in uate how well the resultant phenogramsrepresent birds. the original distancematrix. Sneathand Sokal(1973) provide details on these pheneticmethods. Phylo- The patternsof genic differencescan alsobe genetic trees, also based on Rogers'D-values, were used to infer phylogeneticrelationships (e.g. constructedaccording to the methodsof Farris (1972; Barrowcloughand Corbin 1978,Zink 1982). In Wagner trees) and Fitch and Margoliash(1967; F-M this paperwe examinelevels of genicvariation trees). The Wagner tree is an approximationof the and phylogenetic relationships among 10 most parsimonious tree. The F-M procedure con- speciesfrom 5 of 10 generaof New World quail, structsa number of treesby alteringthe branching an , a partridge, a pheasant, structureand branch lengths. Alternative trees were and a . We use our phylogenetic hy- evaluatedby the percentagestandard deviation (%SD) pothesis to construct an estimate of the evo- and by the numberof negativebranches (the fewer the better). A lower %SD means a better fit of dis- lutionary history of some New World quail. tancesimplied by the tree to the original distance We also evaluateprevious statementsabout the matrix (Fitch and Margoliash 1967). A cladisticanal- taxonomic relationships of these galliform ysis, sensuHennig (1966), using alleles as character birds. states (see Wake 1981), basically corroborated the above methods. The allele in T. pallidicinctus(Te- MATERIALS AND METHODS traonidae) was considered"primitive" when com- We examined 217 specimens of galliform birds paring the pattern of allele distribution in the re- representing 10 species. Species, localities, sample maining taxa (Phasianidae),i.e.T. pallidicinctuswas sizes, and the taxonomic framework used in this pa- used as an "outgroup" to the phasianids. per are given in Appendix 1. Nomenclaturefollows the A.O.U. check-list (1957, 1973). RESULTS Liver, heart, and kidney tissuewere excisedin the field within 4 h of death and frozen in liquid nitro- Protein variation.--Twenty structural pro- gen. Tissueswere homogenizedusing the methods teins and enzymes encodedby 27 presumptive of Selander et al. (1971), and extractswere stored at genetic loci were examined in all individuals. -76øC until used for electrophoresis.Combined tis- Allelic frequenciesfor the 23 variable loci, per- sue extractswere subjectedto horizontal starch gel centagepolymorphism and heterozygosity,and electrophoresisas describedby Selanderet al. (1971). number of alleles per polymorphic locus are Gel and buffer systemsfor the loci listed in Appendix given in Appendix 2. Four loci (Mdh-1, Mdh- 2 were essentiallythe same as those describedby Yang and Patton (1981). More detailed information 2, Lap, Pt-l) were monomorphicand fixed for regardingelectmphoretic conditions is availablefrom the same electromorph acrossspecies. Seven the authors. loci (c•Gpd-1,Got-2, Udh, Gdh, Ldh-1, Ldh-2, We assumethat our electrophoreticallydetectable Pept-2) were monomorphicwithin speciesbut variants (= electromorphs)at a locus differ geneti- exhibited interspecific differences. The re- cally; hence, we refer to them as alleles.Alleles at a maining loci were polymorphicin somespecies locus were coded by their mobility from the origin. and alsoshowed interspecific fixed differences. Thg most anodal allele was designated as a, with We excludethe laboratorystrains of Coturnix slower alleles denoted as b, c, d, etc. Isozyme no- and Alectorisfrom discussionsof within-species menclaturefollows Yang and Patton (1981). Hetero- variation, becausethese levels of variation may January1983] GenicVariation in Galliforms 35

TABLE1. Matrix of geneticdistances between 17 taxa of galliform birds. Distancescomputed by methods of Nei (1978) above diagonal and Rogers(1972) below diagonal.

Species Species I 2 3 4 5 6 7 1. Tympanuchuspallidicinctus -- 1.041 1.308 1.291 1.326 1.201 1.483 2. colchicus 0.649 -- 1.446 1.282 1.310 1.172 1.654 3. Coturnix coturnix A 0.721 0.749 -- 0.064 0.059 1.185 1.340 4. C. coturnix B 0.713 0.714 0.109 -- 0.056 1.136 1.209 5. C. coturnix C 0.729 0.721 0.086 0.101 -- 1.190 1.346 6. Alectoris chukar 0.695 0.682 0.685 0.665 0.689 -- 0.962 7. Lophortyxgambelii 0.769 0.799 0.725 0.692 0.730 0.616 -- 8. L. californicusA 0.773 0.797 0.728 0.690 0.730 0.617 0.028 9. L. californicusB 0.774 0.801 0.732 0.694 0.734 0.617 0.032 10. L. californicusC 0.768 0.795 0.727 0.688 0.728 0.610 0.037 11. Callipeplasquamata A 0.773 0.758 0.752 0.727 0.761 0.652 0.106 12. C. squamataB 0.764 0.750 0.742 0.720 0.752 0.644 0.106 13. Colinusvirginianus A 0.770 0.732 0.741 0.741 0.753 0.687 0.262 14. C. virginianusB 0.772 0.729 0.739 0.738 0.751 0.687 0.262 15. Oreortyxpictus 0.808 0.764 0.751 0.744 0.760 0.649 0.348 16. Cyrtonyxmontezumae A 0.744 0.731 0.746 0.737 0.754 0.684 0.497 17. C. montezumae B 0.735 0.727 0.742 0.734 0.751 0.680 0.492

TABLE 1. Continued.

Species

8 9 10 11 12 13 14 15 16 17

1.495 1.495 1.475 1.493 1.476 1.487 1.495 1.668 1.364 1.335 1.627 1.646 1.617 1.430 1.419 1.342 1.331 1.476 1.334 1.326 1.340 1.343 1.330 1.448 1.429 1.393 1.389 1.450 1.416 1.407 1.204 1.216 1.192 1.339 1.345 1.423 1.413 1.433 1.399 1.390 1.327 1.335 1.312 1.467 1.449 1.427 1.421 1.468 1.435 1.425 0.967 0.966 0.947 1.071 1.054 1.184 1.189 1.056 1.170 1.158 0.005 0.007 0.008 0.089 0.082 0.295 0.295 0.410 0.671 0.653 -- 0.000 0.000 0.115 0.100 0.301 0.298 0.418 0.640 0.624 0.013 -- 0.000 0.121 0.102 0.306 0.304 0.421 0.634 0.617 0.019 0.015 -- 0.120 0.103 0.308 0.306 0.431 0.641 0.623 0.129 0.130 0.131 -- 0.003 0.258 0.262 0.491 0.664 0.645 0.125 0.117 0.126 0.022 -- 0.241 0.244 0.473 0.613 0.593 0.274 0.276 0.280 0.244 0.229 -- -0.002 0.538 0.478 0.462 0.273 0.277 0.283 0.252 0.237 0.015 -- 0.540 0.463 0.448 0.355 0.354 0.367 0.400 0.391 0.424 0.427 -- 0.746 0.744 0.479 0.476 0.479 0.493 0.473 0.395 0.385 0.533 -- 0.001 0.476 0.472 0.472 0.483 0.463 0.389 0.383 0.533 0.018 --

havebeen affectedby prolongedcaptivity. The ble 1. A summary (Table 2) of genetic distance averageproportion of polymorphic loci for the as a function of various taxonomic groupings wild speciesis 14.5% (range0-29.6%). F/ is showsthat /• increasesas taxonomicgroup- 2.6% and rangesfrom 0 to 5.1% (Appendix 2). ings become more inclusive, at least to the Thevalues of F/, percentagepolymorphic loci, subfamily level. This suggeststhat the taxo- and number of alleles per polymorphic locus nomic groupings are "biologically real," based are similar to but lower than thosereported for on our geneticanalysis. Levels of/• for other other groupsof birds (Barrowcloughand Cor- avian taxa are also shown in Table 2 for com- bin 1978, Avise et al. 1980a, Zink 1982). parisonwith the galliform/•'s. Interspecificgenetic distance.--The matrix of At the local population level, the galliforms genetic distancesbetween taxa is given in Ta- sampledhere showless differentiation (/• = 36 GOneRREZ,ZINK, AND YANG [Auk, Vol. 100

•. pallidicinctu$ P. colchicu$ •A t Coturni•coturnix A chukar œ. •ambeh'i ½I œ'califor- Bj nicu$

A0 1 Callipepla$quamata •B I Co#nu$vH•inianu$ O. pictus UPGMA r½½= .99 • I Cyrtonyxmontezumae I ß .8 Rogers' D

I I I I I .58.9 16.0 7.5 2.7 0 Million yeors Fig. 1. UPGMA phenogramderived from the matrix of Rogers'D-values (Table 1). Geographiclocalities given in Appendix 1. Time scaledetermined from the formula ! = 26.3 x 10"D; see text.

0.0007) than other birds, although this may be only slightlybelow the rangegiven by Barrow- due to the closeproximity of some of the sam- cloug_h,0.0078-0.1267, for congenericspecies. ples (Appendix1). The D-valuebetween L. The D between noncongenericspecies in the gambeliiand L. californicus,0.0067, is similarto Odontophorinae, 0.412, is considerablyless that observedbetween subspeciesof other birds thanthe D betweenthese and the other pha- [0.0048 + 0.0049(Barrowclough 1980)], but it is sianids, 1.32. Both of these values are greater

TABLE2. Mean geneticdistance (Nei 1978)as a function of taxonomicrank in somegalliform birds. Taxa includedin eachtaxonomic level are given in Appendix1. D-valuesfrom Table 1. Alsogiven are D-values at comparabletaxonomic levels for other birds (from Barrowclough1980).

Number of compari- Comparable data Taxonomiclevel sons /• -+ SD Range for otherbirds Local population 6 0.0007 -+ 0.0013 -0.00151 to'0.00331 0.0024 -+ 0.0028 Congeneric species(Lophortyx) 3 0.0067 -+ 0.0014 0.00507 to 0.00775 0.0440 _+ 0.0220 Noncongeneric speciesin Odontophorinae 46 0.4116 _+ 0.2021 0.0824 to 0.7460 0.2136 _+ 0.1659 Species in Odontopborinae vs. Phasianus colchicus, Coturnix coturnix, and Alectoris chukar 62 1.3210 _+ 0.1640 0.962 to 1.654 not available Tympanuchuspallidicinctus (Tetraonidae) vs. all other species(Phasianidae) 16 1.4000 _+ 0.1500 1.041 to 1.668 0.6829 •* _+ 0.1970 Basedon study by Barrowcloughet al. (1981)of someprocellariiform birds. January1983] GenicVariation in Galliforrns 37

F. pallidicinctu$ 5.20 t•. colchicus 5.29 .59 5.00 .47 A coturnix .2•/.54 GB tcoturn/x 3.00 ,4. chukar

.59

.64I 1•5 t. •ambe///

.65 ' Lø•B • squamata .o• .58 B vir•inianus 1.50 2.25 O. pictus WAGNER%S D = 9.19 2.46 .•4.•04 AiB montezumaCyrtony x Fig. 2. Wagner tree basedon Rogers'D-values. Branchlengths in units of Rogers'D (x10). The tree is "rooted" at T. pallidicinctus(Farris 1972).

than that given by Barrowclough(1980) for oth- shown (Fig. 1). The Wagner (Fig. 2) and F-M er avian cdnfamilialbut noncongenericcom- (Fig. 3) trees resembleFig. I in terms of the parisons(/• = 0.2136), suggestingthat the overall relationships suggested. Differences Odontophorinae is a distinct group. At the among these three analysessuggest to us that familylevel, the/• observedbetween T. palli- the branching sequence of the subfamilies, dicinctus(Tetraonidae) and the other species(all while themselvesgenically distinct in all anal- in Phasianidae), 1.400, is approximately twice yses, is unresolved by this analysis. That is, the/• reportedfor interfamilialcomparisons of we do not believe that a clearhypothesis of the other birds (0.6829). branching order of the subfamilies emerges Genic relationshipsamong species.--For two from our data. The level of differentiation and reasons,phenetic (UPGMA and WPGMA) and the branchingdiagrams show considerabledi- cladistic(Wagner and F-M) procedureswere vergence, at the structural gene level, among used to constructbranching diagrams (Figs. 1- thesegalliform birds. The branchingsequence 3). First, we wanted to determine whether or of taxa within the Odontophorinae was iden- not the branchingstructure was dependenton tical in all branching diagrams; therefore, we which methods were used (see Presch 1979). feel that it is a robust result. Second, the methods for constructingtrees, as The branching diagrams show that there is opposedto phenograms,are independentof considerabledivergence among taxa within the the assumptionof homogeneityof evolution- Ondontophorinae, and we now discuss rela- ary rates and, hence, provide estimatesof the tionships in that group. Given the patterns in amount of genic change along branches;Fel- Figs. 1-3, it seems unlikely that the level of senstein (1978) discussesthese assumptionsfor divergenceamong local populationswould be Wagner trees. sufficient to alter among-speciespatterns es- The UPGMA and WPGMA phenogramshad tablished here. Thus, we doubt that an analysis equal r,,.'s (0.99) and were topologicallysimi- of geographicvariation, not addressedhere, lar; therefore,only the UPGMA phenogramis would alter our conclusions, which follow. Lo- 38 GUTII2RREZ,ZI•C, A•D YA•O [Auk,Vol. 100

5 27 5 27 7• pall/•Y/c/nctu$ 522 $.22 -.01 .70{ F• colch/cu$ 2.94 •. chu•ar 302 .49

301 3•3 .•7C coturn/x 21.4•1•AB }Coturnl• ]L .11 A •

1.26

F-M

%SD= 5.76 Lneg • ] I •3 O.p/c/us z- M 254 .12A %SD = 7.29 2 negs

Fig. 3. Fitch-Margoliash(F-M) treesbased on Rogers'D-values. Branch lengths in Rogers'D (x10). These two trees,of fourexamined, best summarized the originalmatrix, as judged by thevalues of the percentage of standarddeviation and the numberof negativebranches. These trees, as well asthose in Figs.1 and 2, showthat alternativehypotheses exist regarding the branchingsequence of the five subfamilies. phortyx gambelii and L. californicus,the only Sibley and Ahlquist 1982, Avise et al. 1980c). congenersin our study, consistentlycluster to- We believe that the most reasonablecompari- gether--the only genetic differenceswe found sonsinvolve congeneric,interspecific levels of between them are minor gene frequency dif- differentiation, when comparingacross verte- ferences(Appendix 2). Lophortyxand Callipe- brate classes(Zink 1982). Avian specieslimits pla are the most similar pair of genera in our are usually clearly defined, whereas higher sample of Odontophorinae.The other genera taxonomiccategories are far more arbitrary, es- in the subfamily, Colinus,Oreortyx, and Cyr- pecially across vertebrate classes.Thus, it is tonyx, are each genically distinct, as evidenced important to note that passerinecongeners by the level of separationon phenogramsand show little genic differentiation. Research branch lengthson F-M and Wagner trees. At should address this problem rather than dif- aGpd-1, all members of the Odontophorinae ferencesbetween intergeneric or familial levels are fixed for the same, apparently derived al- of genic divergenceacross vertebrate classes. lele. Various groupsof taxa in the subfamily Unfortunately, our only congenericcompari- are fixed for alleles not found in any non- son, L. gambelii-L. californicus, is between Odontophorine taxon at Udh, Pept-1, Pept-2, probablesibling species. and Est-2. The relationshipsamong P. colchi- Within a major vertebrate classes,compari- cus, C. coturnix,C. virginianus,L. gambelii,and sons of equivalent taxonomiclevels above the L. californicusare similar to thosesuggested by species level may be more appropriate. The Jolleset al. (1979)based on lysozymesequence demonstration(Barrowclough and Corbin 1978, data. Barrowcloughet al. 1981,Zink 1982,this study) that avian genetic distances increase as the DISCUSSION taxonomic unit compared is more inclusive suggeststhat the taxonomichierarchy reflects Levels of genetic differentiation.--It is fairly biological, or phylogenetic, units. We found well established that genetic differentiation that, at the genericand family levels,the gal- among passerinetaxa is low relative to other liforms are considerably more differentiated vertebrates(Avise et al. 1980c,Barrowclough than passerinetaxa (Table 2). We clearly re- et al. 1981). The reasonsfor this are unclear (see quire more comparisonsof nonpasserinetaxa January1983] GenicVariation in Galliforms 39 before we will understand whether or not non- rived characterstates (synapomol'phies) sup- passerinespresent a differentpattern from that port the monophylyof the Odontophorinaeand of passerines.For example,the genetic dis- each of the two subgroups. tanceswe observedmay be amongthe highest Of interest here is Cyrtonyxcooki, an extinct found in nonpasserines,but this remainsto be speciesfrom the mid-Miocene [16 million yr documented. beforepresent (MYBP)] of Nebraska(Brodkorb Rates and datesof divergence.--Thebranch 1964)and a congenerof a speciesexamined by lengthsof the Wagner(Fig. 2) and F-M (Fig. 3) us (C. montezumae).We assume here that this treescan be interpretedas roughestimates of fossil belongs to the monophyletic lineage "rates"of genicdivergence, thereby indicating Cyrtonyx,and it is neitherfrom the Odontopho- lineagesthat havechanged faster or slowerrel- rine stockthat pre-datedthe -Odon- ative to other lineages.Rates are averagesacross tophorussplit, nor is it a primitive (pleiso- loci. The branch lengths (rates)in Figs. 2 and morph) early member of the Dendrortyxgroup. 3 appearhomogeneous, although with missing Thus, the age of C. cooki can be taken as a extant taxa this is difficult to judge, because conservativeestimate of the age of the Cyrto- additional taxamight have evolvedat different nyx lineage, representedin our study by C. rates than the taxa we sampled.The branch montezumae.The averageof D-values from C. lengths (Fig. 2) from the "most recent common montezumaeto its sister taxa in the Odontoph- ancestor" to members of the Odontophorinae orinae (Fig. 1, Table.I), 0.609, is assumedto range from 2.29 (to C. squamata)to 2.83 (to ¸. representminimally 16 million yr (MY). This pictus),and the mean(+SD) equals2.59 + 0.17; resultsin the following conversion:t = 26.3 x thus, we suggestthat evidenceof rate hetero- 106D. We use this calibration to indicate pos- geneity is lacking among thesetaxa. sibledates of divergenceamong the taxa shown Nei's measure of genetic distance can be in Fig. 1. We stressthat this is a rough estimate converted into approximate dates of diver- (but probablyconservative), owing to the vari- gencebetween taxa (Nei 1975).Nei suggested anceof geneticdistances between taxa. It is the a "theoretical" conversion of t = 5 x 106D, first suchapproximation for an avian taxon. where t is time since divergencefrom a com- Holman (1961) discusseda fossil, Lophortyx mon ancestor, and D is Nei's D-value. Yang shotwelli,from Umatilla County, Oregon. This and Patton (1981) used this conversion to es- specimenis associatedwith mammalian re- timate divergence dates among Galapagos mains from the Hemphillian stage,dated at 6 finches.As Yang and Patton and Avise et al. MYBP (Savagepers. comm.). Holman noted that (1980c) noted, other attempts to calibrate this specimenpossessed several characters un- D-values (e.g. Sarich 1977) suggestthat Nei's like modem Lophortyx,and therefore its po- calibration may be low (i.e. too rapid) by a sition in the evolutionaryhistory of Callipepla factor of four. Workers with other vertebrate and Lophortyxis uncertain. If this fossil is a groups (e.g. Maxson and Maxson 1979) have "good" member of either Callipeplaor Lophor- used the fossil record and independent esti- tyx, it would greatlyalter our calibrationof ge- mates from microcomplementfixation studies netic distances. For instance, our estimated date to calibrate electrophoreticdistances and di- of the divergence of Callipeplaand Lophortyx, vergencetimes, and such studieshave tended 2.8 MYBP, would necessarilybe •>6 MYBP, or to supportSarich (1977). twice our presentestimate. Prager et al. (1974) The fossilrecord of the galliforrnsallows what suggestedthat phasianoidtransferrins evolved amountsto the first independentcalibration of at a rate of 0.97/MY. Prager and Wilson (1976) avian genetic distances.Several assumptions gave a transferrinimmunological distance of 65 are made in our calibration. First, based on our for P. colchicus-C.virginianus, or a divergence reading of Holman's (1961, 1964)extensive os- date of 63 MYBP. Our data (D = 1.34) suggest teologicalanalyses, we assumethat the Odon- a divergencedate of 35 MYBP. Unfortunately, tophorinaeis a monophyleticgroup consisting this is the only comparisonin common be- of two subgroups: the ¸dontophorus group tween our studyand Pragerand Wilson's(1976). (consistingof ¸dontophorus,Dactylortyx, Cyr- It is of interest that their estimate of the diver- tonyx, and Rhynchortyx)and the Dendrortyx gencedate for thesetwo taxais twice ours and group (containingDendrortyx, Philortyx, ¸re- in the same direction as our estimate would be ortyx, Colinus,Callipepla, and Lophortyx).De- if L. shotwelliwas a valid Lophortyx.Clearly, 40 GUTIERREZ,ZINK, AND YANG [Auk, Vol. 100

:•'::- Oreortyxp/•tu$ Lophort,vxcah'form•u$ :• Lophort,vx•arnbeh'/ :"'..:.::•Calh•epla squarnata • Colinu$wi•ianu$

Fig.4. Approximatebreeding distributions (excluding introductions) of Oreortyx pictus, Colinus virgin- ianus,Callipepla squamata, Lophortyx gambelii, and L. californicus.These species are all representatives of the subfamilyOdontophorinae. Ranges taken from Leopold et al. (1981). the discrepancyin our estimatesdeserves fu- middleOligocene of Colorado,and the lower ture attention, as does the phylogeneticposi- Miocene (approximately20 MYBP) of South tion of L. shotwelli. We conclude at this time Dakota(Brodkorb 1964). This suggestsa fairly that Nei's (1975)conversion factor given above widespreaddistribution of ancestralOdon- is probablylow by a factorof five for the gal- tophorinestock. Therefore we will applymeth- liforms studied here. We note also that many odsfrom vicarlance biogeography (Nelson and assumptions,such as phylogenetic hypotheses Platnick1981) to generateour biogeographic of fossil and recent forms, need to be tested. hypothesis.We assumefirst that the Dendror- Biogeographyof the Dendrortyx group of the tyx and Odontophorusgroups are distinct, Odontophorinae.--Asmentioned above, the monophyleticlineages (discussed above) that Dendrortyxgroup of the Odontophorinaecon- divergedat least16 MYBP(middle Miocene). sistsof the generaDendrortyx, Philortyx, Ore- Next, we assumethat Dendrortyxand Philortyx ortyx,Colinus, Callipepla, and Lophortyx.Al- are primitiveand/or have not affectedthe dis- thoughwe lackDendrortyx and Philortyx in our tribution of the remaining species. The re- geneticanalysis, we will usethe phylogeny of stricted distribution of Dendrortyx(M6xico to the remainingtaxa and our approximate dating Costa Rica) and Philortyx (M6xico) suggests of dadogeneticevents (Fig. 1) to developan relict status and, therefore, "primitiveness." evolutionaryperspective of the biogeography Based on their skeletal morphology, Holman of thesetaxa. Without an objectiveestimate of (1961) concluded that these two genera were phylogeneticrelationships, it wouldbe diffi- the mostprimitive members of the Dendrortyx cult to evaluate historical evolutionary patterns group. These assumptionsobviously need amongthese New Worldquail given simply a testing. mapof theircurrent distributions (Fig. 4). The remaining stepsin our biogeographic The earliestknown fossil Odontophorinae are hypothesisare outlined diagrammatically in Fig. fromthe earlyOligocene of Saskatchewan,the 5. Thegenetic data indicate that O. pictusorig- January1983] GenicVariation in GalIiforms 41

inated about 12.6 MYBP (Fig. 5a) and probably evolved in western North America, based on iii•0A. 12.6 MYBP/, '• its current distribution (Fig. 4). Perhapsdrying reortyx / 'pre'-' Coh•us- p/ctus / C•///•oep/•- Lophortyx trends at this time (Axelrod 1979) resulted in the invasion of more mesic environments of / / higher elevation by Oreortyx. The southward displacement of land west of the San Andreas Fault at this time might also have isolated this taxon (see Wenner and Johnson1980). Another east-westsplit of an ancestraltaxon resulted in the divergence of Colinus(Fig. 5b), estimated by our data at 7.0 MYBP. The earliest known fossils of Colinus(Holman 1961, Brodkorb 1964) are from the upper Plioceneof Kansas.Because our estimated divergencedate is well before upper Pliocene, the sites of fossil Colinusdo not permit identificationof the area of origin of Colinus. Penultimately, Callipeplaand Lophortyxdi- verged (Fig. 5c), about 2.8 MYBP, or late Plio- cene, when their ancestorwas fragmented in the southwestern aridlands of North America. Axelrod (1979)noted that during the late Plio- ceneisolated arid and semi-ariddesert patches existed in the current Sonoran and Chihua- huan desert regions, and this could have al- lowed the allopatricdifferentiation of Callipe- pla, which presently occupiessuch habitats (Leopoldet al. 1981).Species of Lophortyxdi- C...2.8M YB P I / C.v•;wbnus • verged last, resulting in the current distribu- tion patterns(Fig. 4). It will be of value to as- certainthe phylogeneticposition of L. douglasii. Hubbard (1973)proposed a vicariant biogeo- graphicexplanation for the evolutionof species in the generaLophortyx and Callipepla(consid-

consistingof evergreenchaparral, evergreen-broad- leaf forest, or coniferous forest with shrub under- story in western North America. B. Origin of Colinus virginianus,7.0 MYBP. Presenthabitat is essentially Fig. 5. Diagrammatic representation of hypo- weedy fields bordered by brush or woodlots, al- thetical stagesin the evolution of speciesshown in though in tropical lowlandsit occursin wetter con- Fig. 4. Becausehistorical ranges are unknown,ranges ditions. The exactlocation of the origin of C. virgin- should not be interpreted strictly. The primary in- ianus is uncertain. Note that the range of "pre" formationwe wish to conveyis the isolationevents, Callipepla-Lophortyxcould have extendedfurther east which are indicated by the dashedlines. The exact if C. virginianuswas either isolatedto the south or position of the dashedlines is speculative,however. north. C. Divergence of Callipeplasquamata and Lo- Therefore,whether differentiationamong these taxa phortyx, 2.8 MYBP, probably concurrent with late was parapatric or allopatric and the extent of any Pliocene disjunct patchesof arid and semi-arid des- gaps between taxa during their evolution are un- ert. C. squamataand L. gambeliiare typically found known. After each isolation event, dispersalproba- in desert scrub or arid grasslands,while L. califor- bly occurredacross these boundaries. A. Separation nicusoccurs in oak woodland, chaparral,and brushy of Oreortyxpictus, 12.6 MYBP, presumablyin areas foothills. 42 GUTIgRREZ,ZINK, ANDYANG [Auk, Vol. 100 ered by him congeneric),which involved hy- ed that L. garnbeliiand L. californicusare con- potheticalancestors and their distributionsand specific.The genetic data presentedhere are Pleistocene glaciation cycles. According to consistentwith this idea. Wild hybrids (Hen- Hubbard's scenario, a trichotomoussplit pro- shaw 1885, Miller and Stebbins 1964) between duced C. squarnata,L. douglasii,and "pre-cali- the formsinclude only the F1 generation,how- fornicus-garnbelii"in the Illinoian glacialepoch, ever, which suggestsa lack of introgressionand with L. californicusand L. garnbeliidifferen- panmixia in zones of sympatry. Furthermore, tiating in the Wisconsinianglacial period. Us- hybrids have been reported for other sym- ing the geneticdistance conversion established patrix quail [e.g. Colinusvirginianus x L. cali- above, we estimate the split of L. garnbeliiand fornicus(Aiken 1930);Callipepla squamata x L. L. californicusat 190,000 yr ago, and that of gambelii(Bailey 1928); L. californicusx C. squa- Lophortyx and Callipepla at 2.8 MYBP, or rnata (Jewett et al. 1953); C. virginianusx C. roughly late Pliocene. While Hubbard is cor- squarnata(Johnsgard 1973); O. pictus x L. cal- rect in assumingthat a minimum of two geo- ifornicus(Peck 1911)].Incidence of hybridiza- logic events (e.g. glacial-interglacialcycles) is tion is probably not an accuratepredictor of needed to accountfor the distribution patterns closephylogenetic relationship, becausebirds of extant Callipeplaand Lophortyx,these gen- retain the ability to hybridize despite consid- era probably divergedwell before the times he erable genetic divergence (Prager and Wilson suggested.Illinoian age glacialcycles may well 1975). The apparent lack of prereproductive have effected speciation in Lophortyx,how- isolatingbarriers in sympatry,at leastto F1hy- ever. bridization, does not necessarilyindicate re- This biogeographicreconstruction is a hy- cency of common ancestryor close phyloge- pothesis. We note that sympatry of breeding netic relationships among these quail. species, which is considerable among the Therefore, we reject the notion that hybrids speciesstudied here (Fig. 4), implies dispersal between L. garnbeliiand L. californicusprove (Nelson and Platnick 1981). Thus, it is difficult conspecificstatus. As only the F1 hybrids and to determine the relative importance of vicar- not a hybrid swarmhave been found, theseare iance or dispersalin accountingfor the evo- probably distinct biological species.We also lutionary patterns in this group. Also, geolog- point out that the evidence supporting appar- ical and paleobotanicalevidence is sufficiently ent assortativemating in sympatry overshad- fragmentary (see Axelrod 1979, Wenner and ows our finding a D-value typical of subspe- Johnson 1980) to prevent precise correlations ciesbetween these taxa. Clearlydistinct species with our phylogenetichypothesis. That is, we can be genicallysimilar (Avise et al. 1980b,c). lack a well corroborated "area cladogram" Several authors (e.g. Mayr and Short 1970, (Nelson and Platnick 1981). Johnsgard1973, and referencestherein) have Our phylogenetichypothesis and dating of suggestedmerging Lophortyxand Callipepla. cladogeneticevents suggest that a seriesof three Thesegenera are clearly similar (Fig. 1). The• east-west range disjunctions(Fig. 5) could ex- between our samplesof these two taxa, 0.104, plain the evolution of these genera of New is one-halfthat reportedby Barrowclough(1980) World quail. These patterns need to be corrob- for similar avian intergenericcomparisons; yet orated with studies of other avian groups, as it is within the range (0.0126-1.214).Thus, our well as other groups of vertebrates and non- data provide no clear-cutanswers, and we sug- vertebrates.We believe, however, that our hy- gestthat the decisionregarding their taxonom- pothesis(Fig. 5) is an important first approxi- ic status be made on the basis of other kinds mation of the evolutionaryhistory of thesequail of biological evidence. taxa and makes objective, testable predictions The suggestionby Phillips et al. (1964) and that would otherwise be difficult given only Mayr and Short (1970)that Oreortyxbe merged the map of currentbreeding distributions.We with Callipepla (including Lophortyx) is not also wish to demonstrate that molecular meth- consistentwith our moleculardata (Figs. 1-3). ods of inferring phylogeniescan be combined It would alsorequire inclusionof Colinus;oth- with information from the fossil record to fur- erwise, the new taxonwould not be monophy- ther understandingof evolutionarypatterns in letic. We think that merging these taxa would avian groups. obscure their relatively long, independent, Taxonorny.--Mayrand Short (1970) suggest- evolutionary histories. We recognize, how- January1983] GenicVariation in Galliforms 43 ever, that there can be no absolute value of AMERICAN ORNITHOLOGISTS' UNION. 1957. Check- genetic distanceon which to base taxonomic list of North American Birds, fifth ed. Baltimore, decisions. The external phenotypic resem- Maryland, Amer. Omithol. Union. blance of Mountain and does 1973. Thirty-second supplementto the American Ornithologists' Union check-list of not, in this case, indicate a closephylogenetic North American birds. Auk 90: 411-419. relationship. Furthermore, Hudson et al. (1959, AvisE, J. C., J. C. PATTON,& C. F. AQUADRO. 1980a. 1966) and Holman (1961) provide data on the Evolutionarygenetics of birds. I. Relationships myology and skeletal morphology, respective- amongNorth Americanthrushes and allies.Auk ly, that show that Oreortyxis morphologically 97: 135-147. distinct from other membersof the Dendrortyx , --, & --. 1980b. Evolutionary ge- group. Gutierrez (1980)reported that Oreortyx netics of birds. II. Conservativeprotein evolu- was very differentecologically from Lophortyx. tion in North American sparrowsand relatives. Thus, genetic, morphologic,and ecologicdata Syst. Zool. 29: 323-334. show that Oreortyx is not a close relative of , --, & --.. 1980c. Evolutionary ge- netics of birds. Comparative molecularevolu- Lophortyxor Callipepla,and there seemsto be tion in New World warblers and rodents. J. He- no basis for their generic merger. redity 71: 303-310. The systematicstatus of C. virginianusis un- AXELROr),D. I. 1979. Age and origin of Sonoran certain. It is shown here to be a distinct clade desertvegetation. Occ. Papers,Cal. Acad. Sci. within the Odontophorinae and more similar to Lophortyxand Callipeplathan to Oreortyx AYALA, F.J. 1975. Genetic differentiation during and Cyrtonyx. the speciationprocess. Evol. Biol. 8: 1-78. Some authors(Brodkorb 1964, Mayr and Short BAILEY,V. 1928. A hybrid Scaledx Gambel'squail 1970,Johnsgard 1973) consider the Tetraonidae from New Mexico. Auk 45: 210. a subfamily of the Phasianidae. Our data (Ta- BARROWCLOUGH,G. F. 1980. Genetic and pheno- ble 1) show that the distance from T. pallidi- typic differentiation in a wood warbler ( Dendroica)hybrid zone. Auk 97: 655-668. cinctus(Tetraonidae) and P. colchicus(Phasian- --, & K. W. CORBIN. 1978. Genetic variation idae) to all other species is similar. Thus, it and differentiation in the Parulidae. Auk 95: 691- might be appropriateto considersuch a taxo- 702. nomic scheme. --, & R. M. ZINR. 1981. Genetic dif- ACKNOWLEDGMENTS ferentiationin the .Comp. Bio- chem. Physiol. 69B: 629-632. We thank the following individuals for assisting BRODKORB,P. 1964. Catalog of fossil birds: part 2 in various ways with this project:John Davis, Rich- (Anseriformesthrough ). Bull. Flor- ard D. Sage,Vincent Sarich, Don Straney,Allan C. ida State Mus. Biol. Sci. 8: 195-335. Wilson, and J. W. Wilson. David E. Brown, Richard CORBIN,K. W. 1981. Genic heterozygosityin the Brown, Larry Caudill, Michael Erwin, Jim Gerber, White-crownedSparrow: a potential index to Dale Jones, A1 Woodard, and Tom Zapatka either boundaries between subspecies.Auk 98: 669- helped collectspecimens or provided useful locality 680. information. Ramone Baccus, George F. Barrow- F^RmS,J. S. 1972. Estimating phylogenetic trees clough,John E. Cadle, Kendall W. Corbin, Ned K. from distance matrices. Amer. Natur. 106: 645- Johnson,Allan Larson, A. Starker Leopold, Jill A. 668. Marten, Gareth Nelson, James L. Patton, Duke S. FELSENSTEIN,J. 1978. Cases in which parsimony Rogers,Charles G. Sibley, and David W. Steadman or compatibilitymethods will be positivelymis- providedhelpful comments on variousdrafts of this leading. Syst. Zool. 27: 401-410. paper. Julie Federassisted in the lab, and JeanneA. FITCH, W. g., & E. MARGOLIASH. 1967. Construc- Dawson typed the paper. Gene Christmandrew the tion of phylogenetictrees. Science 155: 279-284. figures.Dr. Donald E. Savageprovided dateson fos- GUTIf•RREZ,R. J. 1980. Comparativeecology of the sil localities. A Ford Foundation Fellowship to the Mountain and California quail in the Carmel first author, the Museum of VertebrateZoology, and Valley, California. Living Bird 18: 71-93. the Union Foundation Wildlife Fund furnished fi- HENNIG,W. 1966. Phylogeneticsystematics. Chi- nancialsupport. RMZ was supportedby NSF grant cago, Illinois, Univ. Illinois Press. DEB-7920694 to N. K. Johnson. HENSHAW,H. W. 1885. Hybrid quail (Lophortyx gambeliix L. californicus).Auk 2: 247-249. LITERATURE CITED HOLMAN,J. A. 1961. Osteologyof living and fossil AIREN, C. E. H. 1930. A Bobwhite x California New World (Aves, Galliformes).Bull. Florida State Mus. Biol. Sci. 6: 131-233. Quail hybrid. Auk 47: 80-81. 44 GUTff•RREZ,ZINK, ANDYANG [Auk, Vol. 100

1964. Osteology of gallinaceous birds. The birds of Arizona. Tucson, Arizona, Univ. Quart. J. Florida Acad. Sci. 27: 230-252. Arizona Press. HUBBARD,J. P. 1973. Avian evolution in the arid PRAGER, E. M., & A. C. WILSON. 1975. Slow evo- lands of North America. Living Bird 12: 155-196. lutionary loss of the potential for interspecific HUDSONßG. E., P. J. LANZILLOTTI,& G. D. EDWARD. hybridization in birds: a manifestation of slow 1959. Muscles of the pelvic limb in galliform regulatory evolution. Proc. Natl. Acad. Sci. 72: birds. Amer. Midl. Natur. 61: 1•37. 200-204. ßR. A. PARKER,J. VANDENBERGE, & P. J. LAN- ß& . 1976. Congruencyof phylogenies ZrLLOTTI. 1966. A numerical analysis of the derived from different proteins. J. Molec. Evol. modifications of the appendicular muscles in 9: 45-57. variousgenera of gallinaceousbirds. Amer. Midi. , A. H. BRUSH, R. A. NOLAN, M. NAKANISHI, Natur. 76: 1-73. & A. C. WrLSON. 1974. Slow evolution of JEWETT,S. G., W. P. TAYLOR, W. T. SHAW, & J. W. transferrin and albumin in birds accordingto ALDRICH. 1953. Birds of WashingtonState. Se- micro-complement fixation analysis. J. Molec. attle, Washington, Univ. Washington Press. Evol. 3: 243-262. JOHNSGARD,P.A. 1973. Grouseand quails of North PRESCH,W. 1979. Pheneticanalysis of a single data America. Lincoln, Nebraskaß Univ. Nebraska set: phylogenetic implications. Syst. Zool. 28: Press. 366-371. JOHNSONßM. S., & J. L. BROWN. 1980. Genetic vari- ROGERS,J. S. 1972. Measuresof geneticsimilarity ation among trait groups and apparent absence and genetic distance. Studies in Genetics VII. of close inbreeding in Grey-crowned Babblers. Univ. Texas Publ. 7213: 145-153. Behav. Ecol. Sociobiol. 7: 93-98. SARICH,V. M. 1977. Rates, sample sizes, and the JOLL•S, J., I. M. IBRAHIMI, E. M. PRAGER, F. neutrality hypothesisfor electrophoresisin evo- SCHOENTGEN,P. JOLLES,& A. C. WILSON. 1979. lutionary studies. Nature 265: 24-28. Amino acid sequence of pheasant lysozyme. SELANDER, R. K., M. H. SMITH, S. Y. YANG, W. E. Evolutionarychange affecting processing of pre- JOHNSON, & J. B. GENTRY. 1971. Biochemical lysozyme. Blochem. 13: 2744-2752. polymorphism and systematicsin the genus LEOPOLD,A. S., R. J. GUTIERREZ,& M. T. BRONSON. Peromyscus.I. Variation in the old-field mouse 1981. North American game birds and mam- (Peromyscuspolionotus). Studies in GeneticsVI. mals. New York, Charles Scribner's Sons. Univ. Texas Publ. 7103: 49-90. MAXSON,L. R., & R. D. MAXSON. 1979. Compar- SIBLEY, C. G., & J. E. AHLQUIST. 1982. The rela- ative albumin and biochemical evolution in tionships of the Yellow-breasted Chat (Icteria vi- plethodontid salamanders. Evolution 33: 1057- rens) and the alleged slowdown in rate of mac- 1062. romolecular evolution in birds. Postilia 187: 1- MAYR, E., & L. L. SHORT. 1970. Speciestaxa of 18. North American birds. Publ. Nuttall Ornithol. SNEATH, P. H. A., & R. R. SOKAL. 1973. Numerical Club 9: 1-127. . San Francisco,W. H. Freeman & Co. MrLLER, A. H., & R. C. STEBBINS. 1964. The lives TEMPLETON,A. R. 1980. Modes of speciation and of desert in Joshua Tree National Mon- inferencesbased on geneticdistances. Evolution ument. Berkeley,California, Univ. Calif. Press. 34: 719-729. NEI, M. 1975. Molecular population genetics and WAKE,D.B. 1981. The applicationof allozymeevi- evolution. Amsterdam, North Holland. dence to problems in the evolution of morphol- 1978. Estimationof averageheterozygosity ogy. Pp. 257-270 in Evolution today (G. G. E. and geneticdistance from a smallnumber of in- Scudder and J. L. Reveal, Eds.). Proc. 2nd Inter- dividuals. Genetics 89: 583-590. natl. Congr. Syst. Evol. Biol. NELSON,G., & N. G. PLATNICK. 1981. Systematics WENNER, A.M., & D. L. JOHNSON. 1980. Land ver- and biogeography.New York, Columbia Univ. tebrates on the California Channel Islands: Press. sweepstakesor bridges:Pp. 497-530in The Cal- NEro, E. 1978. Genetic variation in natural pop- ifornia Islands: proceedingsof a multidiscipli- ulations:patterns and theory. Theor. Pop. Biol. nary symposium(D. M. Power, Ed.). Santa Bar- 13: 121-177. bara, California, Santa Barbara Mus. Nat. Hist. PECK,M. E. 1911. A hybrid quail. Condor 13: 149- YANG, S. Y., & J. L. PATTON. 1981. Genic variabil- 151. ity and differentiationin the Galapagosfinches. PETERS,J. L. 1934. Check- of the world, Auk 98: 230-242. vol. 2. Cambridge,Massachusetts, Harvard Univ. ZINK, R. M. 1982. Patterns of genic and morpho- Press. logic variation among sparrowsin the genera PHrLLrPS,A. R., J. MARSHALL, & G. MONSON. 1964. Zonotrichia,Melospiza, lunco, and Passerella.Auk 99: 632649. January1983] GenicVariation in Galliforms 45

APPENDIX1. Location,sample sizes, and samplecollection dates. Different populationsof the sametaxon are labeledby upper caseletters, which correspondto the lettersused in Appendix2, Table 1, and the figures. The taxonomicframework is from Peters (1934).

Number of indi- Taxon viduals Locality and date

Tetraonidae 13 New Mexico: 8 mi E Milnesand, Roosevelt Co., Decem- (Tympanuchuspallidicinctus ) ber 1975. Phasianidae Ring-neckedPheasant 13 California: Webb Tract, Sacramento Delta, Sacramento (Phasianuscolchicus ) Co., November 1975. 30 Avian SciencesDepartment, University of California, (Coturnix coturnix) Davis, California, A--Big Brown Strain (n = 10). B-- Small Brown Strain (n = 10). C--Albino Strain (n = 10). A, B, and C obtained in February 1976. Chukar 12 Avian Sci. Dept., Univ. Calif., Davis. (n = 10), Febru- (Alectorischukar) ary 1976;5 mi SE Panoche,San Benito Co., California (n = 2), Jan. 1976; samplescombined. Odontophorinae DendrortyxGroup California Quail 36 California: A--5 mi E Shahdon, San Luis Obispo Co. (Lophortyxcalifornicus ) (n = 6), December 1974. B•5 mi N Jolon, Monterey Co. (n = 13), December 1974. C•4 mi E Mercy Hot Springs,Fresno Co. (n = 9), October1975 and (n = 2) in January1976; 5 mi SE Panoche,San Benito Co. (n = 6), January1976. Gambel's Quail 22 New Mexico: 1 mi E Columbus, Luna Co. (n = 19), De- (L. gambelii) cember 1975. Arizona: 10 mi E Green Valley, Pima Co. (n = 3), January 1976; samplescombined. 29 New Mexico: A•8 mi E Milnesand, Roosevelt Co. (n = (Callipeplasquamata ) 7), December 1975. B--1 mi E Columbus, Luna Co. (n = 20), December 1975. Arizona: 10 mi E Green Valley, Pima Co. (n = 2), January1976; combined with population B as there was no differentiation. Bobwhite Quail 15 New Mexico: A--8 mi E Milnesand, Roosevelt Co. (n = (Colinusvirginianus ) 8), December 1975. B--18 mi NE Milnesand, Roose- velt Co. (n = 7), December 1975. 16 California: 2 mi SE Jamesburg,Monterey Co. (n = 16), (Oreortyxpictus) June-August 1975. OdontophorusGroup Montezuma Quail 31 Arizona: A--15 mi E Patagonia, Santa Cruz Co. (n = (Cyrtonyxrnontezumae ) 23), January1976; B--10 mi W Patagonia,Santa Cruz Co. (n = 8), January 1976. 46 GUTIERREZ,ZINK, ANDY^NG [Auk, Vol. 100

APPEND/X2. Allelic frequenciesfor 23 presumptiveloci in 10 speciesof gallinaceousbirds. Numbers in parenthesesare frequenciesof allelesat a locus. A singleletter denotessample fixed for that allele. Full names for loci given in Yang and Patton (1981) or below. Names of taxa and samplesizes given in Ta- ble 1.

Locus T.p. P.c. C.c. A C.c. B C.c. C A.c. L.g. L.c. A PGI b c d (0.65) d (0.60) e c b b e (0.35) e (0.40) ADA • g b d (0.35) f f e a (0.42) a (0.06) f (0.65) c (0.58) c (0.94) PGM-1 a c d d d d b b

MPI h c f e (0.60) f a e e (0.94) f (0.40) g (0.06)

IDH-1 e d c c c d a (0.97) a f (0.03) IDH-2 g b g g g a (0.30) f f c (0.70) c•GPD-1 d c e e e a b b c•GPD-2 c c d d d a b b

ME•' g d (0.12) a (0.25) a (0.45) a (0.40) f f e (0.03) e (0.88) b (0.75) b (0.55) b (0.60) f (0.97) GOT-1 i j h h h a (0.45) g (0.97) g b (0.05) i (0.03) c (0.50) GOT-2 a b b b b a b b GDH •' a c b b b b b b UDH ci d d d d d d c c SDH a e a a a b (0.15) d d c (0.85) LDH-1 a a c c c b b b LDH-2 a a c c c b b b Pept-1 g a b b b a c c

Pept-2 d d e e e a b b Alb g c d (0.95) c (0.40) c b e c (0.12) f (0.05) f (0.60) e (0.88) Est-1 a c (0.08) e d (0.60) e f f (0.05) f (0.03) d (0.92) e (0.40) g (0.95) g (0.97)

Est-2 g g (0.38) f f f e b a (0.03) h (0.62) b (0.97) 6PGD d g f (0.85) e (0.05) f (0.75) c (0.10) d b (0.03) h (0.15) f (0.80) h (0.25) d (0.90) d (0.97) h (0.15) XDH • g f a (0.30) b (0.05) a (0.20) c a (0.08) a (0.03) c (0.70) c (0.95) c (0.80) c (0.92) c (0.97) Percentage poly- morphism 0.0 11.1 22.2 25.9 11.1 14.8 18.5 29.6 Hetero- zygosity 0.0 2.6 4.8 8.9 4.8 5.2 2.5 2.8 Mean number alleles per poly- morphic locus 2.0 2.0 2.1 2.0 2.2 2.0 2.0

"Adenosene de-aminase. t'Malic enzyme(NADP dependent). • Glutamatedehydrogenase. d Unidentifieddehydrogenase. • Xanthine dehydrogenase. January1983] GenicVariation in Galliforms 47

APPENDIX 2. Continued

L.c. B L.c. C C.s. A C.s. B C.v. A C.v. B O.p. C.m. A C.m. B

b b b b b b a b b

c c a a (0.75) a (0.56) a (0.50) e (0.97) c c c (0.25) c (0.44) c (0.50) g (0.03) b b (0.96) c c c c b c c d (0.04) e (0.83) e (0.92) e e (0.75) d d b g (0.91) g (0.94) g (0.17) c (0.04) g (0.25) i (0.09) i (0.06) g (0.04) a a a a a a (0.93) a b b c (0.07) f c (0.11) f f f f f d (0.04) e (0.94) f (0.89) e (0.96) g (0.06) b b b b b b b b b b b b b b b b b (0.85) b (0.87) d (0.15) d (0.13) f f f f b (0.06) e c e e e (0.94) g g (0.92) g (0.71) f (0.05) d (0.06) g g g e (0.13) i (0.08) j (0.29) g (0.70) g (0.94) g (0.87) j (0.25) b b b b b b b b b b b b b b b b b b c c c c c c b a a d d d d d d b (0.63) b b d (0.37) b b b b b b b b b b b b b b b b b b c c c c (0.97) d d c (0.06) e e f (0.03) f (0.94) b b b b b b b c c c (0.08) c (0.15) e e a a e a a e (0.92) e (0.85) g g g g h h f a (0.83) a (0.87) b (0.15) b (0.13) c (0.02) b b b b b b s c b (0.19) c (0.81) d d d d d b (0.07) d d (0.96) d d (0.93) f (0.04)

c c a (0.07) a (0.10) a (0.94) a (0.86) f e e d (0.93) d (0.90) e (0.06) e (0.14)

7.4 18.5 7.4 18.5 14.8 14.8 11.1 18.5 22.2

1.8 3.4 1.6 3.7 2.8 2.6 2.1 2.4 5.1

2.0 2.2 2.0 2.2 2.0 2.0 2.0 2.2 2.0 NEST ATTENDANCE OF PARENT BIRDS IN THE PAINTED SNIPE (ROSTRATULA BENGHALENSIS)

SHIGEMOTO KOMEDA Laboratoryof AnimalSociology, Faculty of Science,Osaka City University, Sugimoto,Sumiyoshi-ku, Osaka 558, Japan

ABSTRACT.--Iobserved 15 pairs of Painted Snipes (Rostratulabenghalensis) in Japanfor 24-h periods in order to compare the nest attendance behavior of males and females and to documentchanges in their behavior during the laying period. From a few days before the laying of the first egg until the day the secondegg was laid, the male and female consorted with eachother (stayedwithin 5 m of each other) more than 90% of the observationtime. They visited the nest repeatedlyon the first and seconddays of laying. On the third day, the consortingratio decreasedto 58.5%, and the male typically started to incubate. He continuedto incubate that night. After laying the third egg, the female usually did not visit the nest exceptto lay the fourth and last egg. The male continuedincubation, and the female met her mate only when she laid the last egg. The absenceof female incubationand the early termination of consortingbehavior are consistentwith the possibility that Painted Snipe femalesare polyandrous.Received 12 January1982, accepted4 October1982.

AMONG the polyandrous charadriiform qualis), and cudweed (Gnaphaliumaffine), grew species, comprehensive studies of pair-bond sparselyin theseabandoned fields. The water depth maintenancehave been conductedon only four in this area varied from 0 to 15 cm. species, the Northern (Jacana spinosa; The main study was conducted in Toyo City, northwestern Shikoku Island, Japan (133ø03'E, Jenni and Collier 1972, Jenni and Betts 1978), 33ø57'N),from March to August1979 (excluding May) the SpottedSandpiper (Actitisrnacularia; Hays and from June to July 1980 (Toyo area). This 40-ha 1972, Oring and Knudson 1972, Maxson and study area of reclaimedland facesthe JapaneseIn- Oring 1980),the Red-neckedPhalarope (Phal- land Sea. A quarter of the area was used as paddy aropuslobatus; Hild6n and Vuolanto 1972),and fields, and the rest was marshy grasslandthat was the Red (Phalaropus fulicaria; Scha- plowed onceor twice a year, usuallyin Juneand July. mel and Tracy 1977).Although a detailedstudy The study area was traversedby six parallel roads of the Painted Snipe (Rostratulabenghalensis) 200m apartand by anotherroad that intersected them has not been conducted, sex role reversal and at right angles.This poorly drained areawas always an apparentlyunbalanced sex ratio suggestthe covered with water. Until it was plowed, dense reed (Phragmitescommunis) or bulrush (Scirpusplanicul- possibility of polyandry in this species(Beven mis) grew there every year. After it was plowed, the 1913, Baker 1935, Lack 1968, Hays 1972, Oring grassesgrew only patchilyor sparselybefore Sep- and Knudson 1972,Jenni 1974).The objectives tember. The Toyo area is about 300 km west of the of this study were (1) to compare the nest at- Biwa area. tendance of male and female Painted Snipes, Most data for the field study were obtained by con- (2) to measurethe amount of time the members tinuous observations. Observations that started after of the pair spend together (consortingratio) 1800 are identified as "nocturnal observations" and during laying and incubationperiods, and (3) were made with the aid of a flashlight.Behavior was to considerthe evidence for polyandry. observedwith binoculars(9x) and a spotting scope (15x). Behavioral actswere recorded in seconds,and the data were later converted to minutes for calcu- STUDY AREAS AND METHODS lations. Of 27 birds observed, 4 males and 2 females Field observations were made in two areas. The were marked with neck bands. Because no other birds preliminary study was done in Chuzu Town, south- of the same sexes as the banded individuals were east of Lake Biwa, central Japan(136ø02'E, 35ø08'N), observedentering their nests, any male or female from May to June1978 (Biwa area). Nest observations who entered a nest was consideredthe progenitor of were conductedin 20-ha abandonedfields among the eggsin the nest. paddy fields at an elevation of about 85 m. Marshy All observationsreported here were made at nests plants, mainly buttercup(Ranunculus quelpaertensis), with four eggs.The four eggswere laid on consec- bitter cress(Cardamine lyrata), foxtail (Alopecurusae- utive days,and the daysof layingwere identifiedas 48 The Auk 100: 48-55. January1983 January1983] NestAttendance in Painted Snipes 49

T^BLE1. The lengthof visitsto the nestduring the layingperiod (minutes + SD).Numbers in parentheses indicate nocturnal visits.

Day I Day 2 Day 3 Day 4

Female Lengthof egg-layingvisits 35.5 + 17.5 62.3 + 33.4 32.8 + 7.4 15.9 + 1.9 Number of egg-layingvisits 2 4 5 6 Length of other visits 3.6 + 3.7 6.1 + 5.2 5.7 + 5.3 I -- Number of other visits 43 50 36 I

Male Length of visits 3.6 + 8.0 8.0 + 12.8 22.6 + 31.7 43.4 + 52.8 (21.6 + 15.1) (36.1 + 10.2) Percentageof stayslonger than 2.0 6.8 26.2 46.8 30 min (35.3) (76.9) Number of visits 99 85 92 47 (17) (13) Number of nests 4 7 9 8 (1) (1)

day 1, day 2, and so on. Thesefour dayscomprise During the laying period, a female laid an egg the layingperiod. A spanof a few daysbefore day every morning between 0700 and 1100(Fig. 1). I is referred to as the prelaying period. The incu- She stayedon the nestfar longerwhen coming bationperiod extends from the day afterday 4 until to lay an eggthan at other times(Table 1). The hatching. female was on the nest for a much shorter du- The percentageof time during which the two ration when she laid the fourth egg (2 = 15.7 membersof a pair stayedwithin 5 m of eachother was termedthe consortingratio. Consortingbehav- min) than when she laid the earlier eggs (2 = ior was studiedat the Toyo areafor a total of 17 days 44.0 min). All other visits to the nest by the (162.0h). Other observationswere madeat both the female during the laying period were much Toyo and Biwa areas. shorter(2 = 5.1 min); 88.5% of thesevisits were Nest-related behavior in the Biwa area was ob- less than 10 min. served from blinds about 10 m from the nests. Four On four additional days when the 1st, 2nd, nestswere observedfor a total of 17 days (190.8 h). or 3rd eggs were laid but egg laying was not Observationswere made at the Biwa area during two observed,females stayed on the nest for 39, 42, nights(22.6 h) of the incubationperiod. All diurnal 46, and 47 min between 0800 and 1210 (Fig. 1; observationsat the Toyoarea were made from a blind 22 June 1979, 21 June 1978, 3 June 1978, and 12 on a carroof 1.5 m abovethe ground.By driving the car alongthe roads,I was able to follow the birds June 1979, respectively).These four stayswere whenthey left their nests.Nine pairswere observed the only long stay on thosedays. It is assumed at the Toyoarea for a total of 36 days(368.7 h) Two that egg laying occurredduring theselong vis- of thesenine pairsabandoned their nestsduring the its. Egg laying was not seen at two nests on layingperiod, so the datafrom thesenests were ex- day 4, but the females entered the nests only cluded from the results(total 7 days).Two nestswere once, and it is assumed that they laid during observeda total of four nights (47.0 h) during the thesevisits (Fig. 1; 30 March 1979 and 21 July layingperiod. Activity at night at the Toyoarea was 1979). observed at intervals of 1 min from about 20 m when Nestvisitation on eachlaying day.•On day 1, a flashlightwas switched on for about5 s and off for about 55 s. both sexesentered the nest frequently during the morning (Fig. 1), but they often left it for

RESULTS a long time during the afternoon. They re- turned to the nest in the evening. The male Clutch size and time of laying.•Of 33 nests stayedabout the samelength of time (Table 1) found, 2 had 3-egg clutches,30 held 4 eggs, and enteredthe nest about as frequentlyas the and only 1 held 5 eggs.Accordingly, the nor- female did. No observations were made on the mal clutch size in this speciesis consideredto night of day 1. be four. On day 2, the duration of diurnal visits in- The time of laying was recorded 17 times. creased but the number of visits decreased 50 $HIGEMOTOKOMEDA [Auk,Vol. 100

TIME OF DAY DAYTIME ACTIVITY 6 9 12 15 18 DAY DATE NEST ,

I 12• JUN '78'79 62 2 AUG ' 79 9 9 .N.JG '79 I0 4JUN '78 2 15 JUN '78 :• 20 JUN '78 4 28 MAR ' 79 5 13 JUN ' 79 6 22 JUN '79 7 3' AUG ' 79 9 I5 MAY '78 I 5 JUN '78 2 16 JUN '78 3 21 JUN '78 4 29 MAR '79 5 I 4 JUN ' 79 6

23, JUN '79 7

4 AUG '79 9 I I AUG '79 I I 16MAY '78 I 6 JUN '78 :• 17 JUN '78 '• 22 JUN '78 4 :30 MAR '79 5 2[ JUL '79 8 5 AUG '79 9 12 AUG '79 I I

NIGHTTIME ACTIVITY 18 21 0 3 6 DAY DATE NEST 2 i8-19 JUN'80 14 2 24-25JUL '80 15 • 25-26JUL '80 15 4 26-27JUL '80 15

Fig. 1. Activity during laying period. Male is shown in upper and female in lower belt of each pair. Blackband shows time spenton nestand shadedband out of nest.Small dot shows the time of egglaying. Day 1 is the first day of egglaying and day 4 the last day. slightly in both sexes.In two nestswatched at returned at 0443. At Nest 14 the male and fe- night, both sexes left the nest site just after male were gone when observationsstarted at dark and returnedjust beforelight. At Nest 15 2100 and returned to the vicinity (within 5 m) the male and female left the nest at 1907 and of the nest at 0426. January1983] NestAttendance in PaintedSnipes 51

TABLE2. The percentageof daylight hours during N,,4 which the male and female of a pair stayedwithin 5 m of each other. t N,,7 Number of Num- hours ber I N-8 ob- of Average (range) served pairs a'o 4'0 do 8b o'o Prelaying % OF TIME period 99.6 (98.5-100.0) 23.6 4 Fig. 2. Average percentageof daylight time on Laying period nest. Upper belt shows male stays and lower belt Day 1 99.0 (97.5-100.0) 34.0 3 female stays. Blackband shows percentageof time Day 2 92.8 (89.3-95.9) 42.6 4 spent on nest and white band out of nest. Number Day 3 58.8 (38.4-81.0) 29.7 3 of nests observed is shown. Day 4 7.9 (3.3-16.7) 32.1 3

On day 3, both sexeschanged their patterns day 4 she stayedon the nest much longer than of nest visitations, but in different ways. Fe- the other females. On her first visit she stayed malesentered the nestrepeatedly before laying for 27 min and laid the fourth egg. She then the third egg but after that did not return to flew off and returned 80 min later and stayed the nest until 1800 in all but two nests. Females on the nest for 5.5 h. Neither the male nor the entered three of four nests between 1800 and female returned after day 4. 1930. At Nest 15 the female entered the nest at Total time spent on nest in laying period.- 1851,left at 1903,and did not return that night, During the whole laying period of daylight but her mate spentthe night on the nest. Dur- hours, males stayed on the nest four times as ing the day malesvisited the nestabout as often much as females(39.8% vs. 9.9%). As egg lay- as the previousday, but the averagevisit was ing advanced,males increasedand femalesde- much longer (Table 1). creasedthe amount of time they spent on the On day 4, all but one of the females entered nest (Fig. 2). On days 1 and 2, females spent the nest only onceand laid the last egg at that slightly more time on the nest than males did time. After leaving the nest, they were never (16.4% vs. 13.2% and 16.7% vs. 16.2%, re- seen near it again. At Nest 9 the female left the spectively).On day 3 somemales spent a long- nest i min after entering it, but she returned er time on the nest than previously, and it was 30 s later and laid the fourth egg.Males entered on this day that the greatestrange was seen in the nest less often than previously but stayed the percentagesof time spent on the nest for muchlonger during each visit. The malesstayed different males. The males on five of the nine on the nestat night, but femalesdid not return nests spent more than 50% of the day on the to the nest. nest (46.2% on average,range 16.3%-71.3%). Two males in the Toyo area abandonedtheir The femaleson day 3 reducedtheir time on the nests after their femaleslaid the third egg, but nest to only 7.3%. On day 4 the males spent the femalesstill laid the fourth eggin the nests. 78.7% of the day on the nest, and femalesspent Pair 12 visited their nest in typical fashion only 3.0%. through day 3. On day 4, the female exhibited Femalesdid not visit their nestsat all at night typical behavior. She enteredthe nest, laid the during the laying period, and males did not fourth egg, left the nest, and never returned. visit at night until after the third egg was laid. The male on day 4, however, never visited the A male spent68.6% of the night followingthe nest, nor was he seenaround the nest. I began laying of the third egg and 86.7% of the night observingNest 13 on day 2. The male did not following the laying of the fourth egg on the enter the nest between 0744 and 1157 on day nest. 2, nor at any time on day 3, and he did not Male nest attendance in the Painted Snipe even visit the vicinity of the nest on day 4. His reached mid-incubation levels on day 3, and female stayed on the nest more than 40% of the incubation started on day 4. In the Red- the total observation time on both days. On necked Phalarope (Hild6n and Vuolanto 1972), 52 SHIGEMOTOKOMEDA [Auk, Vol. 100

DAY I DAY 2

40-

II I- • 60-

DAY 4 40-

20-

o 0-5I 5-10 DAY10-1515-20 20< 0-5 5-10 10-15 151-20 DISTANCE FROM NEST (m)

Fig. 3. Average distanceof female from nest. Shaded bar shows time female spent with male and white bar shows time female spent alone. male visits to the nest become gradually more Because the male and female arrived at the frequent and longer as the laying period pro- nest togetherearly in the morningbefore the gresses.The SpottedSandpiper male incubates laying of the third egg, it is possiblethat they sporadicallybefore day 3 (Maxsonand Oring had consortedduring the precedingnight. The 1980).In the Red Phalarope(Schamel and Tracy female probably left her mate on the nest at 1977, Mayfield 1979), the Wilson'sPhalarope night after the third egg was laid, however, (Phalaropustricolor; Howe 1975), and the Wat- becausethe male began spending considerable tled Jacana(Jacana jacana; Osbone and Bourne time on the nest and the femaledid not stay in 1977), males appear to begin incubating on the vicinity of the nest. day 3. As egg laying advanced,the amount of day- Consortingratio.--A male usuallyfollowed a light time femalesstayed within 5 m of the nest female for three or four days during the pre- decreasedin the Toyo area (Fig. 3). On days 1 layingperiod. The consorting ratio was 99% in and 2, femalesstayed within 5 m of the nest the Toyo areaduring the prelayingperiod. Al- most of the time and consorted with the males though the consortingratio for the whole lay- almostall day. Within 10 m of the nest, females ing periodwas 68%, it decreasedfrom 99% on were occasionallyseparated from the males, day 1 and 93% on day 2 to only8% on the last who stayedon the nest, but when femaleswent day (Table 2). The consortingratios varied more than 15 m from the nest, the malesalways widely among individual pairs on day 3. In followed. On day 3 females spent less time two of the threepairs the femaleconsorted with within 5 m of the nest and often went away the male on day 4 only when she came to lay. from the nest alone. After laying the fourth egg, One female (Nest 9), however, returned to the femalesleft the nest and movedsolitarily, while vicinity of the nest after laying the fourth egg the males stayedon the nest. and stayedwithin 5 m of the incubatingmale. Behaviorduring the incubationperiod.--Dur- January1983] NestAttendance in Painted Snipes 53

TABLE3. Nestattendance of maleduring the incubationperiod. (Females did notvisit the nestduring the incubation period.)

Daytime Nighttime Averagepercentage time on nest(range) 80.4(64.9-94.7) 89.6(88.2-91.1) Averageduration of nestvisits in minutes(range) 39.8(23.9-110.4) 92.9(77.7-111.2) Percentageof visitsexceeding 30 min 53.4 100.0 Numberof days 14 2 Number of hours observed 156.0 22.6

ing the incubationperiod males typically left other hand, only the males incubate in the the nest 1.3 times per hour for about 10 min. Northern Jacana(Jenni and Betts 1978), the Red- Males sometimes remained on the nest for sev- neckedPhalarope (Hild6n and Vuolanto 1972), eral hours, especiallyduring the morning. At and the Red Phalarope(Mayfield 1979), all of othertimes, they stayed away froin the nestfor which have proved to be polyandrous.In the aslong as 40 min. Maleswere on thenest 80% Wilson's Phalarope (Howe 1975), which is of the time during the day and 89% during the probably polyandrous,only males incubate. night (Table3). At night the malesleft the nest The Painted Snipe female decreasedthe time lessfrequently and spentlonger periods on the she spenton the nest as egg laying advanced, nest. and almost all her nest visits occurred before When malesleft the nest,they usuallywalked the third egg was laid and were less than 10 or flew away immediately.In 61 of 85 obser- min long, exceptfor the four egg-layingvisits. vations made during the incubation period, After a male started to incubate, the female did males left the vicinity of the nest within 1 min not visit the nest againexcept to lay the fourth of leaving the nest. This behavior differed egg, nor was she ever seenin the vicinity of markedly froin behavior during the laying pe- the nest. Therefore, incubation is performed riod, becausethe male spent much time in the only by a male.A femaleneither helps her mate vicinity of the nest during the laying period. nor incubatesalone. This suggeststhat the de- A male typicallyflew back to a placeabout 20 gree of sex-rolereversal in incubationis more in froin the nest and then walked to the nest, complete in the Painted Snipe than in the where he preened for a time before entering Mountain Plover, the Sanderling, or the Spot- the nest. ted Sandpiper and is similar to that in the The incubation period averaged 16.8 days Northern Jacanaand the three speciesof phal- (range 16-18 days, n = 6). During the incu- aropes. One Painted Snipe female began in- bationperiod, the femalewas neverseen at the cubatingwhen her mate abandonedtheir nest nest and apparentlynever cameanywhere near but ceasedafter about a half-day.This fact may the nest (confirmedby banded femalesfor 7 suggestthat ancestralPainted Snipe females days). When the male interruptedincubation incubated. to forage,he did not consortwith any females In the speciesin which femalesdo not in- (confirmedfor 6 days).The maleand his chicks cubate,females may deserttheir mates during left the nest within a half-day after all the eggs incubation, as in the Red Phalarope (Schamel hatched,and they never returned to their nest and Tracy 1977, Ridley 1980); they may help (n = 4). their mates, as in the Northern Jacana, the fe- males of which visit the nest, shade the eggs DISCUSSION (Jenniand Betts 1978), and help defend the ter- In speciessuch as the MountainPlover (Cha- ritory againstother birds, even during the in- radrius montanus;Graul 1973) and the Sander- cubationand rearingperiods (Jenni and Collier ling (Calidrisalba; Panneleeand Payne 1973), 1972,Jenni 1974); or they may deserttheir mates which have double clutches, females incubate as soon as or before the clutch is completed, as the last clutch with or without males. In the in the Red-necked Phalarope (Hild6n and Vuo- polyandrousSpotted Sandpiper, females help lanto 1972) and the Wilson's Phalarope (Howe with incubation unlessthey obtain a new mate 1975). In the Painted Snipe, males did not fol- (Hays 1972,Maxson and Oring 1980).On the low their mates on or after day 4, and females 54 SHIGEMOTOKOMEDA [Auk, Vol. 100

did not visit their nestsafter the fourth eggwas She copulatedonly with the later mate during laid. The pair bond of the Painted Snipe ap- the layingperiod of the earliermate (Oring and peared to break down after the laying of the Maxon 1978). In these cases, the earlier male last egg in the same way as the pair bonds of does not always rear his own offspring, al- Red-necked and Wilson's phalaropesdo. It is though Northern Jacana females spend a advantageousto a sequentiallypolyandrous fe- greateramount of time in the earlier male's ter- male that she desert the first mate as soon as ritory just beforeand during the laying period possibleand gather energy reservessufficient and simultaneousmating is exceptionalin the to obtain the next mate and to lay his clutch. Spotted Sandpiper. Other mechanismsmay In the simultaneouslypolyandrous American explain fertilization in specieswith a simulta- Jacana,a female stays with her mate and pre- neously polyandrousmating system.Physio- pares to lay the replacement clutch for him logical and genetic researchon these species (Jenni and Collier 1972). will be necessaryto identify these mecha- Painted Snipe females are unlikely to mate nisms. with anothermale during the prelayingperiod or on the first and seconddays of laying, be- ACKNOWLEDGMENTS causethey consortwith their currentmates most I thank Prof. Donald A. Jenni of the University of of the time on thesedays. Femalesalso consort Montana and Ms. Helen Hays of the American Mu- with their matesduring the night after the sec- seumof Natural History for their criticalreading and ond egg has been laid. During day 3, females revision of the manuscript.I wish to thank the mem- begin spending less time with their mates, but bersof the Laboratoryof Sociology,Faculty by this time it is likely that their fourth egghas of Science,Osaka City University,especially Dr. Sa- alreadybeen fertilized. It is extremelyunlikely toshi Yamagishi, for continuing guidance and en- that even the last egg could be fertilized by couragement.I alsowish to thank Dr. NagahisaKu- roda of YamashinaInstitute for Ornithologyfor his some other male. After laying the last egg and instructive comments. abandoningtheir mates, the femalescould be- gin to consortwith other males. The abandon- LITERATURE CITED ment of the mate and emancipation from in- B,•I(ER, E. C.S. 1935. Nidification of birds of the cubation make it possible for females to mate Indian Empire. London, Taylor and Francis. with other males, thus enabling this speciesto BEVEN, J. O. 1913. Notes and observations on the have a sequentiallypolyandrous mating sys- PaintedSnipe (Rostratulacapensis) in Ceylon.Ibis tem. 54: 527-534. A simultaneously polyandrous female cop- GR•,UL,W.D. 1973. Adaptive aspectsof the Moun- ulates with several males in a short time. In tain Ploversocial system. Living Bird 12: 69-94. contrast, a sequentially polyandrous female Hays, H. 1972. Polyandryin the SpottedSandpip- mates with only one male during a set period. er. Living Bird 11: 43-57. The probability of a male rearing his own off- HILD•N, O., & S. VUOLANTO. 1972. Breeding bi- spring is different in these two polyandrous ology of the Red-neckedPhalarope Phalaropus lobatus in Finland. Ornis Fennica 49: 57-85. sytems. In the sequentially polyandrous HOWE,M. A. 1975. Behavioralaspects of the pair species,a male has only to keep his mate for a bond in Wilson'sPhalarope. Wilson Bull. 87: 248- few days before and during the laying period 270. to ensure that he will rear his own offspring JENm,D.A. 1974. Evolutionof polyandryin birds. (Schamel and Tracy 1977). A simultaneously Amer. Zool. 14: 129-144. polyandrousfemale lays the first set of eggson --, & B. J. BETTS. 1978. Sex differences in nest one nest for the first mate and the second set construction, incubation, and parental behav- of eggs on another nest for the second mate, iour in the polyandrousAmerican Jacana (lacana successively(Jenni and Collier 1972, Maxson spinosa).Anim. Behav. 26: 207-218. and Oring 1980),but she copulateswith several ß & G. COLLIER. 1972. Polyandry in the American Jacana(lacana spinosa).Auk 89: 743- males during that period. Northern Jacanafe- 765. males copulatedwith three or four different L,•ci•, D. 1968. Ecologicaladaptations for breeding males in a single afternoon (Jenni and Collier in birds. London, Methuen & Co. 1972). One Spotted Sandpiper female copulat- MAXSON,S. J., & L. W. ORING. 1980. Breedingsea- ed with two different males for two days, and sontime and energybudgets of the polyandrous she laid for the earlier mate on the next day. Spotted Sandpiper. Behaviour 74: 200-263. January1983] NestAttendance in PaintedSnipes 55

MAYFIELD,H. F, 1979. Red Phalaropesbreeding PARMELEE, D. F., & R. B. PAYNE. 1973. On mul- on BathurstIsland. Living Bird 17: 7-39. tiple broodsand the breedingstrategy of arctic ORING,L. W., & M. L. KNUDSON.1972. Monogamy Sanderlings.Ibis 115: 218-226. and polyandryin the SpottedSandpiper. Living RIDLEY,M.W. 1980. The breedingbehaviour and Bird 11: 59-73. feeding ecologyof Grey PhalaropesPhalaropus --, & S. J. MAXSON. 1978. Instances of simul- fulicariusin Svalbard.Ibis 122:210-226. taneouspolyandry by a SpottedSandpiper Ac- SCHAMEL,D., & D. TRACY. 1977. Polyandry, re- tiris rnacularia. Ibis 120: 349-353. placementclutches, and site tenacity in the Red OSBORNE,D. R., & G. R. BOURNE. 1977. Breeding Phalarope (Phalaropusfulicarius) at Barrow, behavior and food habits of the Wattled Jacana. Alaska. Bird-Banding 48: 314-324. Condor 79: 98-105.

IMPORTANT NOTICE TO AUTHORS

Effectivewith this issue (January1983), The Auk will adhere to use of the commonand scientificnames of North American birds as listed in the Thirty-fourth'Supplement to the American Ornithologists' Union Check-List of North American Birds (1982 Auk 99(3): 1CC-16CC). Please use this nomenclaturewhen pre- paring manuscriptsfor submissionto TheAuk. Exceptionsshould be justified in writing to the Editor. ENDOCRINE RESPONSES TO INCLEMENT WEATHER IN NATURALLY BREEDING POPULATIONS OF WHITE-CROWNED SPARROWS (ZONOTRICHIA LEUCOPHRYS PUGETENSIS)

JOHN C. WINGFIELD,1 MICHAEL C. MOORE, AND DONALD S. FARNER Departmentof Zoology,University of Washington,Seattle, Washington 98195 USA

ABSTRACT.--Plasmalevels of luteinizing hormone (LH), testosterone,and corticosterone were measuredin relation to periods of inclement versusfair weather during the reproduc- tive seasonof the PugetSound White-crowned Sparrow (Zonotrichia leucophrys pugetensis). In 1974, cool stormy weather in spring delayed the onset of breeding by one month and also prolongedthe period of elevatedcirculating levels of LH and testosterone,compared with the fair spring of 1975. Inclement weather in 1974 did not appear to be stressful,as indicatedby body weights and plasmalevels of corticosterone.In late May 1980,however, a storm occurredafter nesting activities had begun and all pairs sampled were feeding young. In this case, plasmalevels of corticosteronewere greatly elevatedabove those of birds sampled at the same time in the warm spring of 1979 and also above those of birds sampledin spring of both 1974 and 1975. In addition, fat depotswere virtually exhausted in birds sampled during the storm of 1980, suggestingthat these birds were stressed.Most pairs lost their brood in May 1980, presumablyto starvation,and renestedafter amelioration of environmental conditions in June. Thesedata suggestthat althoughstorms may modify the onsetand temporalprogression of the reproductivecycle, they are stressfulto adults only when the nesting phase is in progress.Thus, the underlying mechanismsby which inclementweather delaysthe onset of breeding or disrupts the nesting once underway are likely to have different endocrine bases.Received 29 March 1982, accepted31 August1982.

ASEASONALstorms occasionallydisrupt the tempting to suggest that aseasonal storms nesting activity of birds that breed at higher somehowdepress circulating levels of lutein- latitudes and altitudes. For example, periods izing hormone(LH) and sexsteroid hormones, of unseasonabtycold weather, prolongedpre- which regulate sexualand territorial behavior cipitation, or stormsmay reduce the trophic (e.g. Adkins-Regan1981, Harding 1981;see also resources available to Chaffinches (Fringilla Wingfield and Farner1980), and elevateplasma coelebs)to an extent that adults cannot find suf- levels of corticosterone, which mobilizes en- ficient food for themselvesand their nestlings. ergy reserves (see Holmes and Phillips 1976, The latter may die of starvation, or, in some Siegel 1980 for reviews). cases,the nest may be damaged or destroyed Although renestingafter the lossof a clutch (Newton 1964, 1973). In Mountain White- or brood involves well-characterized endocrine crownedSparrows (Zonotrichia leucophrys or- changes(Donham et at. 1976, Wingfield and iantha) aseasonalsnowfall may cover food re- Farner 1979), the effects of inclement weather sourceswithout destroyingnests. Under these and the mechanismsthrough which it disrupts conditions,breeding pairs oftenabandon nests, the nesting phase are by no means clear. and sometimesthe territory, to foragewidely. Storms, concomitant decreases in ambient As conditions ameliorate and food is again temperature, and restrictedavailability of food available, they return to their territories and may be stressful,especially for individuals that begin a secondnesting effort (Morton et at., are incubating or feeding young. Indeed, sev- 1972, Morton 1976). In these instances, it is eral laboratoryinvestigations of reproductively mature birds have shown that a variety of i Presentaddress: The RockefellerUniversity Field stressesalter temporal patterns of secretionof ResearchCenter, Tyrrel Road, Millbrook, New York LH, depress plasma levels of sex steroid hor- 12545 USA. mones, and elevate circulating titers of corti- 56 The Auk 100: 56-62. January 1983 January1983] HormonalResponses to Weather 57

25. Coupville 1974

20- øC 1975

15'

10-

Apr Moy Jun Fig. 1. Temperaturedata (3-daymeans) for Coupeville,Whidbey Island, Island County, Washington (from U.S. EnvironmentalData Service).Coupeville is approximately10 km west of the study siteson Camano Island.

costerone(e.g. Wilson et al. 1979, Scaneset al. MATERIALS AND METHODS 1980, Wingfield 1980, Wingfield et al. 1982; see Holmes and Phillips 1976,Siegel 1980 for Study areas and field techniques.--Male White- review). As far as we are aware, however, there crowned Sparrows were captured in mist nets and have been no investigationsof the endocrine baited live traps in the vicinity of Seattle,Washing- responsesof naturallybreeding populations of ton (48øN),on study sites on CamanoIsland (Island birds to the stressful effects of aseasonal storms. County) in 1974and 1975and near Hart's Lake (Pierce In this communicationwe report the effects County) in 1979 and 1980. Results of investigations at the former site have been publishedpreviously of low ambienttemperature and stormyweath- (Wingfield and Farner 1977, 1978a). er on plasma levels of hormones and repro- Immediately after a bird was captured, a blood ductionin free-livingpopulations of the Puget sample(approximately 400/•1) was removedfrom the Sound White-crownedSparrow (Z. l. pugeten- alar vein into heparinized capillary tubes. Each bird sis). was then banded, color marked, and weighed, and

McMillin Res.

.o. /,,",,\..",, ' ...../,' v'/,,."", ,'"., •,•.,•,'•'• '.,,',,,• '..,,' ',,

Apr May Jun FiG.2. Temperaturedata (3-day means) •ot McMJ]]jnRese•oJt, •ietce County,Washington (•om U.S. •nvJtonmema]Data Se•Jce). McMJiiJn Rese•oJt is approximately25 km northeasto• the studysite at Ha•t's Lake. 58 WXtqCFtELD,MOORE, AND FARbIER [Auk, Vol. 100

TABLE1. Breedingstatus in relationto weatherof naturallybreeding populations of White-crownedSpar- rows (Zonotrichialeucophrys pugetensis).

Year Sampledates Weather Breedingactivity 1974 30 April-20 May Cold, heavy overcast,precipitation, Males on territory and paired, little frequent storms further breeding activity 1975 30 April-20 May Cool, sunny, no storms Most femalesincubating 1974 20 May• June Cold, heavy overcastand precipita- Breedingdelayed by storms,very tion, frequent storms few females with nests, little breeding activity 1975 20 May-6 June Cool, sunny, no storms All sampledpairs feeding nestlings First fledglingsappear 1974 6 June-2 July Warm and sunny Feedingnestlings, first fledglings appear 1975 6 June-2 July Warm and sunny Nesting activity for secondbrood, incubation and feeding nestlings 1979 26 May-31 May Warm and sunny All sampledpairs feeding nestlings of fledglings 1980 27 May-30 May Cold, heavy overcastand precipita- All sampledpairs feeding nestlings tion or fledglings 1980 8 July-28 July Warm and sunny All samplespairs feeding nestlings or fledglings its gonadal state was assessedduring unilateral lap- Passeriformesby Follett et al. (1975). Plasma levels arotomy (Wingfield and Famer 1976). In 1979 and of testosteronewere measuredby the methodsof 1980, estimatesof subcutaneousfat depot in the fur- Wingfieldand Famer(1975), and corticosteroneby a culum and abdomen were also made. All birds were radioimmunoassayprocedure following the chro- then releasedfor field observationsand possiblere- matographic separation procedure described by capture.After centrifugationof blood samplesin the Wingfieldand Famer(1975). These assays have been field, the plasmasamples were frozen on dry iceuntil usedroutinely in this laboratoryfor over8 years(see returned to the laboratorywhere they were storedat Wing field and Famer 1977, 1978a, b, 1980). -20øC until analyzed(Wingfield and Famer 1976). Statistics.--Differencesbetween groups were de- Collectionof samplesin relationto weather.--In 1974 termined by the Mann-Whitney U-test, one-tailed and 1975,blood sampleswere collectedat least4 days (Zar 1974). per week from mid-April to August as part of an investigationof the endocrinecontrol of reproduc- RESULTS tion under natural conditions (Wingfield and Famer 1977, 1978a).The spring of 1974(late April and May) AS noted above,the ambient temperaturewas was marked by continuallyindement weather; heavy 5-10øC lower in early May 1974 than in 1975 precipitation and strong winds (up to 60 km/h) pre- vailed in May, in contrastwith the same period in (Fig. 1). EarlyMay is usuallythe periodof egg 1975 (Fig. 1). laying for the first clutch (Fig. 1, Table 1; see In 1979 and 1980, blood sampleswere collectedat also, Wingfield and Farner 1977, 1978a). Tem- least 3 days per week from mid-April through early peraturesin 1974remained generally lower than July as part of further investigationson the hormonal thoseof 1975,except for a brief period in mid- control of reproduction in birds (Moore and Wing- May, and did not increase appreciably until field 1980, Moore 1982). The springsof 1979 and June. This period of low temperaturewas ac- 1980(April-May) were mild, with no storms,and in companied by heavy overcastand frequent both years breeding began in early May. In 1980, storms with precipitation (Table 1). In 1975 however, an intense storm accompaniedby heavy precipitationoccurred at the end of May, when most weather was generally sunny and without pairs were feeding nestlingsor fledglings.Temper- storms;all pairs studiedwere feedingnestlings ature data for McMillin Reservoir, 25 km northeast or fledglingsby early June. In 1974breeding of Hart's Lake, are presentedin Fig. 2. was delayed,with few femaleshaving nests by Hormoneassays.--Luteinizing hormone was mea- this time. Nestlings were not found until late sured by the double antibody, post-precipitationra- June, and the first fledglingsdid not appear dioimmunoassayfor avian LH as developedby Fol- until 2 July 1974, in comparisonwith 4 June in lett et al. (1972)and modified for useon plasmafrom 1975. January1983] HormonalResponses to Weather 59

0.8T 0.4 -0 120 1

14 0 ß • 4o80B Fott79 80 80 79 80 8O 26- 31 8-26 26 - 31 8 26 Moy July Moy duly

FIG. 4. Fat scoreand plasma levels of LH, testos- terone (T), and corticosterone (B) in male White- 74 75 74 75 74 75 crowned Sparrows (Z. l. pugetensis)sampled during 30Apr- 20Moy- 6Jun- 20Moy 6Jun 2Jul springand summer,1979 and 1980.The stylizedrain cloud indicates the period of stormy weather, and FIc. 3. Body weight, and plasma levels of LH, the sun indicatesthe periodsof fair weather.Vertical testosterone(T), and corticosterone(B) in male White- lines are standard errors of means, and numbers crownedSparrows (Z. 1. pugetensis)sampled during within open bars indicate samplesizes. Asterisks in- spring 1974 and 1975. The stylized rain cloud indi- dicate significantdifference between 1979and 1980, catesthe period of stormyweather, and the sun in- ** = P < 0.001. dicates periods of fair weather. Vertical lines are standarderrors of means,and numberswithin open barsindicate sample sizes. Asterisks indicate signif- At this time in 1980, however, there was a pre- icant differencebetween 1974 and 1975, * = P • 0.05; cipitous decreasein temperature, to as low as ß*= P • 0.02. 10øC below that recorded at the same time in 1979 (Fig. 2), accompaniedby heavy overcast and precipitation (Table 1). In June, conditions There was a downward trend in both years in plasmalevels of LH and testosteronein males ameliorated, temperatures remaining below those of 1979 but similar to those of June 1974 from late April to June or July (Fig. 3), as is and 1975(Fig. 1). During late May 1980,plasma typical for White-crowned Sparrows through levels of corticosteronewere elevated greatly this period (Wingfield and Farner 1977, 1978a). Plasma levels of LH in males in June 1974 were above those of birds sampled at the same time higher than in males during the same period in 1979and at the same stageof breeding (P < 0.001, Fig. 4). By July 1980, when birds were in 1975(P < 0.02), and circulatinglevels of tes- feeding nestlingsof a secondnesting attempt, tosteronewere alsohigher during May and June corticosterone levels had declined (P < 0.02). in 1974 (P < 0.02 and P < 0.05). Plasma levels Subcutaneousfat depotswere virtually deplet- of corticosteronehad increasedby late May and ed in birds sampled during the storm of 1980 June 1975when birds were feedingyoung (P < 0.05). Therewere no differencesin body weight (P < 0.001)but were restoredby July. No dif- ferencesin plasma levels of LH and testoster- between the two years. one in males were observedbetween years. The weather in early spring in both 1979and 1980was warm and sunny, with temperatures DISCUSSION usually remaining above 15øC(Fig. 2). In both years White-crowned Sparrows were feeding Abnormal weather conditionsmay delay the nestlings or fledglings by late May (Table 1). onset of breeding, despite the presence of 60 w•o, Mooa•, ^ND F^aN•a [Auk, Vol. 100 functional or near functional gonads, or dis- temperatures occur late in spring during pa- rupt the nestingphase if breedingis already rental care, however, trophic resourcesdecline underway (for reviews seeMarshall 1949, 1959; at a time when both adultsand a broodof young Lofts and Murton 1968; Famer and Lewis 1971; must be fed. In this case, the storm is stressful, Newton 1973; Murton and Westwood 1977; as indicated by increasesin plasma levels of Wingfield 1980; Wingfield and Farner 1980). corticosteroneand depletionof fat stores.Thus, Our resultsshow that the delay in the onsetof the underlying mechanisms by which incle- breeding causedby prolonged spring storms ment weather delaysthe onset of breeding or in 1974was alsoaccompanied by maintenance disruptsthe nesting phase onceunderway are of elevatedplasma levels of LH and testoster- likely to have different endocrinebases. one, which normally decline from the initial Whether storms influence endocrine func- springpeak during the onsetof incubationand tion througha directeffect of, for example,low feeding of young (Wingfield and Farner 1977, temperatureor an indirect effect on food re- 1978a,b). Thus, the extendedperiod of elevat- sourceshas been a matter of conjecture(Wing- ed plasmalevels of sex hormonesin 1974may field 1980, Wingfield et al. 1982). In do- have maintained sexual and territorial behav- mestic birds low temperaturestend to depress ior so that nesting activities could begin im- testis growth (e.g. Huston 1975) and disrupt mediatelyafter ameliorationof environmental ovulatory cycles (e.g. Riddle and Honeywell conditions. The stormsdid not appear to stress 1924), even though food and water are provid- the adult White-crowned Sparrows because ed ad libitum. Similar treatment also produces plasmalevels of corticosteronewere lower than an elevation of circulating levels of corticoste- in birdssampled at the sametime in 1975,when rone (Nir et al. 1975, Jeronen et al. 1976, Beuv- the birds were foraging intensivelyin order to ing and Vonder 1978). Low environmental feed young. Foragingto feed youngitself may temperature,however, has only a slight effect have been stressful,as suggestedby the grad- on photoperiodicallyinduced gonadalgrowth ual increasein circulating levels of corticoste- in feral species (for review of early literature rone in male Z. I. pugetensisduring parental care see Marshall 1949, see also Lewis and Famer (Wingfield and Farner 1977, 1978a). In 1980, 1973), on plasma levels of LH or testosterone, however, the storm struck after the nesting or on circulatinglevels of corticosteronein male phase had begun and was accompaniedby a Z. I. gambelii(Wingfield 1980,Wingfield et al. dramatic increasein plasma levels of cortico- in press).Moreover, Rowan (1925)showed that sterone well above those measured in 1974 and male Dark-eyedJuncos (Junco hyemalis) housed 1975, and a decrease in subcutaneous fat de- in outdoor aviaries (in Alberta, Canada) under pots,suggesting that thesebirds were stressed. artificiallong days and with free accessto food Curiously, plasma levels of LH and testoster- developedfunctional gonads and cameinto full one were unaffected,probably becausethese song in December, when ambient tempera- levels had already declined from the initial tures were as low as -47øC. Similarly, feral spring peak (see Wingfield and Farner 1977, populations of crossbills(Loxia curvirostra and 1978a). Furthermore, these birds remained on L. leucoptera)can breed at any time of the year territory (or at least in the immediate area) if conecrops of pines (Pinussp.), larches(Larix throughoutthe period of inclementweather, sp.), and spruce(Picea sp.) are sufficient(Bai- although nesting was generallyunsuccessful, ley et al. 1953,Tordoff and Dawson1965, New- as no fledglingswere observeduntil after a sec- ton 1973). ond nesting attempt in Juneand July. Thus, if birds have free accessto food, low The different endocrine responsesto storms environmentaltemperature appears to have lit- in 1974 and 1980suggest the following expla- tle effect on reproductive function in feral nation: If an unseasonablestorm occursearly species.Inanition can have rapid and dramatic in spring,the onsetof breedingis delayedeven effects on reproductive function, however, in though the gonadsmay be functionalor near both domesticatedand feral avian species.For functional, and plasma levels of LH and tes- example, varying degreesof food restriction tosteronein malesremain generallyhigh. The result in gonadal atrophy, decreasedplasma effectsof the storm do not appear to be stress- levels of LH and testosterone, and increased ful, because there are no young that require concentrationsof corticosterone(e.g. Assen- extraforaging by the adults.If stormsand low macher et al. 1965, Assenmacher 1973, Erb et January1983] HormonalResponses to Weather 61

al. 1978, Wilson et al. 1979, Harvey et al. 1980, testosteronein blood plasma of laying hens. Scaneset al. 1980,Wingfield 1980;R. A. Lewis, Poultry Sci. 57: 1042-1051. P. W. Mattocks,J. C. Wingfield, and D. S. Far- EARNER,D. S., & g. A. LEWIS. 1971. Photoperi- ner unpubl. results). Although low tempera- odismand reproductivecycles in birds. Pp. 325- 370 in Photophysiology,vol. 6 (A. C. Giese, Ed.). tures may result in more rapid depletion of New York, Academic Press. energy stores, especially for nocturnal FOLLETT,B. K., C. G. SCANES,& F. J. CUNNINGHAM. thermoregulationand for the feeding or brood- 1972. A radioimmunoassayfor avian luteiniz- ing of young, these stores can be easily re- ing hormone. J. Endocrinol. 52: 359-378. placed if sufficient food is available (Ketterson , D. S. EARNER,& P. W. MATTOCKS,JR. 1975. and King 1977). Thus, it seemslikely that in- Luteinizing hormone in the plasma of White- clement weather is stressfulduring breeding crowned Sparrows, Zonotrichialeucophrys gam- becauseit depressesfood resources,especially belii, during artificial photostimulation. Gen. in speciessuch as the White-crowned Sparrow, Comp. Endocrinol. 26: 126-134. which feeds its young almost entirely on in- HARDING, C. H. 1981. Social modulation of circu- sects. lating hormone levels in the male. Amer. Zool. 21: 223-232. ACKNOWLEDGMENTS HARVEY, W. N., B. J. MERRY, & J. G. PHILLIPS. 1980. Influence of stress on the secretion of corticos- The investigations described herein were sup- terone in the (Anasplatyrhynchos). J. En- ported by grants BMS 74-13933,PCM 77-17690, and docrinol. 87: 161-171. PCM 80-18019 from the National Science Founda- HOLMES, W. N., & J. G. PHILLIPS. 1976. The ad- tion; by Graduate SchoolResearch Fund, University renal cortex of birds. Pp. 293-420 in General, of Washington; by the University of Washington comparativeand clinical endocrinologyof the Training Grant in Neurobiology from the National adrenal cortex (I. Chester-Jones and I. W. Hen- Institutes of Health, GM 07108; by the Chapman derson, Eds.). New York, Academic Press. Fund, AmericanMuseum of Natural History; and by HUSTON, T. M. 1975. The effects of environmental SigmaXi. We are gratefulto EstherArruza who typed the final draft. temperature on fertility of the domestic . Poultry Sci. 54: 1180-1183.

LITERATURE CITED JERONEN,E., P. ISOMETSA,g. HlSSA, & A. PYORNILA. 1976. Effects of acute temperature stresson the ADKINs-REGAN,E. 1981. Hormone specificity,an- plasma catecholamine,corticosterone and me- drogen metabolism, and socialbehavior. Amer. tabolite levels in the pigeon. Comp. Biochem. Zool. 21: 257-271. Physiol. 55C: 17-22. ASSENMACHER,I. 1973. Reproductive endocrinol- KETTERSON,E. D., & J. R. KING. 1977. Metabolic ogy: the hypothalamo-hypophysialaxis. Pp. 158- and behavioralresponses to fastingin the White- 191 in Breedingbiology of birds (D. S. Farnet, crowned Sparrow (Zonotrichialeucophrys gam- Ed.). Washington, D.C., National Academy of belii). Physiol. Zool. 50: 115-129. Sciences. LEWIS, R. A., & D. S. EARNER. 1973. Temperature --, A. TIXIER-VIDAL, & H. ASTIER. 1965. Effets modulationof photoperiodicallyinduced vernal de la sous-alimentationet du j•une sur la gon- phenomena in White-crowned Sparrows (Zo- adostimulation du Canard. Ann. Endocrinol. notrichialeucophrys). Condor 75: 279-286. (Paris) 26: 1-26. LOFTS,B., & R. K. MURTON. 1968. Photoperiodic BAILEY, A.M., R. J. NIEDRACH, • A. L. BAILEY. and physiologicaladaptations regulating avian 1953. The Red Crossbills of Colorado. Publ. breedingcycles and their ecologicalsignificance. Denver Mus. Nat. Hist. 9: 164. J. Zool. London 155: 327-394. BEUVING, G., & G. M. A. VONDER. 1978. Effect of MARSHALL,A. J. 1949. Weather factorsand sper- stressingfactors on corticosteronelevels in the matogenesisin birds. Proc. Zool. Soc. 119: 711- plasmaof laying hens. Gen. Comp. Endocrinol. 716. 35: 153-159. 1959. Internal and environmental control DONHAM,g. S., C. W. DANE,& D.'S.EARNER. 1976. of breeding. Ibis 101: 456-478. Plasma luteinizing hormone and the develop- MOORE, M. C. 1982. Hormonal response of free- ment of ovarian follicles after loss of clutch in living male White-crowned Sparrows to exper- female Mallards (Anas platyrhynchos). Gen. imental manipulation of female sexualbehavior. Comp. Endocrinol. 29: 152-155. Horm. Behav. 16: 323-329. ERB, R. E., A. N. DE ANDRADE, & J. C. ROGLER. , & J. C. WINGFIELD. 1980. Hormone-be- 1978. Interrelationships between diet and ele- havior interactionsduring reproductionin White- vated temperatures(cyclic and constant)on con- crowned Sparrows (Zonotrichia leucophrys). centrationsof progesterone,estradiol-17B, and Amer. Zool. 20: 788. 62 WINeFIELD,MOORE, AND FARNER [Auk, Vol. 100

MORTON,M. L. 1976. Adaptive strategiesof Z0- Effect of sexual experience, location, malnutri- notrichiabreeding at high latitude or high alti- tion and repeatedsampling on concentrationsof tude. Pp. 322-336 in Proc. 16th International Or- testosteronein blood plasmaof Gallusdomesticus nithological Congress (J. H. Frith and J. H. roosters.Poultry Sci. 58: 178-186. Calaby, Eds.). Canberra, Australian Acad. Sci. WINGFIELD,J. C. 1980. Fine temporal adjustment --, J. L. HORSTMAN, & J. M. OSBORN. 1972. Re- of reproductivefunctions. Pp. 367-389 in Avian productive cycle and nesting successof the endocrinology (A. Epple and M. H. Stetson, mountain White-crowned Sparrow (Zonotrichia Eds.). New York, Academic Press. leucophrysoriantha) in the centralSierra Nevada. --, & D. S. FARNER. 1975. The determination Condor 74: 152-163. of five steroidsin avian plasma by radioimmu- MURTON, R. K., & N.J. WESTWOOD. 1977. Avian noassay and competitive protein binding. Ste- breeding cycles.Oxford, Clarendon Press. roids 26: 311-327. NEWTON,I. 1964. The breeding biology of the --, & ---. 1976. Avian endocrinology--field Chaffinch. Bird Study 11: 47q58. investigationsand methods.Condor 78: 570-573. 1973. Finches. New York, Taplinger. , & 1977. Zur Endokrinologie einer NiR, I., D. YAM, & M. PEREK. 1975. Effects of stress brutenden Population von Zonotrichia leuco- on the corticosteronecontent of the blood plas- phryspugetensis. Vogelwarte 29: 25-32. ma and adrenal gland of intact and bursectom- , & -- 1978a. The endocrinologyof a ized Gallusdomesticus. Poultry Sci. 54: 2101-2110. natural population of the White-crowned Spar- RIDDLE, O., & H. E. HONEYWELL. 1924. Studies row (Zonotrichialeucophrys pugetensis). Physiol. on the physiologyof reproductionin birds. XVIII. Zool. 51: 188-205. Effects of the onset of cold weather on blood , & --. 1978b. The annual cycleof plas- sugar and ovulation rate in pigeons. Amer. J. ma irLH and steroid hormonesin feral popula- Physiol. 67: 337-345. tions of the White-crownedSparrow, Zonotrich- ROWAN,W. 1925. Relation of light to migration and ia leucophrysgambelii. Biol. Reprod. 19: 1046-1056. developmentchanges. Nature 115: 494-495. ß & . 1979. Some endocrine correlates SCANES, C. G., G. F. MERRILL, R. FORD, P. MAUSER, of renesting after loss of clutch or brood in the & C. HOROWITZ. 1980. Effects of stress (hy- White-crowned Sparrow, Zonotrichialeucophrys poglycemia,endotoxin, and ether) on the pe- gambelii.Gen. Comp. Endocrinol.38: 322-331. ripheral circulating concentrationof corticoster- , & . 1980. Control of seasonalrepro- one in the domestic fowl (Gallus domesticus). duction in temperate zone birds. Prog. Reprod. Comp. Biochem.Physiol. 66C: 183-186. Biol. 5: 62-101. SIECEL,H. S. 1980. Physiologicalstress in birds. ß J. P. SMITH, & D. S. FARNER. 1982. En- BioScience 30: 529-534. docrine responsesof White-crowned Sparrows, TORDOFF,H. B., & W. R. DAWSON. 1965. The in- Zonotrichialeucophrys, to environmental stress. fluenceof day length on reproductivetiming in Condor 88: 399-409. the Red Crossbill. Condor 67: 416-422. ZAR, J.H. 1974. Biostatisticalanalysis. Englewood WILSON, E. K., J. C. ROGLER,& R. E. ERB. 1979. Cliffs, New Jersey,Prentice Hall. SEXUAL DIFFERENCES IN THE BEHAVIOR OF ADULT GREAT BLACK-BACKED GULLS (LARUS MARINUS) DURING THE PRE- AND POST-HATCH PERIODS

RONALDG. BVTLER• A•r) SASKIAJAhIE$-BL•TLEI• The Mount DesertIsland Biological Laboratory, Salsbury Cove, Maine 04672 USA

ABSTRACT.--Sexual differences in the behavior of adult Great Black-backed Gulls (Larus marinus)were examinedduring the pre- and post-hatchperiods of the 1980breeding season. Followingclutch completion,females invested more time than males in both territorial attendanceand incubation, but there were no sexualdifferences in egg-shiftingor grass- collection.In contrast,males engaged in moreagonistic behavior and long-called and yelped in all contextsmore than femalesbefore chick hatching. Males also displayedmore mate- oriented mews and chokesthan did females at this time. During the post-hatch period, there were no sexualdifferences in territorial attendance,brooding, or frequencyof chick feedings,but malescontinued to exhibit higher frequenciesof agonisticacts than did fe- males, as well as higher levels of long-callsand yelps. After chick hatching,both sexes demonstratedincreased levels of agonisticacts, long-calls, yelps, mews, and yeows,while malesexhibited decreased frequencies of chokesand head-tosses.The resultsof this study stronglysuggest that parentalinvestment by male L. marinusmay equal that of females. Received10 August1981, accepted 28 May 1982.

GVLLspecies (Laridae) are generallymonog- feed chickssignificantly more often than do fe- amous, and members of both sexes exhibit the males (Pierotti 1981). Burger (1981a) reported samebehavior patterns related to parentalcare that, althoughadult male and femaleHerring during the breeding season(e.g. incubation, Gulls contribute equal parental investment in brooding, chick feeding, territorial attendance the post-hatchperiod, sexualdifferences in in- and defense).Male and female gulls are often vestment patterns developed as chicks grew difficult to distinguishvisually, and many de- older. Even sexualdifferences in spiteful and scriptions of their behavioral ecology have altruistic behavior have been observed in both simply "lumped" sexes in data analysis. Re- Herring and Western gulls (Pierotti 1980). cent studies, however, have shown that defi- Clearly, the investmentpatterns of adult larids nite differences are evident in the behavior of vary qualitatively, if not quantitatively, and male and female larids. For example, Tinber- additional data are needed to evaluate the rel- gen (1956, 1960) reported that male Herring ative significanceof male and female contri- Gulls (Larus argentatus)were more active in butions in terms of parental effort. collectingnest material and were involved in We examined the behavior of adult Great territorial defense more often than females. Black-backedGulls (L. marinus)during the in- Similarly, male and female Laughing Gulls (L. cubation and post-hatchphases of the repro- atricilla;Burger and Beer 1975)and Ring-billed ductive cycle. The specificobjectives of this Gulls (L. delawarensis;Southern 1981)have been study were to: (1) investigatepossible sexual observedto exhibit differencesin agonisticbe- differencesin a number of agonisticand non- havior toward territorial intruders. Male West- agonistic behaviors exhibited by breeding ern Gulls (L. occidentalis)are not only respon- adults and (2) determine how the transition sible for the bulk of territorial defense, but also from the pre- to post-hatchperiods influenced the behavior of males and females.

METHODS

• Present address:Department of BiologicalSci- This study was conductedon Little Duck Island, ences, Duquesne University, Pittsburgh, Pennsyl- Hancock County, Maine between April and August vania 15282 USA. 1980. Approximately520 pairs of L. marinusnested 63 The Auk 100:63-75. January1983 64 BUTLERAND JANEs-BuTLER [Auk, Vol. 100

TABLE1. Sexualdimorphism in GreatBlack-backed Gulls?

Bill depth Weight Tarsus Culmen at nares Sex (g) (mm) (mm) (mm) Males (n = 19) 1,963+ 11b 97.9 + 0.4b 66.1 + 0.6b 25.4 + 0.2b Females (n = 11) 1,577 + 19 90.7 + 0.7 59.7 + 0.4 22.2 + 0.1

Data expressesas mean _+SE. Males > Femalesin t-test comparison,P < 0.001.

in two large meadowson this 35-ha island. Breeding tact with another conspecific),chase (aerial pursuit habitat consistedof severalspecies of grasses(Grarn- of one adult by another),attack (physical assault upon rnineae),stinging nettle (Urtica dioica),angelica (An- one adult by another, which generallyinvolved bit- gelicalucida), raspberry (Rubus idaeus), and exposed ing), fight (prolongedbout of mutualattacking), long granite outcroppings. The island was uninhabited call (Fig. lb; Moynihan 1958,Tinbergen 1959),yelp and supportedno mammalian predators. (Fig. lc; Stout et al. 1969, Hand 1981),mew (Fig. ld; A blind was erectedon masonaryscaffolding at the Tinbergen 1959,Hand 1981),yeow (plaintive call of periphery of the colony during the pre-incubation Moynihan 1958,plaintive yeow of Hand 1981),gak- stageof the GreatBlack-backed breeding cycle. One ker (Moynihan1958, Eh-Eh call of Hand 1981),choke or both adults on each of 19 nest sites were color (Fig. le; Moynihan 1958,Tinbergen 1959), head-toss markedwith leg-streamersduring the 1978,1979, and (Fig. lf; Hand 1981), shift or roll eggs, grass-collect, 1980breeding seasons (and all adultswere marked feed chick, present grass to chick, and reconsume with temporarydyes), all pairs were sexedon the regurgitatedfood. In addition, the frequency and basis of size differences, and sexes were confirmed duration of time on the territory, incubation bouts, through observedcopulations. In all 19 cases,the and brooding bouts were noted. Data collectionwas largermember of the pair was a male, and morpho- facilitated by our use of binoculars (10x), digital metric data (see Table 1) indicated sexual dimor- stopwatches,cassette tape-recorders, and super-8 mm phism similar to that reported by Ingolfsson(1967). movie cameras. We determined territorial bound- Studynests were selectedto includesites from both aries by meansof aerial photographstaken in May high- and low-densityareas within the colony(But- 1980 and measurements made with an optical tape ler and Trivelpiece 1981). Data were collectedonly measure (see Butler and Janes-Butler 1982). from those territories that were entirely visible for For the purposeof examiningsexual differences in the duration of the breeding season and those in behavior,frequency scores from eachadult pair were which both members of the pair were in attendance analyzedusing the Wilcoxonsigned-ranks test. In- until chicksfledged. Chicks were consideredfledged trasexual comparisionsbetween the pre- and post- at 55 daysfollowing hatching and/or when they were hatch periods were analyzed in the same fashion. observedflying stronglyin the vicinity of the colony. Proportionaldifferences were treated with the Chi- Behavioral observationsof birds on their nesting squarestatistic applied to raw data rather than per- territories were conducted during daylight hours centages.Morphometric data were analyzed using (0500-2100) over the courseof the breeding season. the t-test. To facilitate analysisand discussion,be- Two observers watched an average of 10 territories havioral data were broken down into acts related on eachof 18 days(1,877 nest hours) during the pre- either to attendanceand care of offspring, intrapair hatch period and an average of 7 territories during behavior, or agonisticinteractions. Behaviors such 20 days (1,701nest hours) of the post-hatchperiod. asthe long-call,which occurredin eachof the above We recordedthe frequenciesand identities of adults categories,were analyzedaccording to their contex- engaged in the following behaviors(where neces- tual appearance. sary,definition or sourceof previousdescription for other gull speciesnoted): upright aggressive(Fig. la; RESULTS Moynihan 1958),grass-pull (Tinbergen 1959, 1960), face-off (behavior that occurred in an agonistic con- Attendanceand care of offspring.--Beforetheir text in which one or more adults assumed a recum- chicks hatched, adult females were observed bant position oriented either facing or parallel to a occupying territories (t = 51.1, n = 19, P < neighboringconspecific within i m of a territorial boundary,possibly related to a squat-and-freezede- 0.05) and incubating (t = 32.0, n = 19, P < scribedby Moynihan 1958),jab (gaping-jabof Moy- 0.005; Fig. 2) significantly more than were nihan 1958), charge(lunge or running approachby males. There were no differences, however, an adult with wings generally extended; charging between males and females in the mean (+SE) bird alwaysterminated behavior before making con- frequencies of egg-shifting (• = 0.30 + 0.05/h January1983] SexualDifferences in Larus marinus Behavior 65

b

Fig. 1. Behaviorpatterns in which male and femaleGreat Black-backedGulls differed over the courseof the breeding season:(a) upright aggressive(and gakker) by severaladults near territorial boundary, (b) long-call,(c) yelp, (d) mew, (e) chokeby pair followinggrass presentation by standingbird, (f) head-toss (adult femaleon left) before regurgitationby mate to chick. and 0.34 + 0.04/h, respectively)or bouts of nest- Males and femalesdid not differ significantly substratecollecting (•: 0.18 + 0.05/h and 0.14 in terms of which hours of the day they were + 0.08/h, respectively).The mean nest-relief in attendance during either the pre- or post- interval was 5.3 + 0.6/h, and males fed females hatch periods. The overall adult attendance at the nest site only infrequently following pattern did change, however, following hatch- clutch completion (• = 0.007 + 0.006/h). Fol- ing (Fig. 3). During the pre-hatchperiod, max- lowing chick hatching, males and femalesdid imum numbers of adults were in attendance not differ in terms of time in attendance on the on territoriesduring early morning, late after- territory, and both sexesparticipated equally noon, and evening hours, and territories were (• = 13 + 5 min/h and 18 + 5 min/h, respec- generallyleft unattendedfor no longerthan 1- tively) in broodingchicks during the first week 2 min. During these brief absences,the adult following hatching (Fig. 2). in attendancegenerally flew out over the col- 66 BUTLERAND JANES-BUTLER [Auk, Vol. 100

lOO PRE-HATCH POST-HATCH

li!ilMALES ß T• I FEMALES 50 ,.T...,I*$1I , T •- ß•, P<0.01P

TIME ON INCUBATION TIME ON BROODING TERRITORY TIME TERRITORY TIME

MALE AND FEMALE TIME INVESTMENT Fig. 2. Pre-hatch(1,877 nest-hours)and post-hatch(1,701 nest-hours)comparisons of male and female time investment in territorial attendance,incubation, and brooding. Significant Chi-square values for in- tersexualcomparisons (*) are indicated. ony, defecated,and then returnedto incubate. lumpfish (Cyclopterus lumpus), mackeral In contrast,during the post-hatchperiod, one (Scombarscombrus), sculpin (Myoxocephalus or sometimes both mates left the territoW in sp.), and unidentified fish remains] or pelagic early morning (presumably to forage to feed invertebrates [krill speciesand squid (Loligo their chicks), and maximal numbers of adults borealis)l,organisms relatively independent of were observedon the territories only in late tidal cycles.There were no sexualdifferences afternoon and evening. TemporaW adult de- in the types of food items fed to chicks.Males sertion of territories increasedsteadily from and femalesdid not differ in mean frequencies 0.2% of the total observation hours in the first of chick feedings (• = 0.50 +_0.06/h and 0.45 +_ 10 days of June to 46.3% of the observation 0.06/h, respectively), reconsumptionof food hours in the last 10 days of July. Somechicks offered to chicks (• = 0.12 + 0.04/h and 0.13 + were left unattendedfor periods of up to 8 h, 0.03/h, respectively), or presentation of grass and presumablyboth adults were out foraging or other substrate to chicks (• = 0.13 + 0.03/h at these times. Although this pattern varied and 0.17 + 0.05/h, respectively). between high- and low-density nesting areas Adult vocalizations that were directed at (Butler and Janes-Butler 1982), there were no chicks included long-calls, mews, and yelps. sexual differences in the proportion of time Chicks that mandibulated the bills of recum- spent away from the territoW by pairs in these bant adults and/orcontinuously solicited feed- areas.Attendance patterns during both the pre- ings often elicitedlong-calls from their parents. and post-hatchperiods in 1980 were not af- Adults alsolong-called when chicksleft the ter- fected significantlyby the occurrenceof high ritoW and/or when they returned to the terri- and low tides. This may be related to the fact tow, especiallyif neighboringadults were in that, during 1980breeding season, over 87% the area. Males long-calledmore than females of the food items identified (47% of all chick in this context (t = 32.5, n = 19, P < 0.01). feedings)consisted of fish [alewife(Pomolobus Adults yelped and mewed most often before pseudoharengus),herring (Clupea harengus), feeding chicks. These vocalizationswere also January1983] SexualDifferences in Larusmarinus Behavior 67

AGONISTIC INTERACTIONS 0,6 J• MALES

02 f 'FEMALES

TOTAL ADULTS PRESENT

0.4 e • p

AGON ISTIC INTERACTIONS

0.2

TOTAL ADULTS PRESENT

ONG •EW YELP C ALL 0500 1200 20O0

TIME OF DAY CHICK ORIENTED Fig. 3. Diurnal changesin both territorial atten- dance (expressedas percentagetotal adults in atten- VOCALIZATIONS dance)and the mean frequenciesof agonisticacts for Fig. 4. Comparisonsof meanfrequencies of chick- adult gulls during the pre- and post-hatch periods. oriented vocalizationsexhibited by adult males and Blackenedportions of attendancebars representper- females. Results of Wilcoxon signed-rankstests for centageof territoriesleft unattendedduring the post- intersexual comparisons(*) are indicated. hatch period. used by adults, however, when chicks were mate feeding during the early pre-hatch pe- out of sight in densevegetation or when chicks riods, as well as during attemptsby femalesto wandered too close to a territorial boundary or eat food being presented by males to their a neighboringadult. Malesyelped significantly chicksduring the post-hatchperiod. There were more often than females (t = 29.5, n = 19, P < no sexual differencesin head-tossingduring 0.01), but there were no sexual differences in the pre-hatchperiod (Fig. 5), but malesmewed the frequencyof chick-orientedmews (Fig. 4). (t = 46.0, n = 19, P < 0.025) and choked (t = Pair-bond related behavior.--Adults directed 28.5, n = 19, P < 0.01) more than their mates. long-calls, mews, yelps, chokes, and head- Both sexesexhibited significant decreasesin tossesat their mates during both the pre- and mate-directedmews followinghatching (males: post-hatchperiods (Fig. 5). Long-callswere t = 13.0, n = 19, P < 0.005; females: t = 18.0, generallygiven to or by landing mates, and n = 19, P < 0.005), but only maleschoked (t = males long-called more than females during 53.5, n = 19, P < 0.05) and head-tossed sig- both the pre-hatch(t = 35.0, n = 19, P < 0.01) nificantly less frequently (t = 25, n = 19, P < and post-hatchperiods (t = 41.5, n = 19, P < 0.005) during this period. Finally, males and 0.025). Both males and females exhibited in- femalesyelped before their own or their mate's creasedfrequencies of mate-directedlong-calls aerialdeparture from the territory.Males yelped following chick hatching (t = 30, n = 19, P < in this contextsignificantly more often than fe- 0.005 and t = 0, n = 19, P < 0.005, respective- males, however, both before (t = 19.5, n = 19, ly). Mate-orientedmews and chokes,and to a P <0.005) and after (t = 11.5, n = 19, P < lesser extent head-tosses,appeared primarily 0.005) chick hatching. Intrasexualcomparisons in the context of nest relief in both sexes. Head- of yelps before and after hatchingrevealed no tossingalso appeared in the contextof limited significant changes(Fig. 5). 68 BUTLERAND JANES-BUTLER [Auk, Vol. 100

PRE-HATCH I•i::il,• A L E S ß FEMALES

0.05

POST- HATCH I,--

z

ß FEM^L•S

POST-HATCH T::. ttp

0.05

MATE ORIENTED VOCALIZATIONS

Fig. 5. Pre- and post-hatchcomparisons of mean frequenciesof mate-orientedvocalizations exhibited UPRIGHT FACE'OFF GRASS-PULL by adult males and females. Results of Wilcoxon AGGR. signed-rankstests for intersexualcomparisons (*) and intrasexualcomparisons between the pre- and post- hatch periods (?) are indicated. TYPE OF AGONISTIC ACT Fig. 6. Pre- and post-hatch comparisons of the mean frequenciesof low-level agonisticacts by adult Agonistic behavior.--Agonistic interactions males and females.Results of Wilcoxonsigned-ranks accounted for 34% of all social interactions tests for both intersexualcomparisons (*) and intra- sexual comparisonsbetween the pre- and post-hatch during the pre-hatchperiod and 35% of all en- periods (ñ) are indicated. countersduring the post-hatchperiod. Males and females exhibited marked sexual differ- encesin agonisticbehavior during both pe- males (t = 36.0, n = 19, P < 0.01), although riods (Fig. 6 and 7). During the pre-hatchpe- there were no other significantdifferences in riod, males exhibited significantly greater frequenciesof high-level agonisticacts follow- frequenciesof low-level agonisticacts than did ing chick hatching (Fig. 7). The frequency of females,including upright aggressivedisplays fights that males (2 = 0.005 + 0.004/h)and fe- (t = 13.5, n = 19, P < 0.001), face-off encoun- males(2 = 0.002 + 0.002/h)engaged in did not ters (t = 12.0, n = 19, P < 0.001), and grass- differ significantly. pulling bouts (t = 12.0, n = 19, P < 0.001;Fig. Intrasexualcomparisons of the pre- and post- 6). There were no statistical differences in fre- hatch phases indicated that, following hatch- quenciesof high-levelagonistic acts in the pre- ing, males demonstrated significantly higher hatch phase of the breeding cycle.During the levelsof upright aggressivebehavior (t = 28.5, post-hatchperiod, males again displayed sig- n = 19, P < 0.005), charges t = 35.5, n = 19, nificantly greater frequencies of upright ag- P < 0.01), and chases(t = 0, n = 19, P < 0.001; gressivebehavior (t = 27.5, n = 19, P • 0.01), Figs. 6 and 7). Femalesshowed changesonly face-off interactions (t = 36.5, n = 19, P < in the frequency of low-level agonistic acts 0.01), and grass-pullingbouts (t = 12.0, n = during the post-hatch period. These included 19, P < 0.001;Fig. 6). Malesalso charged con- significantincreases in upright aggressiveacts specificssignificantly more often than did fe- (t = 18.5, n = 19, P < 0.005), face-off behav- January1983] SexualDifferences in Larus marinus Behavior 69

0.10 PRE'HATCH 01 PREHATCH

POST-HATCH

CHARGE CHASE JAB ATTACK LONG YElP CHOKE YEOW GAKKER

TYPE OF AGONISTIC ACT VOCALIZATIONS IN AGONISTIC CONTEXTS Fig. 7. Pre- and post-hatchcomparisons of the Fig. 8. Comparisonsof pre- and post-hatchmean meanfrequencies of high-levelagonistic acts for adult frequenciesof vocalizationsthat were exhibited by malesand females.Results of Wilcoxonsigned-ranks adult males and females in agonistic contexts. Re- testsfor both intersexualcomparisons (*) and intra- suits of Wilcoxon signed-rankstests for both inter- sexualcomparisons between the pre- and post-hatch sexualcomparisons (*) and intrasexualcomparisons periods(?) are indicated. betweenthe pre- and post-hatchperiods ('•) are in- dicated. ior (t = 36.5, n = 19, P < 0.01), and grass- pulling(t = 25.0,n = 19, P < 0.01;Fig. 6). The adult landed on its own territory, when other frequenciesof agonisticbehavior displayed by adults gave long-callsor yelps, when two or both sexes showed an abrupt and sustained moreneighboring territory holders that shared increaseover the courseof the entire post-hatch a commonboundary with a callinggull inter- period.Examination of the daily occurrenceof acted, or at times in responseto a non-neigh- agonisticinteractions from 0600 to 0900, 0900 boring intruder before a more high-level re- to 1200, 1200 to 1500, and 1500 to 1800 revealed sponseby the territory holder. The yeow and no substantialdifferences during the pre-hatch gakker, which appearedto function more as period (Fig. 3). Followinghatching, however, alarmor alertingcalls, were observed most often a distinct decreasein agonisticinteractions for when a high-levelagonistic interaction was in both sexes was observed between 1200 and 1500 progressnear the territorialboundary of the (Fig. 3). This decreasecorresponded to gener- callingbird, when adultsand chicksfrom two ally higher ambient temperaturesand the ten- different territorieswere in closeproximity near dencies for both adults and chicks to sleep or their mutual boundaries, when the calling remain inactive during theseperiods. bird's chickwas in an adjacentterritory, when Vocalizationsthat were frequently exhibited a calling bird was intruding into an adjacent by adult gulls in agonistic contextsincluded territory, or on windy dayswhen large num- the long-call, yelp, choke, yeow, and gakker. bers of adultswere soaringover a callinggull's The long-call, yelp, and choke occurred most territory.During the pre-hatchperiod, males frequentlyduring conspecificoverflights of the displayedsignificantly higher frequenciesof territory, during interactions involving adja- long-callsthan their mates (t = 51.5, n = 19, cent territory holders, when a neighboring P < 0.05) but did not differ with regardto oth- 7O BUTLERAND JANEs-BUTLER [Auk, Vol. 100

NEIGHBOR

lOO PRE- HATCH POST-HATCH

LOW

5o I HiGH

• HIGH

,. p - o,o5 NON-NEIGHBOR H•GH t P • 0.05 I ::]::•ttt ...ttt p.o.oo,P- 0.00' , [:j.ii HIGH HIGH I

5o ,

' Low !i::i 0 iiiiil , MALES FEMALES MALES FEMALES

SEX OF TERRITORIAL DEFENDERS Fig. 9. Percentageof low- and high-level agonisticinteractions between neighboring and non-neigh- boring intrudersand adult males(total of 442) and females(total of 192), during the pre- and post-hatch periods.Results of Chi-squaretests for both intersexualcomparisons (*) and intrasexualcomparisons be- tween the pre- and post-hatchperiods (?) are indicated.

er agonisticvocalizations (Fig. 8). Similarly, bor). Both males and femalesdisplayed signif- during the post-hatchperiod, males showed icantly more high-level agonisticresponses to significantlyhigher frequenciesonly of long non-neighboringintruders than to neighbor- calls (t = 4.0, n = 19, P < 0.001). Both males ingintrudersduringbothpre-(x2= 29.56,df =1, and femalesdisplayed intrasexualincreases in P • 0.001 and X• = 6.06, df = 1, P • the frequenciesof both long-calls(males: t = 0.05, respectively)and post-hatchperiods (X 2 = 0, n = 19, P • 0.001; females: t = 0, n = 19, 147.22, df = 1, P < 0.001 and X2 = 26.07, df = P < 0.001) and yeows (Males: t = 12, n = 19, 1, P < 0.001, respectively;Fig. 7). During the P < 0.001; females: t = 11, n = 19, P < 0.001) post-hatchperiod, however, malesdisplayed a following chickhatching (Fig. 8). significantlygreater proportion of low-levelre- Males and females also displayed some dif- sponsesto neighbors(X • = 11.58,df = 1, P • ferencesin agonisticbehavior related to terri- 0.001) and high-level responsesto non-neigh- torial defense,depending on both the sex and bors (Xz = 4.68, df = 1, P < 0.05) than did fe- type of intruder (i.e. neighbor or non-neigh- males (Fig. 9). Comparisons of the pre- and January1983] SexualDifferences in Larus marinus Behavior 71

LOW LEVEL 0.01, respectively;Fig. 10). Female-femalein- teractionswere rare and accountedfor only 6% 100 PRE-HATCH • POST-HATCH of the agonisticencounters between neighbor- ing pairs. These data reflect the type of re- sponsedirected at approachingmales and fe- malesby defendersof the sameor oppositesex regardlessof the phase of the breeding cycle. Males approaching or intruding within a ter- 50 , ritorial boundary generallyelicited a low-level agonistic act from defenders of either sex. Ap- proachingor intruding females,however, were significantlymore likely to stimulate a high- level response(e.g. attack, charge,or chase)by both males (X• = 19.69, df = 1, P < 0.001) and

o HIGH LEVEL females(X • = 5.16, df = 1, P • 0.05; Fig. 11). Pre- and post-hatchcomparisons of male and female reactions to the involvement of their mates in an agonistic interaction (while both members of the pair were in attendance) re- vealed significant increasesin the proportion 50 of agonisticresponses by both males(X'-' = 6.46, df = 1, P < 0.05) and females (X'-'= 6.03, df = 1, P • 0.05). In addition, males were much more likely to exhibit agonistic responsesto aggressiveinteractions involving their mates than were females(X • = 4.93, df = 1, P • 0.01). M/M M/F F/F M/M M/F F/F Male involvement in these mate-neighbor in- teractions was often closely preceded or fol- SEX OF PARTICIPANTS lowed by the female mate's withdrawal and Fig. 10. Percentageof low- and high-level terri- mewing return to the vicinity of the chicks. torial interactionsthat occurredbetween three pos- Intermate agonistic interactions were infre- sible dyads; M/M = male-male (240 interactions), quent and generallyinvolved a male attacking M/F = male-female (147 interactions), F/F = female- its mate as she attempted to eat food regurgi- female (27 interactions). Resultsof Chi-square tests tated by the male to its chicks. Females were of comparisons(*) of the distribution of low- and never observed attacking their mates under high-level interactions between the pre- and post- similar circumstances. hatch periods are indicated. DISCUSSION

The resultsof the presentstudy of adult Great post-hatch periods revealed no significant Black-backed Gulls indicate that: (1) females changesin the frequenciesof intrusionsby invested more time than males in territorial at- either neighboring (,• = 0.12 + 0.03/h and tendanceand incubation during the pre-hatch 0.16 + 0.04/h, respectively)or non-neighbor- period; (2) there were no sexualdifferences in ing conspecifics(,t = 0.08 + 0.02/h and 0.10 + other types of offspring care or in territorial 0.03/h, respectively). attendance during the post-hatch period; (3) Analysis of agonistic interactionsbetween males were more aggressive than females neighboring territorial adults (in which the throughout the season;(4) both males and fe- sexes of both members of a dyad could be males displayed increasedinvestment in ago- identified) indicatedthat, during the pre- and nistic behavior during the post-hatch period, post-hatchperiods, low-level agonistic inter- as well as increased levels of long-calls and actionsoccurred primarily between males and yeows in agonisticcontexts; and (5) both sexes high-level encountersoccurred most often be- showeda significantincrease in the use of long- tween male and female defenders (X• = 14.89, calls, yelps, and mews in offspring-oriented df = 1, P • 0.001 and X• = 11.93, df = 1, P • contexts. 72 BUTLIiRAZaD JAZalis-BuTLER [Auk, Vol. 100

INTRUDING MALE INTRUDING FEMALE 100

NOR LOW HIGH NOR LOW HIGH AGON AGON AGON AGON

TYPE OF RESPONSE TO INTRUDER BY TERRITORIAL ADULTS Fig. 11. Type of response(NOR = no obvious response,low-level agonistic, or high-level agonistic) directedat intruding malesand femalesby defendingadults. Resultsof Chi-squaretests of comparisons(*) of the distribution of responsesdirected at intruders by males(317 responses)and females(114 responses) are indicated.

Agonistic behavior is an important invest- fenders of either sex, while female intruders ment for adults during the breeding season, were more likely to stimulate high-level re- due to the potential risk of injury during high- sponses.The increasedfrequency of agonistic level agonisticinteractions. Although attacks acts,as well as vocalizationsgiven in agonistic and fights are relatively rare, some adults do contexts,by both sexesduring the post-hatch sustain eye and wing injuries during such en- period was probably associated with move- counters (pers. obs.). When compared with ments of chickswithin the natal territory, the their mates,males consistently displayed higher presence and movements of chicks (and cor- frequenciesof agonisticbehavior, were more responding movements of adults) on neigh- likely to exhibit high-level agonisticbehavior boring territories, the relative increase in the towards non-neighboring intruders and female number of adult overflightsrelated to foraging neighbors, generally exhibited an aggressive trips to feed chicks, and/or the increased in- responseto agonisticinteractions between their vestment adults have in older offspring. Pre- mates and other adults, and were always dom- viously reported post-hatch increasesin terri- inant over their mates during interpair, ago- tory size in Glaucous-winged Gulls (L. nistic interactions. Pierotti's (1981) observa- glaucescens)(Hunt and Hunt 1976), Herring tions of female Western Gulls attempting to Gulls (Burger 1980), and Great Black-backed relieve their mates on the nest in response to Gulls (Butler and Janes-Butler 1982) are also the landing of an intruder concurs with our probably related to increasedoffspring mobil- post-hatch data concerning similar situations ity. involving L. marinuspairs. This phenomenon Any explanationof sexualand seasonaldif- may actuallyrepresent a "manipulation" on the ferences in vocalizations (and their associated part of the female for further investmentby the behavior patterns) necessitatessome interpre- male. In fact, increased aggressivenesson the tation of their functional significance.This is part of male gulls may extend throughout the sometimesdifficult, becausegull displays and year, as Monaghan (1980) observedthat male vocalizations may have multiple, context-de- Herring Gulls were dominant over femalesand pendent messagesand meanings (Beer 1975, juvenilesduring feeding interactionsin winter 1980), and it is not always possibleto deter- months. The relative risk to territory holders mine context definitively. Choking, mewing, of engagingin agonisticinteractions with nor- yelping, and long-calling occurredin a variety mally more aggressivemales may be reflected of contextsover the courseof the breedingcycle. in the fact that intruding males generallyelic- Greater pre-hatchfrequencies of male mewing ited low-level responsesfrom neighboringde- and choking (in mate-oriented contexts)may January1983] SexualDifferences in Larus marinus Behavior 73

PRE-HATCH lOO this context. Although the significanceof con- sistent sexualdifferences in yelping and long- callingduring both the pre- and post-hatchpe- riods remains unclear, the use of the latter in terms of advertisement and individual identi- fication (Beer 1975, 1980), as well as the pos- 50 sible functionof the yelp in this context,should not be overlooked. Stout's (1975; Stout et al. 1969) observationsthat the yelp of Glaucous- winged Gulls followed the termination of ag- onistic interactions are not inconsistent with this suggestedfunction. Gakkering and yeowing in Great Black- POST-HATCH backedGulls both appearedto functionas alert/ alarm vocalizations. Gakkering occurred fre• quently in agonisticcontexts. It was also ob- served, however, when a human intruder ap- t P < 0.05 proachedthe colony.Adults generallyyeowed 50 in responseto high-level agonisticinteractions nearby or when a chick wandered into the vi- cinity of a neighboringadult. This vocalization also was given in response to human intru- sionsinto the colony,however, and during rare overflights of Ravens (Corvus corax). Both

NOR AGON NOR AGON vocalizationsappeared to be highly sociallyfa- cilitated and often resulted in alerting large

MALE RESPONSE FEMALE RESPONSE portions of the colony. The typical chick's re- sponseto these calls was to walk quickly out Fig. 12. Type of response (no obvious re- of sight into densevegetation on the territory. sponse= NOR or agonisticresponse) of adult males We suggestthat the post-hatchincrease in fre- (42 responses)and females (108 responses)to ago- quenceof yeowing was relatedto the increased nistic interactionsinvolving their mates. Results of Chi-square testsfor both intersexualcomparisons (*) vulnerability of mobile offspringto both con- and intrasexual comparisonsbetween the pre- and specificaggression and predation. post-hatchperiods (?) are indicated. Head-tossing and its accompanyingvocali- zation have typically been ascribedthe func- tion of food-begging in female gulls during have been related to increased time investment courtship (Tinbergen 1959, 1960). Smith (1980) by females in incubation and corresponding has recentlysuggested that this type of behav- increasednest-relief attempts by males. The ior is actually an investment "demand" on the post-hatch increases in frequency of both part of the females of some monogamous mewing and yelping by both sexeswere prob- species. In L. marinus adults, males typically ably relatedto a major shift from agonisticand fed femalesduring the pre~eggperiod, but both mate-related encounters to parent-offspring members of the pair exhibited head-tossing interactions.Both vocalizations(often given in during the initial phasesof courtship-feeding combination) appeared in function to attract sequences.The behaviorpersisted in both sexes and/or locate chicks before feeding, when a throughoutthe incubationperiod (even though chick was separatedfrom the adult(s)in dense courtship-feedingdid not) and was most fre- vegetation, when a chick had wandered out of quently observed when one adult joined its its natal territory, or when a chickwas in the mate on the territory. Following chick hatch- vicinity of a neighboring adult. Evans (1980) ing, the frequency of male head-tossingde- has suggestedthat parent-chickrecognition of clined significantly,but femalesgenerally dis- individuallydistinctive mew callsin Ring-billed played this behavior during chick-feedingby Gulls is reinforced through feeding, and the the male. Continued female head-tossingfol- same may be true of the yelp of L. marinusin lowing chick hatching may represent further 74 BUTLERAND JANES-BUTLER [Auk, Vol. 100

"manipulation" of males that facilitates close that data in the present analysiswere collected approach of their mates to the site of chick after clutch completion and that territorial de- feedings.Males did not generallyreact aggres- fense appears to be related to chick survival sively to approachingfemales that displayed (Butler and Janes-Butler 1982), we considered head-tossing,although males often rapidly re- agonisticbehavior of Great Black-backedGulls consumedfood packagesunder these circum- in this study to be primarily parental invest- stances.Females were jabbed at or attackedby ment. Although L. marinusfemales did invest their mates only when they attemptedto con- statisticallymore time than males in territorial sume the food regurgitatedfor chicks.Females attendance(67% and 59%, respectively)and that obtained food in this manner were often incubation (57% and 43%, respectively),these observedfeeding chicksthemselves within an differences were small, and the activities in- hour. volved were relatively risk-free. In contrast, Trivers (1972)defined parental investmentas males appeared to maximize their parental ef- "any investmentby the parentin an individual fort by investing significantlymore in agonis- offspring that increasesthe offspring's chance tic behavior than did females during both the of surviving (and hence reproductivesuccess) pre- and post-hatchperiods. In the absenceof at the cost of the parent's ability to invest in sexualdifferences in other typesof parentalcare other offspring." In this regard, much of the (e.g. post-hatch attendance,brooding, chick- behavior displayedby an adult during the in- feeding, etc.), the results of the present study cubation and chick phasesof the reproductive strongly suggestthat parental investment by cycle(other than those that promote adult sur- male Great Black-backedGulls may equal that vival) may be consideredinvestment in off- of females. Pierotti (1981) observed a similar spring. Trivers hypothesizedthat females(due investment pattern in Western Gulls, but fur- to the larger size of their gametes)already have ther investigationwill be necessaryto confirm a greater investment in their offspring at fer- whether this is typical for larids in general. tilization and therefore should be willing fur- ther to invest more energy than males in pa- ACTONOWLEDGMENTS rental care. Gladstone (1979), however, has This researchwas funded in part by NIH grants questioned the actual "cost" of female gamete ES-00920and S07-RR-05764,NSF grantDEB-7826821, production relative to male energy expendi- and a Sigma Xi award. We are indebted to Dr. David tures and risks with regard to such activities Miller for sponsoringthis work, to the Mount Desert as territorial defense. In fact, recent studies of Island BiologicalLaboratory for the use of its facili- the breeding biology of Northern Gannets ties, and to the National AudubonSociety for its (Morus bassanus;Montevecchi and Porter 1980) continued cooperationin our research.Earlier drafts of the manuscriptbenefited substantially from the and Black Skimmers (Rynchopsniger; Burger clarifyingcomments of Drs. J. Byers,W. Montevec- 1981b) have stated that male parental invest- chi, R. Pierotti, E. O. Price, and N.J. Volkman. Fi- ment may equal or actually exceedthat of fe- nally, our specialthanks to Capt. Steen L. Mery- males. Care must be taken, however, in des- weatherfor his excellentlogistical support in the field. ignating behavioralacts related to reproductive effort as either parental investment or mating LITERATURE CITED effort. For example, male territorial defense be- BEER,C.G. 1975. Multiple functionsand gull dis- fore clutch completionmay aid in avoidanceof plays. Pp. 16-54 in Function and evolution in cuckoldry,and agonisticbehavior during this behavior--essays in honor of ProfessorNiko period would be consideredprimarily mating Tinbergen. (G. P. Baerends,C. G. Beer, and A. effort. Manning, Eds.). Oxford, ClarendonPress. Moreover, interpretations of the relative in- 1980. The communicationbehavior of gulls vestments of males and females in their off- and other . Pp. 169-205, in Behavior of spring must be formulatedcautiously because marine animals,vol. 4: marine birds. (J. Burger, B. Olla, and H. Winn, Eds.). New York, Plenum of the difficulty of equating suchfactors as the Press. cost of gamete production, time spent in ter- BURGER,J. 1980. Territory size differencesin rela- ritorial attendanceand incubation, the energy tion to reproductive stage and type of intruder and risks involved in agonisticbehavior, and in Herring Gulls (Larusargentatus). Auk 97: 733- future benefits accrued from territorial defense 741. by site-tenaciousspecies. Due both to the facts 1981a. On becoming independent in Her- January1983] SexualDifferences in Larus marinus Behavior 75

ring Gulls: parent-young conflict. Amer. Natur. rental investmentsby seabirdsat the breeding 117: 444-456. area with emphasison Northern Gannets, Morus 1981b. Sexualdifferences in parentalactiv- bassanus.Pp. 323-365, in Behaviorof marine an- ities of breedingBlack Skimmers. Amer. Natur. imals, vol. 4: marine birds. (J. Burger, B. Olla, 117: 975-984. and H. Winn, Eds.). New York, Plenum Press. --, & C. BEER. 1975. Territoriality in the MOYNIHAN, M. 1958. Notes on the behaviour of Laughing Gull (L. atricilla). Behaviour 55: 201- some North American Gulls, II: non-aerial hos- 220. tile behaviour of adults. Behaviour 12: 95-182. BUTLER,R. G., & S. JANES-BUTLER. 1982. Territo- PIEROTTI,R. 1980. Spite and altruismin gulls.Amer. riality and behavioralcorrelates of reproductive Natur. 115: 290-300. success of Great Black-backed Gulls. Auk 99: 58- 1981. Male and female parental roles in the 66. Western Gull under different environmental , & W. TRIVEL?IECE.1981. Nest spacing, re- conditions. Auk 98: 532-549. productive success, and behavior of the Great SMITH, S. M. 1980. Demand behavior: a new in- Black-backed Gull (Larus marinus). Auk 98: 99- terpretation of courtship feeding. Condor. 82: 107. 291-295. EVANS,R.M. 1980. Developmentof individual call SOUTHERN, L. K. 1981. Sex-related differences in recognition in young Ring-billed Gulls (Larus territorial aggressionby Ring-billed Gulls. Auk delawarensis):an effectof feeding.Anim. Behav. 98: 179-181. 28: 60-67. STOUT,J. F. 1975. Aggressivecommunication by GLADSTONE,D. E. 1979. Promiscuityin monoga- Larusglaucescens. III. Descriptionof displaysre- mous colonial birds. Amer. Natur. 114: 545-557. lated to territorial protection. Behaviour 54: 181- HAND, J. L. 1981. A comparisionof vocalizations 208. of Western Gulls (Larus occidentalisoccidentalis --, C. WILCOX,& L. CREITZ. 1969. Aggressive and L. O. livens). Condor 83: 289-301. communicationby Larusglaucescens. I. Sound HUNT, G. L., JR., & M. W. HUNT. 1976. Gull chick communication. Behavior 34: 29-41. survival:the significanceof growthrates, timing TtNBERGEN,N. 1956. On the functionsof territoW of breeding,and territoW size. Ecology57: 62- in gulls. Ibis 98: 401-411. 75. 1959. Comparative studies of the behav- INGOLFSSON,A. 1967. The feedingecology of five iour of gulls (Laridae):a progressreport. Behav- speciesof largegulls (Larus) in Iceland.Unpub- ior 15: 1-70. lishedPh.D. dissertation.Ann Arbor, Michigan, 1960. The Herring Gull's world. New York, Univ. Michigan. Basic Books, Inc . MONAGHAN,P. 1980. Dominanceand dispersalbe- TmvERs, R. L. 1972. Parental investment and sex- tween feeding sites in the Herring Gull (Larus ual selection.Pp. 136-179in SexualSelection and argentatus).Anim. Behav. 28: 521-527. the Descentof Man (B. Campbell,Ed.). Chicago, MONTEVECCHI,W. A. & J. M. PORTER. 1980. Pa- Aldine. NEST MICROCLIMATE, WATER-VAPOR CONDUCTANCE, AND WATER LOSS IN HERON AND TERN EGGS

CAROLMASTERS VLECK, 1 DAVID VLECK,1 HERMANN RAHN, AND CHARLES V. PAGANELLI Departmentof Physiology,State University of New Yorkat Buffalo,Buffalo, New York 14214USA

ABSTRACT.--Ratesof water loss (/•lH•o) were measured in eggsof sevenspecies of tree- nestingCiconiiformes and three speciesof ground-nestingCharadriiformes during natural incubation.Measurements of eggtemperature, conductance of the eggshell to water vapor, and nestand ambienthumidity allow oneto calculatethe differencein water-vaporpressure (Ap) betweenegg and nest and betweennest and ambient air. Ap betweenegg and nest was significantlylower in ground-nestingspecies (23 torr) than in tree-nestingspecies (31 torr). We suggestthat low Ap in ground-nestingspecies was due to relativelyhigh humidity (ca. 27 torr) in nestsbuilt on the ground comparedto the humidity in loosestick nestsof tree-nestingspecies (ca. 14 torr). Water-vaporconductance of eggsfrom more humid nests is relativelyhigher than that of eggsfrom lesshumid nests(107% and 78% of predicted values, respectively),resulting in similar fractionallosses in massfor the whole incubation periodin bothgroups of birds.Mn2o was not alwaysconstant throughout incubation. We suggestthat changes in Mn•omay be dueto changesin ambienthumidity and/or egg tem- perature. Received29 December1980, resubmitted5 January1982, accepted5 July1982.

AVIAN eggslose water by diffusion through sticksand twigs and are sometimesso flimsy pores in the shell. Total water lossduring in- that the eggscan be seenthrough the bottom cubation must be within certain limits for suc- of the nest. The and gulls nest on the cessfuldevelopment. Hatchability of chicken ground. The ternslay their eggsin a scrapeon eggsis optimal when total water lossis 10-12% exposedsand, whereas the gullsconstruct nests of initial eggmass (Lundy 1969),and water loss of vegetationthat are usuallyplaced under cov- of naturally incubated eggs of a wide variety er of grassor small shrubs. of birds averagesabout 15% (Drent 1975, Ar We measured rates of water loss and water- and Rahn 1980). The rate of water loss from vapor conductanceof naturallyincubated eggs eggs is determined by: (1) the pore geometry in each speciesand also measured egg tem- (length, cross-sectionalarea, and number of perature, nest humidity, and ambient humid- pores)of the eggshell,which determineswater- ity in four of thesespecies. Results of this study vapor conductance;(2) egg temperature,which suggest that the difference in water-vapor determines water-vapor pressure within the pressurebetween an egg and its microenviron- egg; and (3) parental behavior, nest structure, ment is less in the ground nestsof gulls and and ambient conditions, which interact to de- terns than it is in the tree nests of herons. termine water-vapor pressure in the nest. We investigatedthe interactionof thesefac- MATERIALS AND METHODS torsin the water economyof 10 speciesof birds. Birdsand study area.--We studied eggsof 7 species Although all the speciesstudied nest in colo- of herons, egrets, and ibis (Table 1), and 3 species nies on small islands off the coast of North Car- of gulls and terns. The Ciconiiformes were nesting olina, their nesting habits fall into two distinct in trees on two small islands in the estuary of the categories.The herons, egrets, and ibis build Newport River nearBeaufort, North Carolina.Royal their nests in trees, usually 2-5 m above the Terns (Sterna maxima), Sandwich Terns (S. sandvi- ground. Their nests consistof loosely woven censis),and LaughingGulls (Larusatricilla) nested on Morgan Island, a dredge spoil island near the tip of Cape Lookout,North Carolina. The two terns nested • Presentaddress: Department of Ecologyand Evo- togetherin densecolonies on sparselyvegetated or lutionary Biology, University of Arizona, Tucson, baresand, and LaughingGulls nested on the ground Arizona 85721 USA. in nearby vegetation. 76 The Auk 100:76-83. January1983 January1983] WaterLoss in Heronand Tern Eggs 77 • • •r

r• 30 , , o œ300 œ400 f00100 Time (hours)

top center.,r recorder F•G. 1. Great Egret egg temperature.The top trace is the temperatureat the top of the egg, adjacentto the brood patch, and the lower traceis the temperatureat the centerof the egg. Egg position was maintained in the nestby the thermocouples,which were threadedthrough the nest and run to the chartrecorder.

Ratesof water Ioss.--The changein egg mass dur- with the top thermocoupleadjacent to the brood patch ing incubation is due almost entirely to water loss of the adult. The egg was anchoredin place by the (Drent1970). We determinedrates of waterloss (Mn2o) thermocoupleleads passing through the bottom of from naturally incubatedeggs by weighing the eggs the nest. Royal Terns usually lay a single egg, and in the field at 2- to 6-day intervals with a Torbal orientation of the thermocoupleegg in the nest cup torsion balancereadable to _+0.01g. was fixed by burying the thermocoupleleads in the Physicalcharacteristics of the egg.--The water-va- sand under the egg. por conductance(Gmo) of eggs was measured using Egg temperaturewas recorded continuouslywith the method of Ar et al. (1974) in which water loss of an Omega thermocoupleD.C. millivolt amplifier and eggs in a desiccatorat constanttemperature is mon- a battery-operated Linear chart recorder. We cali- itored by measuringchange in mass.G•t•o varies with brated the recorder-thermocouplesystem before and egg temperatureand barometric pressure.To permit after each series of measurementsagainst an NBS- comparisonwith other published data we converted certified mercury thermometer.All the speciesstud- all our reported values to 25øC and 760 torr (101.3 ied incubated essentiallycontinuously, so egg tem- kPascals)using the proceduresdescribed by Paga- peraturesdid not vary by more than a few degrees. nelli et al. (1978). The Gn•oat an incubation temper- The egg temperatureswe report are means of tem- ature of 37øC is 1.02 times the value at 25øC. peratures taken at 30-min intervals from continuous The initial mass of eggs that were collectedwas recordsmade during uninterrupted incubation. measured by refilling the air cell with water at the Water-vapor pressure.--Water-vapor pressures in conclusionof the experiment. For eggs that were not the macroenvironment around the nests (P•) were collected,we calculatedinitial mass from the egg's measured in two ways. We calculatedP• from the dimensions by the method of Hoyt (1979), using relative humidity and temperaturerecord of a clock- species-specificweight coefficients(K,.) determined driven Serdexhygrothermograph situated about 0.5 from the collectedeggs. m above the ground in the vicinity of the nests.The Egg temperatureduring incubation.--We measured hygrothermographwas calibratedusing saturatedsalt egg temperatureswith forty-gaugecopper-constan- solutions (Winston and Bates 1960), and was found tan thermocouplesimplanted in eggsof Great Egrets to be accuratewithin 1 torr (= 133 Pa) in the range (Casmerodiusalbus), Cattle Egrets (Bulbulcusibis), of temperatures and relative humidities encoun- White Ibis (Eudocimusalbus), and Royal Terns. Two tered. We alsomeasured mean Pt using "egg hygro- thermocoupleswere glued into the bird's own egg meters" (Rahn et al. 1977). These consisted of chick- using fast-setting epoxy, one at the top of the egg en eggshellsof known GH•o,filled with silica gel and and one in the centerof the egg (seeFig. 1). In egret weighed to the nearest0.1 mg in the laboratory be- and ibis nests the thermocoupleegg was placed as fore and after the period of measurement. Because closeas possibleto the centerof the 2- to 4-egg clutch, 'he desiccantholds humidity inside the shell near 78 Vn•ca •t ^n. [Auk, Vol. 100 January1983] WaterLoss in Heronand Tern Eggs 79

' TABLE2. Ambient and egg temperaturesmeasured 340 at the center and top of the egg from two Great Egretnests and one nestof eachof the other species. Numbers shown are meansof temperaturestaken WhiteIbis at 30-min intervals + SD; number of hours of ob- 300 servationin parentheses.

Temperature(øC) 260 Egg Species Ambient Center Top Great Egret 16.7 34.3 + 1.4 36.9 + 1.6 (27.5) (28.5) 21.7 36.8 + 0.8 37.5 + 1.1 180{..._..._•/ • rat (•) (25) White Ibis 20.5 -- 37.1 + 1.0 140 C• (20) Royal Tem 24.3 37.8 + 0.7 39.2 + 1.2 11.6 13.5 18.4Ambient PH20 (13) (9.5) IO0• 4 • i• lb 2• DaysAfter First Weighing declined between intervals and declined sig- FIG. 2. Ratesof water lossfrom eggsas a function nificantly (P < 0.001) during the last interval. of time. Lines connectmean valuesduring each in- Therewas no consistenttrend in Mn2oof the terval plottedat the midpoint of the interval. Vertical other species.Mean valuesfor eachspecies are bars extendonly +1 standarderror for clarity of pre- listed in Table 1. sentation. Number of eggs in each group are: Royal Tem (12), SandwichTem (12), CattleEgret (13), White The GH2ovalues for the various species are Ibis (8), and LaughingGull (13). Numbers under the listed in Table 1. White Ibis eggshad an un- values for Cattle Egret eggs are the mean ambient usuallywide range of GH•ovalues, from 3 to 15 water-vapor pressuresfor the same interval rag(day-torr) L Mean temperature in the middle of the egg rangedfrom 34.3øCin GreatEgret eggs to 37.8øC zero, meanwater-vapor pressure (PH2o) around such in RoyalTern eggs(Table 2). A significantver- a hygrometeregg can be calculatedfrom the equation tical temperaturegradient existed through all PH2o= Mn2o/Gu2o,where MH•ois the rate of water of the eggsmeasured. Temperature at the top absorptionof the eggin mg.day-] and GH,ois water- of the egg, that part in closestcontact with the vaporconductance in mg(day.torr)-L We placedegg brood patch of the incubating adult, averaged hygrometersunder nestsof Great Egretsand White 1.6øChigher than at the middle. Temperatures Ibis for 2-3 days to determinemean Pt and at the sametime placedanother egg hygrometer in the nests were recordedsimultaneously at the centerand to determinewater-vapor pressurearound the eggs top of one Great Egretegg for 14 h, and a mean in the nest (Px). The chicken-egghygrometers dif- gradientof 4.5øC+ 1.2 (.• + SD) wasfound over fered in colorand were slightlylarger than the birds' a distanceof 20.5 mm (Fig. 1). own eggs,but with the exceptionof oneGreat Egret Ambient and nest water-vapor pressuresfor that destroyeda hygrometer,the adult birds incu- CattleEgret, Great Egret, White Ibis, and Royal bated hygrometereggs along with their own. Tern nests are shown in Table 3. Mean P• val- ues determined with the hygrothermograph RESULTS and with egg hygrometersplaced directly un- der heron nests over the same time interval Mean daily ratesof water lossvaried directly were not significantlydifferent (P > 0.5), and with egg size, ranging from 121 mg'day -• in no distinction is made between the two meth- Snowy Egret (Egretta thula) eggs to 321 mg. ods. day-• in RoyalTern eggs(Table 1). The M,2o The Px values were calculated as follows. of Sandwich Tern and Royal Tern eggs in- Water lossfrom an egg can be expressedby a creasedsignificantly (paired t-test, P < 0.01and form of the Fick diffusion equation as P < 0.001,respectivel.y) from each interval to the next (Fig. 2). The MH2oof CattleEgret eggs •u2o = Gu•o(P•- Ps), (1) 80 VLECKET AL. [Auk, Vol. 100

T,•SrE 3. Mean water-vapor pressurein eggs (PA), in nests around eggs (Px), and in ambient air around nests (Pt) during incubation in four speciesof birds. T,,.,,,is estimated egg-surfacetemperature. See text for discussion of calculations.

Estimated Px (torr) T,... PA Ap = p.• - Px' P4 -- Ap Egg hy- Pt Species (øC) (torr) (torr) grometer (torr) Great Egret 35.6 44 30 14 15 12 Cattle Egret 37.2 48 30 18 -- 16 (17-22) • White Ibis 37.1 47 36 11 14 11 Royal Tern 38.5 51 24 27 -- 15 (26-30) •

Range of P,• given range in •,P (Table 1). where P.• is the water-vapor pressure in the tree-nesting Ciconiiformeswas 30.7 + 3.3 torr egg and P,vis the water-vapor pressurein the (• _+SD), whereas the mean for the three Cha- nest around the egg (Rahn and Ar 1974). The radriiformeswas only 23.0 + 0.6 torr. The dif- term ZIP = (P.4 - Px) is the driving gradient ferencebetween the two groups is significant for water loss and can be estimated as ZIP = (t' = 5.95, P < 0.01). MH2o/GHzo(Table 1). The Px is then equal to

(P..•- AP) and can be calculatedin two ways. DISCUSSION First, for hygrometereggs incubated with the naturalclutch, P.a = 0, soP.•. = MH•O/GH•O.Here, Nest humidity.--The data we present can be Mn.,ois the rateof wateruptake, not waterloss. used to comparethe microclimatearound eggs Second,for a natural egg, P.l approximatesthe in the loose stick nests of herons, egrets, and water-vapor pressure at saturation at egg tem- ibis with that around eggsin the ground nests perature,so Px = P,4- MH2o/Gn.•o.Both tech- of gullsand terns.Our resultsshow that ground niques yield similar estimatesof Px in Great nests are significantlymore humid than tree Egret and White Ibis nests.We used only the nests in the same macroenvironment. second method to estimate Px in Royal Tern The difference in zIP between tree- and andCattle Egret nests. Because average MH.,O, ground-nesters(Table 1) must result from dif- and thus zIP, varied over time in these two ferences in P•, Px, or both between the two species,we calculateda range of Px using the classes of nests. We did not measure P.• di- highestmean /•iH•O and the lowest mean rectly, but used egg temperaturesto estimate (Table 3). P.• values (see Results). If incubation temper- Egg-surfacetemperatures used to estimate atures of gull and tern eggs were 3-4øC lower were assumedto be midway between temper- than incubationtemperatures of egret, ibis, and atures recorded from the top and middle of heron eggs,P.• and zIP of the former would be continuouslyincubated eggs (Table 2), except reducedby enoughto accountfor the observed for White Ibis eggs, for which we measured difference (7.7 torr) in zIP. This is not the case, temperatureat the top of the eggonly. We sug- however. The averagetemperature of a Royal gest later that egg temperature,and thus P.•, Tern egg was actually higher than tempera- may rise during incubation. If a temperature tures of Great Egret, Cattle Egret, or White Ibis of 35øC was associated with the earliest and eggs (Table 2). For seven speciesof gulls and lowestmeasured M•o and a temperatureof terns for which egg temperatures have been 39øCwith thelatest and highest measured MH•o reported (Howell and Bartholomew 1962, Drent in the Royal Tern, then Px would changefrom 1970, Morgan et al. 1978, Rahn and Dawson 21 to 28 torr over the course of incubation. 1979, Pettit et al. 1981), average egg tempera- The zIP values for each of the 10 specieswe tures are in no case lower than those we mea- studiedare reporte.d in Table1. Forthe three sured in egretsand ibis. speciesin which Mn.mwas not constant, the We conclude that the difference between zIP range in zIP is shown. The mean zIP for seven values for the two groups must be due to a January1983] WaterLoss in Heronand Tern Eggs 81 differencein Px. That is, the ground nests of Pt - Px between ground nestersand tree nes- gulls and terns are more humid than tree nests ters is marginallysignificant (Mann-Whitney of herons, egrets, and ibis in the same mac- U = 19, P = 0.1). If we include data from the roenvironment. In four species, we have the four speciesreported here, the differencebe- data necessaryto estimate Px itself, and not comeshighly significant(P < 0.01). just ZIP(Table 3). The estimatedmean P,v around Eggwater-vapor conductance.--Nest humid- a Royal Tern egg is 9-16 torr higher than that ity is of physiologicalsignificance to a devel- around Great Egret, Cattle Egret, or White Ibis opingembryo, because it affectsMn2o and the eggs. hydrationstate of thatembryo, but Mn2o de- High Px in ground-nesters initially seems pends on Gn2oas well as on zIP. In order to surprising. To the human observer the bare compareGn2o values between different species, open sand on which the terns nest is a more it is necessaryto take the effectsof egg size desiccating environment than are the often and incubationperiod into account.Hoyt (1980) shaded nest sites of herons. Both groups of presentedan equation that expressesthe in- birds incubate nearly continuously, so differ- terrelation between these parametersfor 143 ences in parental attentivenessare unlikely to speciesof birds: account for the differences in Px. Rather, the G,•o = 2.32 (Mø'97•/1ø'7:•), (2) likely explanationfor the observeddifference can be found from a consideration of the sources where M is initial egg mass in g and I is in- of water vapor in a nest and the avenues for cubationperiod in days. If our measuredGn2o water loss from the nest. Tree-nesting herons valuesare expressedas a percentageof the val- build loose stick neststhat do not presentmuch ue predictedfor eggsof the sameM andI from of a barrier to convectiveand/or diffusive gas equation2, G,•o averages78 + 11% (2 + SD) exchange.Water vapor added to the nest from of the predicted values for 7 tree-nestersand the egg or incubating adult is quickly lost to 107 + 6% for 3 ground-nesters.These two per- the ambient air. When ambient conditions are centagesare significantlydifferent (t'= 5.36, similar, ground nestshave higher Px than tree P < 0.01). That is, eggs laid in stick nests in nests for three reasons:(1) the ground itself is trees, which are incubated in less humid mi- an important source of water vapor (Geiger croenvironments, have relatively lower G n2o 1965), (2) the relatively solid surface of the than eggsof ground nesters,which are incu- ground reduces water loss from the nest by bated in more humid microenvironments. convectionand diffusion, and (3) wind speed Lomholt (1976)and Birchard and Kilgore (1980) on the ground is less becausenests are in the reported a similar relationshipbetween Gmo boundary layer. The Px of the Great Egret and and nest humidity in severalother speciesof White Ibis nests measuredwith egg hygrom- birds. eterswas only 0-3 torr higher than water-vapor Because Gn.•o is high in ground-nesting pressure in the macroenvironment(Pt), while species,where zIP is low, and low in tree-nest- the calculatedmean Px in the Royal Tern nests ing species,where zIP is high, the total water was 12 torr higher than water-vapor pressure lossduring incubation, expressedas a fraction in the macroenvironment (Table 3). of initial mass(F), is not significantlydifferent Our data provide direct evidence that nest betweenthe two groups(Table 1, t = 0.44, P > site and structurecan have a significanteffect 0.5). Estimatesof F dependon Mn2o,which is on the gradient in water-vapor pressure be- not alwaysconstant (see below). The measure- tween the nest and the macroenvironment. Data mentsof Mn.•othat we usedto estimateF ex- compiled by Walsberg (1980) provide an in- tended over 45-90% of the incubation period dependenttest of this hypothesis.In 10 ground- for eachspecies, however, so it is unlikely that nesting speciesthe mean gradientof water-va- Mn.•ocould vary enoughduring the remainder por pressure between nest and ambient con- of the incubation to change the conclusions ditions (Pt - P.•) was 8.2 + 4.3 torr (,t + SD). above. For two tree-nesting species(Anous tenuirostris Changesin ratesof waterloss from eggs.--In and Gygisalba) the mean gradientswere 3 and the RoyalTern, SandwichTern, andCattle Egret 0 torr, respectively.Despite differencesamong daily rates of water loss during natural incu- investigators,their techniques,and the habi- bation changedwith time (Fig. 2). Becausefew tats in which they worked, the difference in studiesof egg water losshave been carriedout 82 VLECI(ET AL. [Auk, Vol. 100

over the whole incubation period, it has gen- a gradual increasein Pt, the averagehumidity erally been assumedthat A;IH2oremains con- in the macroenvironment(Fig. 2). The close stant. Drent (1970), however, showed that the correlation between P,•. and Pt in ciconiiform A;iH2oincreased during the earlystages of in- nests(Table 3) suggeststhat Px might have in- cubationin the Herring Gull (Larusargentatus), creasedby an equivalentamount, about 7 torr. andA•iH•o of eggsof fourother gulls also varies All other things being equal, this would reduce over time (Morgan et al. 1978, Rahn and Daw- AP, the gradient driving water loss from the son 1979).Change in rate of water lossfrom an egg, and thereforereduce A•l•2o (see equation egg can result from one or more of the follow- 1). The predictedreduction in M•o closelyap- ing: (1) changein G•.,o; (2) changein nest hu- proximatesthe observedchanges. midity, P.v;or (3) changein PA, due to change The observed increases of 20% and 17% in in surfacetemperature of the egg. •I•o in Sandwichand Royal terns, respective- Water-vapor conductanceof eggs increases ly (Fig. 2), cannotadequately be accountedfor during the first few days of incubation in sev- by changesin Pt. At least part of the increased eral speciesof passerinesbut is relatively con- rate of water loss may be due to an increasein stant after the development of the chorioallan- surface temperature of the egg. Drent (1970) tois (Carey 1979, Hanka et al. 1979, Birchard found that temperature in the nest of the Her- and Kilgore 1980, Sotherlandet al. 1980). Most ring Gull and in the center of the egg rose of our measurementsof G•2owere made using steadily during the first two-thirds of incuba- eggs that had been incubatedat least 10 days tion. In addition, heat production of the de- (the sameeggs in which we measured•lH•o). veloping embryo late in incubationis sufficient Exceptionswere the LaughingGull eggs,which to raiseegg-surface temperature in somespecies were all freshlylaid, and someWhite Ibis eggs. (Drent 1970,Vleck unpubl.). Even at a constant We could detectno significantchange in Gmo nest humidity, the observed17% increasein of White Ibis eggs during development (t = A•i•oin RoyalTern eggscould be accounted 0.98, P • 0.2). The G•o of 12 freshlylaid eggs for by only a 1.3øCrise in mean surfacetem- was 8.59 + 2.20 mg(day'torr) • (27+ SD), and peratureof the egg. A completeunderstanding the G•o of 13 eggscontaining well-developed of thecauses of variationin •I•.,owill require embryoswas 7.00 + 2.49 rag(day'tort) -•. It may measurementof G•2o, P.¾.and egg temperature be that G•2oincreases with developmentof the throughout the incubation period. chorioallantoisonly in small (<10 g) eggs, as Birchardand Kilgore (1980)suggested. Even if ACKNOWLEDGMENTS GH2o changes early in incubation,such c.hanges We are indebted to Drs. Joseph and Celia Bona- cannot account for the variations in Mmo of ventura and the Duke University Marine Laboratory Royal Tern and Cattle Egret eggsvery late in for the use of laboratoryspace and facilities during incubation. our stay in Beaufort,North Carolina, and to Phyllis In the absence of any evidence that G•o Parisi and John Fussell, who provided assistancein changesduring middle to late incubation,these the laboratoryand in the field. This study was sup- changesin •lmoare presumably due to changes ported by NSF Grant PCM-7620947. in nest humidity, egg temperature(and hence Pit) or both. Our measurementsof egg tem- LITERATURE CITED peratures and nest humidities did not extend AR, A., & H. RATON.1980. Water in the avian egg: overthe entireperiod of A•i•2omeasurements, overall budget of incubation. Amer. Zool. 20: so we cannot be certain of what caused the 373-384. observedchanges in A•i•o.The available data, , C. V. PAGANELLI, R. B. REEVES, D. G. however, suggestthat changesin P.v may be GREENE,& H. RAHN. 1974. The avian egg: sufficientto accountfor the variationin •I• water vapor conductance, shell thickness, and in Cattle Egret eggs, but not in the two tern functional pore area. Condor 76: 153-158. BIRCHARD,G. F., & D. L. KILGORE. 1980. Conduc- species. tance of water vapor in eggsof burrowing and For 7 of the 10 speciesin which we measured nonburrowing birds: implications for embry- A;i•2o,neither A;iH2o nor ambienthumidity onic gas exchange.Physiol. Zool. 53: 284-292. changed significantly during the period of CAREY, C. 1979. Increase in conductance to water measurement.The gradualdecline in A;i•o of vapor during incubation in eggs of two avian CattleEgret eggs, however, was correlatedwith species.J. Exp. Zool. 209: 181-186. January1983] WaterLoss in Heronand Tern Eggs 83

DRENT, R. 1970. Functional aspectsof incubation MORGAN, K. R., C. V. PAGANELLI,& H. RAHN. 1978. in the Herring Gull. Behav. Suppl. 17: 1-132. Egg weight loss and nest humidity during in- 1975. Incubation. Pp. 333420 in Avian bi- cubation in two Alaskan gulls. Condor 80: 272- ology,vol. 5 (D. S. Farner and J. R. King, Eds.). 275. New York, Academic Press. OPPENHEIM,g.W. 1972. Prehatching and hatching GEIGER, g. 1965. The climate near the ground. behavior in birds: a comparativestudy of altri- Cambridge,Massachusetts, Harvard Univ. Press. cial and precocialspecies. Anim. Behav. 20: 644- HANKA, L. g., G. C. PACKARD,P. g. SOTHERLAND, 655. t. L. TAIGEN, t. J. BOARDMAN,& M. J. PACKARD. PAGANELLI,C. V., R. A. ACKERMAN,&: H. RAHN. 1979. Ontogenetic changesin water-vapor con- 1978. The avian egg: in vivo conductancesto ductance of eggs of Yellow-headed Blackbirds oxygen,carbon dioxide, and water vapor in late (Xanthocephalusxanthocephalus). J. Exp. Zool. 210: development. Pp. 212-218 in Respiratory func- 183-188. tion in birds, adult and embryonic (J. Piiper, HOWELL, T. R., & G. A. BARTHOLOMEW. 1962. Ed.). New York, Springer-Verlag. Temperature regulation in the Sooty Tern Sterna PETTIT, T. N., G. S. GRANT, G. C. WHITTOW, H. fuscata. Ibis 104: 98-105. RAHN, & C. V. PAGANELLI. 1981. Respiratory HOYT, D. F. 1979. Practicalmethods of estimating gas exchangeand growth of White Tern em- volume and fresh weight of bird eggs. Auk 96: bryos. Condor 83: 355-361. 73-77. RAHN, H., & A. AR. 1974. The avian egg: incuba- 1980. Adaptation of avian eggs to incuba- tion time and water loss. Condor 76: 147-152. tion period: variability around allometric regres- , & W. R. DAWSON. 1979. Incubation water sions is correlated with time? Amer. Zool. 20: loss in eggs of Heermann's and Western gulls. 417425. Physiol. Zool. 52: 451460. LANGHAM, N. P. E. 1974. Comparative breeding , R. A. ACKERMAN,&: C. V. PAGANELLI. 1977. biology of the Sandwich Tern. Auk 91: 255-277. Humidity in the avian nest and egg water loss LOMHOLT,J. P. 1976. Relationship of weight loss during incubation. Physiol. Zool. 50: 269-283. to ambient humidity of bird eggs during incu- SOTHERLAND,P. R., G. C. PACKARD,T. L. TAIGEN, bation. J. Comp. Physiol. 105: 189-196. &: T. J. BOARDMAN. 1980. An altitudinal cline LUNDY, H. 1969. A review of the effects of tem- in conductance of Cliff Swallows (Petrochelidon perature, humidity, turning, and gaseousenvi- pyrrhonota)eggs to water vapor. Auk 97:177 ronment in the incubator on the hatchability of 185. the hen's egg. Pp. 143-176 in The fertility and WALSBERG,G. E. 1980. The gaseousmicroclimate hatchabilityof the hen's egg (T. C. Carter and of the avian nest during incubation. Amer. Zool. B. M. Freeman,Eds.). Edinburgh,Oliver & Boyd. 20: 363-372. MAXWELL,G. R., & H. W. KALE,II. 1977. Breeding WINSTON, W., &: D. H. BATES. 1960. Saturated so- biology of five speciesof herons in coastalFlor- lutions for the control of humidity in biological ida. Auk 94: 689-700. research. Ecology41: 232-237. BEHAVIOR, PREY SELECTION, AND ENERGETICS OF SNAIL KITES IN GUYANA: CONSUMER CHOICE BY A SPECIALIST

STEVEN R. BEISSINGER l Departmentof Zoology,Miami University,Oxford, Ohio 45056 USA

ABSTRACT.--Thehunting behavior, snail size selection,and time-activitypatterns of non- breedingSnail Kites (Rostrhamus sociabilis sociabilis) were studied in Guyanarice fields. Kites spent62% of the photoperiodperching, 19% foraging,13% in maintenanceactivities, and 6% flying. As the day progressed,the percentageof daylighthours spent perching increased significantly,while foraging decreasedsignificantly. Kites successfullycaptured and in- gestedPomaeea snails in 78% of the foraging bouts observed.The mode of hunting was evenly split between coursing(50.7%) and still-hunting (49.3%). Searchingand returning times were related to time of day, as significantlymore coursing hunts and still-hunts occurredin mornings and late afternoons, respectively.Prey handling time was the most time-consumingcomponent of a foraging bout. Time spent searchingfor prey, returning, or handling prey was not related to snail size. The size distribution of captured snails differed significantly from that of available snails; kites selectedmore medium snails and fewer smallsnails and took large snailsin equal frequencyto that at which largesnails were available. Kites capturedthe same-sizedsnails when coursingas when still-hunting. As rice grew, kite utilization of rice fields declined, and the frequency of still-hunting decreased while course-huntingattempts increased.The daily caloric intake of kites was estimated to be 104.2 kcal, the daily energy expenditure85.7 kcal. Resultsare discussedin relation to consumerchoices of specialists.The most important decision a foraging Snail Kite may make is what patch to searchin and how long to searchbefore abandoningpatches. Received 14 January1982, accepted28 June1982.

LITTLEis known about the hunting behavior detail how kites extract and consume snails, and foraging ecologyof the Snail Kite (Rost- but there is no quantitative information on the rhamussociabilis). Long recognizedas an ex- behavior and energeticsof kite foraging.In this treme diet specialist, the Snail Kite feeds paper, ! describethe foragingecology of the almost exclusively on Pomaceasnails (Haver- Snail Kite (R. s. sociabilis) on the basis of field schmidt 1962, Snyder and Snyder 1969, Voous observationsof hunting behavior, activity pat- and van Dijk 1973;for exceptionssee Sykes and terns, and prey selectionin rice fields in Guy- Kale 1974, Beissingerin prep.). A kite captures ana, South America. a snail by flying over a marsh, extendinga foot, and grabbing the molluscwith long toes near STUDY AREA AND METHODS the water surface. It then perches in order to Studies were conductedfrom July-August 1977 at extract the body from the shell with its thin, the RiceResearch Station and surroundingrice fields sharply hooked bill. Two modes of hunting on the coastallowlands (1.4 m below sea level) at the have been observed(Snyder and Snyder 1969, Mahaicony and Arbary Rice DevelopmentScheme Haverschmidt 1970): "still-hunting" kites vi- (MARDS), Burma, Guyana, South America (6ø28'N, sually searchfrom a perch and capture a snail 57ø45'W).The freshwatermarshes or mangrovesthat once covered this region have been almost totally near that perch after a short flight; "course- replaced by rice cultivation (6,721 ha in 1976; Ken- hunting" kites fly 3-5 m above a marsh, usu- nard pers. comm.),with few fieldsremaining fallow. ally facing into the wind, visually searching There are two rainy and two dry seasons(Cummings until a captureoccurs. Snyder and Snyder (1969) 1965). Rainfall during the study period totaled 14.76 and Voous and van Dijk (1973) described in cm in July, the end of the long rainy season,and 3.84 cm in August, the onsetof the shortdry season.Dai- ly minimum and maximum temperaturesranged from ' Present address: School of Natural Resources, 22-31øC.For a more detailed descriptionof the study University of Michigan, Ann Arbor, Michigan 48109 area, see Giglioli (1959), Osborneand Bourne(1977) USA. and Bourne and Osborne (1978).

84 TheAuk 100: 84-92. January 1983 January1983] SnailKite Foraging 85

A 26-ha study site of seven fields was selectedfor birds were sitting upon a fence post, vegetation, or behavioral observations of kites. Strip censuses of mudflat. It was not possible to determine reliably Snail Kites and Limpkins (Aramusguarauna) were whether perched birds were engaged in searching made one to five times weekly from 0700 to 0800. activities(i.e. still-hunting),and no attemptwas made Fieldwork was initiated near the time of sowing. Rice to separateperched birds into thosesearching or those had reachedfull height by the end of the study 6-8 not searchingfor snails. Thus, still-hunting kites vi- weeks before harvest. sually searchingfor prey while perchedwere includ- I marked 14 fence posts on the border between two ed in this category. (2) Maintenance activities in- fields and clearedthe surroundingarea of old snail cluded preening, scratching, ruffling, wing shellsso that I could samplethe size of snailsselected stretching,wing drying, and bill cleaning. (3) Flying by kites. As many as 14 kites were observedforaging was defined as movementsfor the purposeof chang- from these postsduring strip censusesbut four was ing location of perchesor hunting areas, as well as the average. Postswere selectedbecause of easy ac- flights to or from roostingareas. (4) Aggressionwas cessibility and use by foraging kites. Fence posts observed mainly in the form of the supplanting of ranged from 7 to 17 cm in diameter and 1 to 1.5 m one individual by another or by active chasing of high. Empty snail shellswere collectedfrom beneath anotherkite. Finally, (5) foragingincluded all activ- each post on 14 occasionswith periods of 24-48 h ities associatedwith the capture and ingestion of between collections.Aperture length was measured snails. A foraging bout was composedof searching to the nearest millimeter, and after one week three (flight in the pursuit and captureof snails, including size classeswere delineated: small (11-24 mm), me- both coursing and still-hunting), returning (flight dium (24-28 mm), and large (29-36 mm). from the point of captureto any perchwhere the prey I attempted to estimate the relative abundance of was consumed), and handling (the extraction of a the three size classes of snails in the two fields from snail from its shell and ingestion). The time between late July through early August. Becausewater tur- foraging bouts ("between") was calculated as the bidity was great, I could not detect snailsvisually. I amount of time that elapsedbetween ingestion of a collected snails by walking or crawling a transect snail and the initiation of the next searchingflight. through the field, tactilely searchingthe mud and For each capture clearly observed(79% of the cap- plant surfaceswith both hands. A total of 20 h was tures), I visually estimated the snail size class.From spent sampling snails in kite feeding areas. Thirty- experiencegained while measuringover 1,000 snail four of thesesnails (13 large, 9 medium, and 12 small) shells, I usually was able to estimate the aperture were collectedfor caloric analyses.After extraction length within 1 mm. A comparisonof my estimate from the shell, albumen glands were discarded(as of the distribution of snail size classes that kites fed kites do not generallyingest them; see Snyder and upon during periods of behavioral observationand Snyder 1969), and each snail was oven dried for 28 the distribution of snails collectedbeneath feeding h at 100øC,when dry weight becameconstant. Ten perchesrevealed no significantdifference (X 2 = 0.94, snailswere individually grotrod in a Wiley Mill, and df = 2, 0.50 < P < 0.70). This supportedmy ability caloric determinations were made on 40-mg samples to assigncaptured snails correctlyto the three size of each through use of a Phillipson Microbomb cal- classes. orimeter. Statistical analyses were conducted using para- I monitored Snail Kite activitiesby observingfocal metric procedureson the StatisticalAnalyses System individuals (Altmann 1974) chosen at random for 15- (S.A.S.) at Miami University and Wayne State Uni- min periods from 20 to 75 m away through spotting versity. Assumptionsof normality and homoscedas- scopeand binoculars.As there was no breeding col- tiscity of parametric modelswere tested by inspect- ony of kites in the vicinity and kites that were ob- ing scatterand residualplots, skewnessand kurtosis servedextracted and ate snailsin the study area, none coefficients,and comparingsample variances. Non- of the focal individuals was assumedto be breeding parametric statistics were used when assumptions during the time of the study. Activitieswere record- were violated. Becausepercentage data were not nor- ed directly into a tape recorderand later timed to the mally distributed and often outside the 30-70% in- nearestsecond with a stop watch. Three time periods terval, they were arcsine transformedbefore analy- were chosena priori to homogenizeair temperature ses.Differences were designatedas significantif P < trends within periods (morning: from sunriseat 0630 0.05. to 1100; midday: from 1100 to 1500; and late after- noon; from 1500 to 1830, just before sunset). A total RESULTS AND DISCUSSION of 60 focal samples was distributed almost evenly throughoutthe daylight hours (morning 19, midday Dail!/ activit!/.--Snail Kites usually arrived 21, late afternoon 20). I tried not to sample the same individually at foraging stations, presumably kite more than once a day by sampling during only from a communal roost to the north, between one time period a day. dawn (0630) and 1100, with occasionalarrivals Activities were categorizedas follows. (1) Perched or departures during the course of the day. 86 STEVENR. BEISSINGER [Auk, Vol. 100

MORNINGI MIDDAYJLATE AFTERNOON TIME OF DAY Fig. 1. Theactivity pattern of SnailKites in Guyanarice fields relative to timeof day.Different letters within activitytypes indicate significant differences between time periodmeans as testedby Duncan's Multiple Range test.

Twice, I observedgroup movements away from perching while visually searchingare nearly the study area during the day. It appeared, identical (King 1974). however, that most of the individuals foraging Despite fluctuationsduring the day in the in the area remained there for the entire day amountof time spentin maintenanceand flying until departingfor the roostagain 0.5-1 h be- activities,no statisticallysignificant trends were fore sunset.Although kites were not individ- noted (Fig. 1). Maintenancewas highestin the ually marked,certain fence posts were consis- morning, most likely in responseto frequent tently occupiedduring someperiods of the morning rainstorms.Morning valuesfor flying study, perhapsby the sameindividual. were lowest but would have been higher if The diurnal activity patternof Snail Kites is flights from the roost to the foraging grounds presentedin Fig. 1. Kitesspent the majority of could have been observed. thedaylight hours perching (62%), followed by Hunting behavior.--I observed74 foraging foraging(19%), maintenance (13%), and flying bouts by kites, and 82% of these were suc- (6%). Less than 0.1% of the photoperiodwas cessful. All food items were Pomacea snails. In spentin aggressivebehavior, which was ob- only 78% of the boutswas the snail completely servedonly twice during the 60 samplingpe- ingested, however, as kites dropped snails riods. Although aggressionassociated with three times during the processof handling them feeding territorieshas been occasionallyob- on a perch. On one of theseoccasions, a young- servedin Florida (Snyderand Snyder1970), I of-the-year kite (identified by char- saw no evidence of this in Guyana. Because acteristics;Haverschmidt 1968) had difficulty aggressivebehavior was a minorproportion of perchingwith the snailin its talons,a behavior Snail Kite activities, it was excluded from the oftenseen in youngkites in Florida(pers. obs.). remainderof the analysesin this paper. I do not know in the other two cases whether Activitypatterns changed with time of day: the snailswere rejectedpurposely (e.g. a cap- the percentageof photoperiodspent perching ture of an empty shell)or droppedaccidently. increasedsignificantly from morning to mid- In summary,Snail Kite predatoryefficiency was day and late afternoon,while foraging de- similiar to the mean value of 73% that Collopy creasedsignificantly during this time (Fig. 1). (MS) noted for other invertebrate-eatingbirds, As it was not possibleto determinereliably the mostsuccessful avian predators,but much whethera perchedkite was searchingvisually higherthan he foundfor fish-eaters(45%) and for prey (still-hunting)or simplyresting, all those hunting primarily mammals (27%) or perched kites were assumedto be resting, birds (8%). When snails were available, kites which resultedin lower foragingand increased had little trouble capturing them. restingvalues. This bias doesnot applyto en- Both hunting modeswere equally employed ergybudgets constructed later from these data, by kites:50.7% of the foragingbouts observed however•as the energeticcost of perchingand were coursinghunts and 49.3% were still- January1983] Snail Kite Foraging 87

TABLE1. Descriptivestatistics for the componentsof a Snail Kite foraging bout.

Number i'% + SD (s) of obser- of complete Foragingcomponents vations :• -+ SD (s) Range(s) bouts Coursing Searching 36 88 + 69 16-303 44 + 13 Returning 28 17 + 15 1-79 11 + 8 Handling 27 73 _+27 32-150 45 + 12 Still-hunting Searching 35 4 _+4 1-13 5 -+ 5 Returning 30 4 + 3 1-12 7 + 5 Handling 33 66 _+34 21-200 88 + 7 All hunts Searching 71 47 + 64 1-303 24 + 21 Returning 58 11 + 12 1-79 9 + 7 Handling 60 69 _+31 21-200 67 + 23 Between 43 146 -+ 172 10-710 --

hunts. In Florida, young kites have been ob- turning times decreasedsignificantly through- served to still-hunt, but adults rarely do so ex- out the day. Searchingtimes for coursinghunts cept in times of food stressduring regional showeda similarbut not significanttrend. De- drought,when kites are forcedto foragealong creasedsearch times probablywere relatedpri- woody lake marginsor along canals(Snyder marily to the proportion of still-hunts and and Snyder1969, Beissinger in prep.). Snyder coursing hunts that occurred in the three time and Snyder (1969)attribute the high incidence periods:58% of the coursinghunts occurred in of still-huntingby kitesin Guyanato the avail- the morning, while 46% of the still-hunts oc- abilityof moresuitable perches along rice fields curred in the late afternoon (from Table 2). The and canals.The hunting successof both meth- distributionsof still-huntsand coursinghunts ods was similar, as 81% of 36 coursing-hunt throughout the three time periods were sig- attemptsand 83% of 35 still-hunt attemptswere nificantly different (X2= 9.38, df = 2, P < successful. 0.01). Three explanationscould accountfor this Descriptive statistics associated with the shift in foraging mode: (1) as ambient air tem- componentsof a Snail Kite foragingbout are perature increasedduring the day, the costly summarized in Table 1. The mean searching exertion of long coursingflight was avoided time for coursinghunts was 22 timesmore than becausethe energeticrequirements for flight for still-hunts, but this was partly due to the increase proportionally with increasesin am- manner in which these parameters were mea- bient air temperature (Kendeigh et al. 1977); sured, concealing still-hunt searching time (2) by coursingin the morning,Snail Kites may within perching time and thus disproportion- have been investing time exploring the prey ally loweringsearching times for still-hunts.In base of the chosenforaging patch, a tactic re- comparisonto the other componentsof a for- ported for other avian predators (Smith and aging bout, prey handling was the most time Sweatmen 1974, Davies 1977); and/or (3) with consuming.For all hunts,handling constituted increasing water temperature, snails become an averageof nearly 68% of a bout. Even the more active (McClary 1964) and inspire near handling time for coursing hunts (45.2%), the water surfacemore often, increasingthe which may be the most meaningfulvalue be- encounterrate (and still-hunting success)for cause searchingtime is totally measured, in- perchingkites. Neither handling time nor time dicated that handling a snail is as time con- between captures(a measureof feeding rate) sumingas finding one. showedany significanttrends in relationto time The amount of time spent searchingand re- of day (Table 2). turning by kites was related to the time of day No significantdifferences were found in the (Table 2). When observationsof both hunting amount of time required for foraging-bout modes were pooled, mean searching and re- components in relation to snail size (Table 3). 88 STEVENR. gElSSINGER [Auk,Vol. 100

TABLE2. Variationin the componentsof a Snail Kite foragingbout during threetime periods.

Mean + SE (n) s by time of day Foragingcomponent Morning Midday Late afternoon Coursing Searching 95 + 16 (21) 94 + 27 (8) 60 + 14 (7) 0.628 Returning 17 + 2 (20) 24 + 14 (5) 8 + 4 (3) 0.245 Handling 69 + 5 (19) 70 + 11 (5) 101 + 25 (3) 0.462 Still-hunting Searching 4 + 2 (9) 6 + 1 (11) 2 + 1 (15) 0.107 Returning 4 +_1 (8) 6 + 1 (9) 4 + 1 (13) 0.282 Handling 57 + 4 (10) 90 + 19 (9) 58 + 5 (14) 0.708 All hunts Searching 68 + 14 (30) 43 +_15 (19) 21 + 7 (22) 0.020 Returning 13 +_2 (28) 12 + 5 (14) 4 + 1 (16) 0.003 Handling 65 +_4 (29) 83 +_13 (14) 65 + 7 (17) 0.734 Between 122 + 66 (21) 160 + 55 (10) 175 +_66 (12) 0.448 Significancelevels based on Kpaskal-Wallistest (Chi-square approximation).

Searchingand returning times decreasedrela- time (and presumablyenergy) required to find tive to snail size, but this was not statistically a snail, return to a perch,and extractand ingest significant.The trend of decreasedsearch times the snail did not differ significantlyfor snails for larger snails seems to run counter to the of different sizes,the profitability of a snail can potential encounterrate, which would be low- be defined merely by its total energeticvalue er for the less common, larger snails (Fig. 2). (Krebs 1978, Schoener 1979). The possibility that kites foragedby searching An interactionbetween time of day and snail for larger snailsbut took smallersnails after a size might have been responsiblefor the out- time without capture in order at least to stay come of the above results.Despite not meeting even with energy expendituresmight account the assumptionof normality, a two-way anal- for the shortersearching times observed for less ysis of variance (ANOVA) model was con- common,larger snails. For all hunts, handling structed to test for interaction. One-way AN- time was very consistent,with a mean for the OVA models were examined simultaneously three size classes of 67 s. Because the cost in with the Kruskal-Wallis tests previously re-

TABLE3. Variation in the componentsof a Snail Kite foraging bout in relation to the size classof snail captured.

Mean +_SE (n) s for snail size Foragingcomponent Small Medium Large pa Coursing Searching 78 + 15 (8) 54 + 9 (10) 53 + 21 (2) 0.402 Returning 15 + 3 (8) 17 + 3 (10) 7 +_2 (2) 0.336 Handling 60 +_6 (8) 65 + 11 (10) 93 +_6 (2) 0.168 Still-hunting Searching 6 + 2 (8) 3 + I (9) 3 +_I (6) 0.245 Returning 6 + I (8) 4 +_I (9) 4 + I (6) 0.507 Handling 72 + 16 (10) 71 + 9 (10) 58 + 4 (6) 0.440 All hunts Searching 42 +_12 (16) 30 + 8 (19) 16 + 9 (8) 0.340 Returning 10 +_2 (16) 11 + 2 (19) 4 + I (8) 0.200 Handling 66 + 9 (18) 68 + 7 (20) 67 + 6 (8) 0.604 Between 131 + 51 (10) 145 + 52 (12) 259 + 119 (6) 0.444 a Significancelevels based on Kmskal-Wallistest (Chi-square approximation). January1983] SnailKite Foraging 89

and Hazelwood, 1977); and (4) the density of vegetation covering the water surface. I sam- pled snail availability in a manner that as- sumed that factors affecting availability act similarly on all sizes of snails. If not, then the size distribution of snails that was collected from the field might not be the same as that availableto kites. For instance,large snails may be less vulnerable to predation, becausethey may haverespiratory rates that are lessaffected by changingwater temperatures,causing them to surfaceless often, or they may frequentdense vegetation more often than small snails.

SMALL MEDIUM LARGE The choiceof snail size couldbe affectedby one other factor--snail parasite loads. Gross- SNAIL SIZE man and Hamlet (1964) state that kites are in- Fig. 2. The sizesof Pomaceadolioides potentially fected by "lung flukes" carried by Pomacea available (striped bars) to Snail Kites and sizestaken snailsbut offer no direct evidenceof parasitism (dark bars). or references.At least eight trematode species are described from Pomacea snails, and birds are definitive hosts in all three cases where life ported (Tables 2, 3), and in all casesparametric historiesare known (Nasir and Rodriguez1969; and nonparametrictests yielded similar values. Nasir et al. 1969a, b; Nasir and Silva 1972; Gas- From the two-way ANOVA model, there was con 1975;Hanning and Leedom 1978). Gigan- no evidence that a time of day and snail size tism, increased shell thickness, and shell ab- interaction occurred for searching (F = 0.70, normalities in molluscs sometimes result from P = 0.596) or returning times (F = 0.07, P = parasitism (Cheng 1971, Hanning 1978) and 0.991). might offer kites a mechanismto rejecthighly Snail size selection.--I collected 959 Pomacea parasitizedindividuals. Becauseparasitic loads shellsthat had been eatenby kites. Over 99% do increasewith snail size in Pomaceapaludosa of the snails were Pomaceadolioides; P. glauca (Hanning 1978),kites may not havetaken large composedthe remainderof the sample.Snyder snails more often in order to avoid higher in- and Snyder (1969) also noted Snail Kites in fection rates. Guyana feedingheavily on P. dolioidesnear the Coursinghunts require far moreflying, a very water surfaceand rarely upon P. glauca, a less energeticallyexpensive activity (Table 4), than abundant and more aquatic bottom dweller. do still-hunts. Course-hunting Snail Kites From transect searches, I collected 155 P. do- would be expectedto chooselarger snails in lioidesto serve as a sample of the size distri- order to gain a greaterenergetic payoff to offset bution of snails potentially available to kites. increased costs (Schoener 1969, 1979), espe- Figure 2 comparesthe sizes of snails available cially as searchingtime is unrelated to snail and those chosenby kites. There was a signif- size. No differences in the distributions of snail icant difference between the frequency of size size classesselected by coursingor still-hunt- classes available to kites and those selected ing kites occurredwhen all three size classes (X2 = 7.632, df = 2, P = 0.022). It appearsthat were considered (X2 = 1.85, df = 2, 0.60 < P < Snail Kites avoided the more abundant small 0.70) or when medium and large size classes snails, preferred the medium sized snails, and were combined and tested against small (X2 > took the large snails in approximatelyequal 0.750, df = 1, P > 0.750). frequencywith those that were available. Effectsof ricegrowth.--Figure 3 illustratesthe Becausesnails are captured by kites on the declining use of rice fields by Snail Kites dur- wing, snail availability is a function of: (1) snail ing the study period. When rice fields were density; (2) snail depth below the water sur- drained, kite numbers were highest. As irri- face; (3) the effects of water temperature and gation and rice growth commenced,kite use dissolvedoxygen on Pomaceainspiratory and significantly declined, until no birds were ob- surfacing behavior (McClary 1964, Freiburg served in mid-August. A highly significant 90 STEVENR. BEISSINGER [Auk,Vol. 100

TABLE4. A simplifiedenergy budget for nonbreed- IRRIGATION ing SnailKites. Equation10 from Koplin et al. (1980) RICEMATURATION was used to calculateenergy cost,assuming an av- eragedaily air temperatureof 27øC(pers. obs.) and a mean body weight of 367.6 g, the mean of seven observations from Haverschmidt (1968, 1970), and specimensfrom the Museum of Zoology, Univer- 2 1 sity of Michigan.

Proportion Energy cost z Activity of 24-h day (kcal) Nonflight 0.945 59.1 Flight 0.055 26.6 WEEK 1 2 3 4 I 2 3 JULY AUGUST Total 1.000 85.7 Fig. 3. Weekly mean densityof Snail Kites (striped bars) and Limpkins (dark bars) determined by strip censusesin Guyana rice fields. The number of significantly(X '• = 5.93, df = 1, P • 0.025)from weekly counts appearsabove each bar. Upper hori- 0% (n = 9) in the first 2 weeks to 47% (n = 15) zontal bars indicate the duration of sowing (S), rice in the last 3 weeks. These results are opposed growth and maturation, drainage (D), and irrigation. to predictions from models by Norberg (1977) that predatorsshould employ lessenergy-con- suming hunting methods as prey density de- (P = 0.002, r '•= 0.747, n = 20) linear correla- creases. tion betweendaily countsand week was found Energybudget.--The daily foodintake of Snail for Snail Kitesbut not for Limpkins (P = 0.620, Kites was calculatedby proportionatelyreduc- r '•= 0.125, n = 18). As rice grows, foraging ing the distribution of the three size classesfor becomes more difficult for kites because: (1) the 68 snailsconsumed during the 15 h of time- the areaof water surfacethat a kite canvisually activity observationsto that which would have scan while hunting decreasesand (2) to make occurred in a 12-h Guyana photoperiod. Ca- a capture, a kite's descent to the water must be loric valuesper gramdry weight for P. dolioides unhindered by vegetation. For effective for- were relatively constant(;? + SD = 4.04 + 0.09, aging, Snail Kitesneed sizeablepatches of open n = 10). Using mean dry weightsfor eachsnail water with floatingbut not tall emergentvege- size classand assumingan assimilationeffi- tation. Limpkins are able to searchtactilely for ciency of 0.9, a value chosenbecause snail tis- Pomaceasnails while wading (Snyderand Sny- sue is almostentirely digestableand raptor as- der 1969)and can foragesuccessfully in heavily similation efficiencies peak near this value vegetatedareas. In the studyarea, they utilized (Sarkerand Naulleau1981), ! calculatedthe daily fallow fieldsuntil rice reachedapproximately a energy intake by Snail Kites to be 104.2 kcal. 0.5-m height. An averagedaily energyexpenditure for Snail Assuming relatively stable snail popula- Kites was calculatedin the following manner. tions, snail availability should decreaseas rice As kites spent 12 h a day roosting,means from grows, and the encounterrate of still-hunting the 60 time-activitysamples (Fig. 1) were pooled kites should decrease.In response,Snail Kites to determinethe mean percentageof eachdaily should course-hunt more often to increase en- activity for a 24-h period. BecauseSnail Kites counter rates by searching a greater area. To spentless than 20% of the photoperiodin flight test this, ! examined the proportion of still- activities,equation 10 from Koplin et al. (1980) hunts that occurredduring the first 2 weeksof was used to compute the daily energy budget behavioral observations, when rice was absent (Table 4). Daily energy expenditure for Snail or low, and the last 3 weeks, when rice was Kites was estimated to be 85.7 kcal (Table 4). highly emergent. Becausehunting mode was My calculationsindicate that kites easily uti- related to time of day, ! used only the 24 hunts lized ricefields to meettheir daily energyneeds, that occurred in the late afternoon, when still- realizing a daily net energy gain of 18.5 kcal. hunting was most frequent for this analysis. The high proportion of snailscaptured by en- The proportion of coursing hunts increased ergeticallyinexpensive still-hunting probably January1983] SnailKite Foraging 91 accountedfor this large net energy gain. Dur- from the still-hunting perch. Factorsthat affect ing the study period, I saw no evidence that snail-encounter rates can cause a shift in hunt- kites were ever stressedto find food or that any ing mode: the vegetative structureof a patch were malnourished. On the contrary, the small may be a prime determinant, as still-hunting proportionof time spent flying by kites in rice increasedwith rice growth during the 2 months fields (Table 4) may indicate that this was a of study, and environmental factors (e.g. air very profitable patch to exploit. and water temperature, dissolved oxygen, and Consumerchoice by a specialist.--SnailKites wind speed) that affect both snail-encounter are a classicexample of a "specialist." Preying rates and kite metabolic costs probably ac- almostsolely on one genusof freshwatersnails counted for the diurnal shift by kites in the (oftenone speciesin someportions of its range), mode of hunting that was observed. kites need not distinguish among a number of BecauseSnail Kites selectively avoid small different items to be included in the diet. Un- snails,a kite might be expectedto prefer large der most conditions a kite may need only to snails.The diet shouldvary from patchto patch, chooseamong food items by size as it relates however, depending upon the distribution of to the energeticbenefit of caloricintake and the size classesavailable and the energeticcosts of energeticcost of searchingand handling times. capture, which are both functions of the en- In this study, no increasedcost was associ- counter rate and hence of patch choice. Patches ated with the pursuit, capture, handling, and do vary greatly in snail density and size dis- ingestion of prey in relation to prey size. No tribution (Bourne and Berlin 1982). Thus, the additional energeticbenefit in terms of calories most important decisiona foraging Snail Kite per capture was accrued during costly coursing must make is what patch to searchin and how hunts as compared to still-hunts. Snail Kites long to searchbefore abandoning one patchfor avoided capture of smaller snails, selectively another. preyedon medium-sizedones, and did not take ACKNOWLEDGMENTS large snailsmore frequently than they were en- countered.In addition, patch use and hunting I am especiallyindebted to David R. Osbornefor mode were affectedby rice growth. Rice fields advice in the field, companionship, and the oppor- proved to be an easy place for kites to meet tunity to conductthis study. The Guyanesegovern- their energeticneeds. ment and the Rice Research Station at Burma made MacArthur (1972) partitioned the processof housing and lab facilities available, and T. Runyan food gathering into four phases: (1) where to provided housingduring much of the data analyses. J. Diana, L. Flath, and T. Grimshaw kindly gave me search; (2) how to search and what to look for; bomb calorimetryinstruction. StevenSeilkop assist- (3) whether to pursue or ignore a food item; ed with computerprogramming. Thanks to G. Be- and (4) the capture attempt and ingestion. Kite lovsky,G. Bourne,G. Fowler,T. Grimshaw, S. Lima, morphologyis so specialized(Snyder and Sny- E. Werner,two anonymousreviewers, and especially der 1969) that, in order to maximize snail cap- M. Collopy and N. Snyder for reviewing various ture and ingestionproficiency (4), the decision stagesof this manuscript.Work was supportedby of what foods to searchfor (2) has already been the Alumni Foundationand Departmentof Zoology determined. A Snail Kite need only decide on at Miami University and the Schoolof Natural Re- where to search (1), how to search (2), and (3) sources,University of Michigan. Finally, I would like whether or not to include an item in its diet to thank the peopleof Guyanafor their gracioushos- on the basis of size. These decisions may be pitality. based on the encounter rate within a given LITERATURE CITED patch. In the choiceof hunting modes, kites might ALTMANN,J. 1974. Observationalstudy of behav- be expected to still-hunt exclusivelyif they ior: samplingmethods. Behaviour 49: 227-267. BOURNE,G. R., & J. A. BERLIN. 1982. Predicting could find enough food. When insufficient Pomaceadolioides (Reeve) (Prosobranchia: Am- numbers of snails per unit time are available pullariidae)weights from linear measurements from a perch, kites should shift to coursing. of their shells. Veliger 24: 367-370. Thus, the effectivenessof still-hunting may be --, & D. R. OSBORNE. 1978. Black-bellied primarily controlledby the snail-encounterrate WhistlingDuck utilizationof a riceculture hab- within the patch as a function of the distance itat. Interciencia 3: 152-159. 92 STEVENR. BEISSINGER [Auk, Vol. 100

CI-IENG,T. C. 1971. Enhanced growth as a mani- MCCLARY,A. 1964. Surface inspiration and cil- festation of parasitism and shell deposition in iary feeding in Pomaceapaludosa (Prosobran- parasitized mollusks. Pp. 103-138 in Aspectsof chia: Mesogastropoda:Ampullariidae). Malaco- the biologyof symbiosis(T. C. Cheng, Ed.). Bal- logia 2: 87-104. timore, Maryland, University Park Press. NASIR, P., & L. RODRIGUEZ. 1969. Studies on the CUMMINGS,L. P. 1965. Geography of Guyana. freshwater larval trematodes.XX. The life cycle London, Collins. of Stephanoproraparadenticulata n. sp. (Trema- DAVIES, N. B. 1977. Prey selection and the search toda, Echinostomatidae). Zool. Anz. 182: 230- strategy of the Spotted Flycatcher (Musciapa 244. striata),a field studyon optimalforaging. Anita. , & J. S•LVA. 1972. Freshwater larval trema- Behav. 25: 1016-1033. todes. XXIX. Life cycleof Guaicaipuriaparapseu- FREIBURG, M. W., & D. H. HAZELWOOD. 1977. doconciliasp. n. Proc. Helminthol. Soc. Wash- Oxygen consumptionof two amphibious snails: ington 39: 142-146. Pomaceapaludosa and Marisa cornuarietis(Proso- , M. T. DIAZ, & M. G. DIONE. 1969a. Fresh- branchia:Ampullariidae). Malacologia16: 541- water larval trematodes. XXVI. Life cycle of 548. Guaicaipuriapseudoconcilia (Nasir, Diaz and Le- GASCON,A. 1975. DoS ciliadosdel genero Parasi- mus de Guevara,1969) comb. n., gen. n., subfaro. cuophoraparasitos de Pomaceacanaliculata (La- n. Proc. Helminthol. Soc. Washington36: 190- marck).Rev. Biol. Urg. 3: 111-126. 193. G•GL•OLLE.G. 1959. histories and field in- --, & L. J. HAMANA S. 1969b. Studies vestigations, 1951-1957. Georgetown, Guyana, on'freshwater larval trematodes. XXV. Two new British Guyana Rice Development Co. Ltd. speciesof echinostomecercariae. Proc. Helmin- GROSSMAN,M. L., & J. HAMLET. 1964. Birds of prey thol. Soc. Washington 36: 175-177. of the world. New York, Potter Inc. NORBERG,g.g. 1977. An ecologicaltheory on for- HANNING, G.W. 1978. Aspectsof reproduction in aging time and energeticsand choiceof optimal Pomaceapaludosa (Mesogastropoda: Pilidae). food-searchingmethod. J. Anita. Ecol. 46: 511- Unpublished M.S. thesis. Tallahassee, Florida, 529. Florida State Univ. OSBOI•E,D. R., & G. R. BOUI•E. 1977. Breeding , & W. S. Leedom. 1978. Schistosomedermati- behavior and food habits of the Wattled Jacana. tis from Pomaceapaludosa (Say) (Prosobranchia: Condor 79: 98-105. Pilidae). Nautilus 92: 105-106. SARKER,S. W., &: G. NAULLEAU. 1981. •tude HAVERSCHMIDT,F. 1962. Notes on the feeding quantitativedes fienteset de l'assimilationchez habits and food of some hawks in Surinam. les Rapaces. L'Oiseau 51: 161-165. Condor 64: 154-158. SCHOENER,T.W. 1969. Models for optimal size for 1968. Birdsof Surinam. Edinburgh, Oliver solitarypredators. Amer. Natur. 103:277-313. and Boyd. 1979. Generality of the size-distance rela- 1970. Notes on the Snail Kite in Surinam. tion in modelsof optimal feeding.Amer. Natur. Auk 87: 580-584. 114: 902-914. KENDEIGH, S.C., V. R. DOL'NIK, &: V. M. GAVRILOV. SMITH, J. N.M., & H. P. A. SWEATMAN. 1974. Food 1977. Avian energetics.Pp. 127-204 in Graniv- searchingbehaviour of titmice in patchy envi- orous birds in ecosystems(J. Pinowski and S. ronments. Ecology55: 1216-1232. C. Kendeigh,Eds.). New York, CambridgePress. SNYDER, N. F. R., &: n. A. SNYDER. 1969. A com- KING, J. R. 1974. Seasonal allocation of time and parative study of mollusc predation by Limp- energy resourcesin birds. Pp. 4-85 in Avian en- kins, EvergladeKites and Boat-tailed Grackles. ergetics (R. A. Paynter, Ed.). Publ. Nuttal Or- Living Bird 8: 177-223. nithol. Club, No. 15. , & --. 1970. Feeding territories in the KOPLIN, J. R., M. W. COLLOPY, A. R. SAMMANN, & EvergladeKite. Condor 72: 492-493. H. LEVENSON.1980. Energeticsof two winter- SYKES,P. W., JR., & H. W. KALE, II. 1974. Ever- ing raptors. Auk 97: 795-806. glade Kite feed on nonsnail prey. Auk 91: 818- KRESS,J.R. 1978. Optimal foraging:decision rules 820. for predators.Pp. 23-63 in Behavioralecology (J. Voous, K. H., & T. VAN DIJK. 1973. How do Snail R. Krebs and N. B. Davies, Eds.). Sunderland, Kites extract snails from their shells? Ardea 61: Massachusetts,Sinauer Associate, Inc. 179-185. MACARTHUR,R. H. 1972. Geographicalecology. New York, Harper and Row. MOLLUSK PREDATION BY SNAIL KITES IN

NOEL F. R. SNYDERl"• AND HERBERT W. KALE IP 1SebringField Station, Patuxent Wildlife Research Center, U.S. Fishand WildlifeService, Box 911, Moorehaven,Florida 3347• USA, and :•FloridaAudubon Society, 1101 Audubon Way, Maitland, Florida 32751 USA

ABSTRACT--SnailKites (Rostrhamus sociabilis sociabilis) observed along the lower Rio Mag- dalenain northernColombia fed heavily on Pomaceachemnitzi but alsotook Marisacornu- arietis,a smallerspecies. The kitescaptured Pomacea preferentially over Marisa and had difficultiesextracting Marisa from its shell. They also failed to extractsome large specimens ofP. chemnitzi.Failures with both species were apparently related to problems in removing opercula.These observations indicate that R. sociabiliscan no longerbe considereda strict specialiston Pomacea.Received 21 August1979, resubmitted 8 April 1982,accepted 15 October 1982.

MOLLUSKSare a major dietary componentfor RESULTS a surprisingly diverse assemblageof birds, in- cluding various , shorebirds,rails, , Snail Kites characteristicallybring their prey to conspicuousfeeding perches for extraction and raptors (Snyder 1967). Of particular inter- procedures,and the shells accumulatingun- est, becauseof its endangeredstatus in the derneath give a record of the foods eaten. We United States,is the Snail, or Everglade,Kite (Rostrhamussociabilis), a speciesrenowned for soon learned, however, that someshell piles in the study area yielded biased representations its dependenceon fresh-watersnails of the ge- of kite diet because of differential losses of shells nus Pomacea.Snyder and Snyder (1969) dis- cussedsome of the peculiar anatomical and be- after deposition. Shell piles under perches standing in water retained their Marisa shells havioral specializations of the Snail Kite that well but lost most of their Pomaceashells, ap- enable it to feed on Pomaceapaludosa in Flor- parently becausethe former are heavily calci- ida. Here, we present information on interac- tions of the specieswith snailsof northern Co- fied and tend to sink, while the latter readily float and can drift away on water currents. lombia, as observed28 April to 5 May 1978, the start of the rainy season. Thus, for example,we found only 28 Pomacea shells among 699 Marisa shells at one perch Our study area was a backwater of the Rio that obviouslyhad been inundated recently. Magdalenanext to the Barranquillaairport. Two In contrast,we found a preponderanceof Po- large snailsof the family Pilidae were found in abundance here, Pomacea chemnitzi and Marisa maceaover Marisa in shellpiles abovethe water line. The overall Pomacea to Marisa ratio at four cornuarietis. Pomacea chemnitzi, with its shal- such percheswas 3.0, and it varied from 1.8 to low spire, closely resembles P. paludosa in 7.2 among the perches (Table 1). The four shape,but, at leastin this locality,it commonly perches were located within 100 m of each oth- reachesa size considerablylarger than typical er in an area of mixed brushy swamp and individuals of the Florida species.Marisa cor- flooded pasture immediately adjacentto an ac- nuarietis is coiled in a plane and looks like a tive kite nestingcolony. Using the percheswere giant ramshornsnail, but, like Pomacea,it pos- sessesa hard operculumwith which it can close a minimum of three brown-plumagedkites (fe- off the entrance of the shell. males and/or young males). Under all four of the above percheswe were surprised to find evidence that the kites were failing to extractsome Pomaceaand Marisa, as '-'Present address: California Field Station, Patux- we found whole uneaten individuals of both ent Wildlife Research Center, U.S. Fish and Wildlife species, some still alive and others in various Service, 87 North Chestnut Street, Ventura, Califor- stagesof decay. In addition, an inspection of nia 93001 USA. the shellsrevealed that many of the snailswere

93 The Auk 100:93-97. January1983 94 SNYDERAND KALE [Auk, Vol. 100

TABLE1. Shellsfound under active perchesof Snail Kites in Colombia?

Perch number Total i 2 3 4 shells

Pomacea chemnitzi Empty shells 35 35 28 28 126 Shells with viscera 5 15 2 5 27 Whole snails 6 1 6 1 14 Total shells 46 51 36 34 167 Marisa cornuarietis Empty shells 5 1 0 5 11 Shells with viscera 7 7 1 8 23 Whole snails 5 6 4 6 21 Total shells 17 14 5 19 55

Pomacea/Marisa ratios 2.7 3.6 7.2 1.8

All shell piles above the water line. only partially eaten (Table 1). The failure rates bee in February 1979, only four still contained for the two snail specieswere highly signifi- small fragments of snail viscera. On the other cantlydifferent (X 2 = 25.47,df = 1, P < 0.001). hand, recently fledged Florida Snail Kites do Of 167 Pomaceashells only 14 (8%) still con- fail occasionallyin extractionsof P. paludosa. tained whole, uneaten snails, compared to 21 For example, on 21 May 1979we collectedshells of 55 Marisa shells (38%). from a screen installed below a perch being In the cases of uneaten P. chemnitzi, failure used almost exclusively by banded kites less appeared to be related to snail size (Fig. 1A), than 2 monthsbeyond fiedgingage. Of the 114 as uneaten snails (averaging57.2 mm in great- shells collected, 9 (7.9%) contained whole, est length) were significantlydifferent in length unextracted snails. The uneaten snails did not from eaten snails (averaging 52.8 mm), as differ significantlyin sizefrom eatensnails (t.• = shown by a Mann-Whitney U-test (t., = 3.182, 1.524, 0.10 < P < 0.20; Mann-Whitney U-test), P < 0.'005).Opercula of the uneaten specimens and, in the casesin which we directly observed were scratched with the usual bill marks one young kites dropping whole snailsat perches, finds on kite-extracted Pomacea.Apparently, the causewas quite clearly nothing more than the kites had tried to extract the uneaten Po- clumsiness. The birds had obvious difficulties maceabut had failed to grasp the opercula or holding the snails properly in their feet and had been unable to pull the opercula free from sometimeslost control of the shells in attempt- the snails. In watching severalunsuccessful at- ing to position them. tempts, we saw no signs that the birds might Whether the failures of Colombian Snail Kites have been deliberately rejecting individual in extractionof P. chemnitzimight similarly be snails that were noxious to them. The kites related to the age of the kites is conjectural.The simply worked away at thesesnails with their kites we observed at the perches were all bills in the usual fashion, appeared to lose in- brown-plumaged and none appeared to be terest after a few minutes, and eventually al- mated--characteristicsconsistent with youth. lowed the snails to fall from the perches. All were fully independent,however, and none The failure of the kites to extract all P. chem- was clumsyin its snail-extractionefforts. These nitzi bears further comment. With P. paludosa factors, together with the size-dependenceof in Florida we have never, in many hundreds their failures with P. chemnitzi,suggest that of direct observations, seen an adult kite fail, their difficulties may have resulted from more and only very rarely have we found evidence than just inexperience. In any event, we be- of less than complete extractions of softparts. lieve it would be unwise to assume that the For example, of 529 P. paludosashells collected few birds we observed were representative of from beneath a perch being used exclusively the kite population as a whole. by adult and subadult kites on Lake Okeecho- Failures with M. cornuarietis, like failures January1983] SnailKites in Colombia 95

A

A 60,

o 40,

20, CMA

Z 60, B

40'

20'

3o •$ •o •$ so $$ 6o

Millimeters Fig.1. Sizedistributions of Pomaceachemnitzi and Marisa cornuarietisshells found under active perches of SnailKites: A, greatestlength of eaten(clear) and uneaten (shaded)Pomacea chemnitzi; B, greatestdi- Fig. 2. Diagrammaticrepresentation of position ameterof eaten(clear) and uneaten(shaded) Marisa of columellar muscle attachment (CMA) and oper- cornuarietis;C, greatest diameter of all Pomacea culum (O) of: A, Marisa cornuarietis;B, Pomacea chemnitzi. chemnitzi. with P. chemnitzi,appeared to be causedby an which must be severed to allow a complete ex- inability of the kitesto removeopercula, as the tractionof the softparts,lies much closer to the operculaof whole,uneaten snails under perch- entrancein Pomacea(Fig. 2B). The great ma- es showed clear scratchmarks. The high fre- jority of P. chemnitzifrom which kites were quencyof failure may have been due to this able to remove operculawere fully extracted snail'sability to pull far back into its shell: in (126 of 153 cases).By comparison,only 11 of Pomaceathe operculumusually forms a tight 34 Marisa from which the opercula had been fit just insidethe aperture;in Marisathe oper- removedwere fully extracted,a highly signif- culum can be withdrawn for a considerable icantlydifference (X 2 = 33.0,df = 1, P < 0.001). distance inside where it is difficult to grasp. Generally,the softpartsof Marisa snailswere Failure with Marisa was not a function of snail broken near the columellar muscle attachment, size, as unextracted snails did not differ sig- and all of the viscera remained in the shells. nificantlyin maximumdiameter from extracted Breakageprobably resultedfrom the kites' snails (t., = 0.468, P > 0.5; Mann-Whitney pullingon the softpartsfrom the shellentrance U-test). after failing to severthe columellarmuscle. Even when kites were able to remove the Direct observations of the snail-extraction operculaof Marisa, they commonlyfailed to behavior of the minimum of three kites using get more than the foot of the snail out of the the activeperches listed in Table1 confirmed shell,probably because the columellarmuscle a highersuccess rate with Pomaceathan with attachment of this snail is about as far inside Marisa, although the successrates were not the entranceas the long hookedbill of the birds significantlydifferent (X 2 = 1.13,df = 1,0.25 • can reach (Fig. 2A). This muscleattachment, P < 0.50) for the two snails, probably due to 96 S•¾DERAfar) K,•LE [Auk, Vol. 100

small sample size. Of six casesof the kites at- The size distributions of captured Pornacea tempting to extractMarisa, only two were suc- and Marisa snails in Fig. 1A and B are pre- cessfulto the point of the birds' getting some sented in terms of maximum linear dimen- meat out of the shell. With Pornacea the success sions: total length from top of spire to base of rate was eight of 11 attempts. aperture for Pornaceaand greatestdiameter for Not only were the kites failing more fre- Man'sa. In these terms the sizes of some Marisa quently with Marisa than with Pornacea,they shells fall below the sizes of all Pornacea. At were also receiving much lessreward from this first sight it might seempuzzling that the kites snail when successful.The wet weight of a capturedthe smallestMarisa, becausethey were Marisa's softpartsgenerally does not exceed 2 apparently rejecting Pomaceaof similar linear g, whereas a typical P. chernnitziyields 15-20 extent (but much greater mass). Under field g of meat. When the total and partial failure conditions, however, the maximum linear di- rates in extractions from Table 1 are included mensions of a Pornacea will often be hidden in calculations,the yield in soft tissues from from the view of a kite flying overhead, be- an averageP. chernnitziruns about 15 times as cause snails near the water surface are not al- great as from an average M. cornuarietis. ways oriented the same way with respect to From the above comparison one might ex- gravity. In fact, a Pornaceaviewed by looking pect to find the kites selectivelyforaging for down on its spire looksvery much like a Mar- Pornacea in preference to Marisa. That this isa, and in this orientation its greatestvisible might be true was apparent from an inspection linear dimension is a maximum diameter, not of foragingareas near the activefeeding perch- a total length. If instead of plotting Pornacea es. In walking several transectsthrough the accordingto total length, we plot these snails marsh we were able to tally between 10 and 20 accordingto greatestdiameter (Fig. 1C), we find times as many adult Marisa as Pornaceanear that the Pornaceacurve now overlapsthe Mar- the water surface (where they would be vul- isa curve completely. Thus, the potential ease nerable to the kites). Yet, we found a Marisa of discrimination between the smallest Porna- to Pornacearatio of about 1:3 at the four feeding cea captured and the smallestMarisa captured perches and directly observed a 6:11 Marisa: disappears. Pornaceacapture ratio near the perches. DISCUSSION One may questionwhy the kites were taking Marisa at all in view of its much inferior re- The extractionsof P. chernnitziby Snail Kites ward potential. Possibly,the capturesof Mar- followed the pattern describedby Snyder and isa representednothing more than mistakes in Snyder(1969) and supportedby Voousand Van discrimination between the two snails. From Dijk (1973). They bore no resemblanceto the the 3-6-m heights at which the kites generally extractionprocedures described by Lang (1924) forage, distinguishing between Pornaceaand and Murphy (1955). We have now observed Marisa may present problems, especiallywhen kites eating three different speciesof Pornacea shellsare coveredwith algae. Once a kite has (including P. dolioides,on which Lang's de- gone to the trouble of capturing the "wrong scription was based), and all have been han- species," the reward, though relatively low, dled in the sameway: removal of the opercu- may still be high enough to justify an extrac- lum with the bill, cutting of the columellar tion attempt rather than a rejectionand a con- muscle attachment with strokes of the upper tinued searchfor Pornacea.Also, assumingthe bill, and pulling of the freed softpartsout of observed 10-20-fold superiority in abundance the shell with the bill. Unlike Lang and Mur- of Marisa over Pomaceamight apply inversely phy, we have not seen kites waiting for vol- to the time and energy investment needed to untary extensionof snailsfrom their shells,we locate the two snails, the net reward superi- have no evidencethat they ever ingestopercula ority of a typical Pomaceaover a typical Marisa in the wild, and we have seennothing to sug- may be far less than that calculatedabove for gestthat they pierce the snailsin a "nerve plex- snailsalready captured. Unfortunately, we did us," causingthe snails to releasetheir holds on not record information on the average time it their shells.We believe that the Lang-Murphy takes to capture the two snail species, so we description,while colorful, is incorrect.Unfor- could not evaluate this question in any direct tunately, this description has become en- fashion. trenched in the popular literature on kites and January1983] SnailKites in Colombia 97

may never be evicted,as there are now so many speciesto have complexcompetitive and pred- referencesreiterating it. atory relationshipswith other snail species(see On several occasions we observed Snail Kites Demian and Lutfy 1965,Ruiz-Tib•n et al. 1969), discardingthe yolk glands of female P. chem- the establishmentof Marisa may prove to be a nitzi, and we found discarded yolk glands of mixed blessingfor the Snail Kites of Florida. this speciesunder kite feeding perches.Very likely the yolk and eggsof this snailare highly ACKNOWLEDGMENTS distasteful,as are the yolk and eggsof P. do- Our studies in Colombia were made possibleby lioides and P. paludosa, which are also com- support from the U.S. Fish and Wildlife Service and monly rejectedby the kites and other predators the Dade County, Florida Aviation Department. The (see Snyder and Snyder 1969, 1971). Like the primary purposeof the trip was a study of the impact orange eggs of P. dolioidesand the pinkish- of the Barranquillaairport on the nearby Snail Kite white eggsof P. paludosa,the yellowish-white populations. We thank Bonnie Ann Frank, David eggsof P. chemnitziare conspicuousand pre- Dahmen, Edgar de Castro, and German Solano for assistance rendered in field studies in Colombia and sumablyaposernatic. All three specieslay their Steven Beissinger and Gary Falxa for assistancein eggsout of the water on emergentvegetation. kite studies in Florida. Alan Solem of the Chicago Marisa cornuarietis,in contrast,has aquatic Field Museum identified the Pomacea snails. eggs, which are not highly conspicuous.We did not note discardedMarisa yolk glandsun- LITERATURE CITED der kite perches and have no direct evidence BERG,C. O. 1961. Biology of snail-killing Scio- that the yolk and eggsof this speciesmight be myzidae (Diptera) of North America and Eu- distasteful. rope. Internationaler Kongressffir Entomologie Although with most fresh-watersnails there Wien 1960, Bd. I: 197-202. is little evidencethat operculamight serve as DEMIAN, E.S., & R. G. LUTFY. 1965. Predatory an effectiveadaptation to thwart predators(see activity of Marisa cornuarietisagainst Bulinus Snyder 1967), the operculaof both M. cornu- (Bulinus)truncatus, the transmitterof urinary arietisand P. chemnitziappear to have value in schistosomiasis.Ann. Trop. Med. Parasit. 59: reducingkite predationin Colombia.Some kite 331-336. perchesare over water, and snailsthat are not EDMONDSON,W. T. (ED.). 1959. Fresh-water biol- successfullyextracted by kites get a second ogy, 2nd Ed. New York, John Wiley and Sons, Inc. chanceat life if they fall back into the water on LANG, H. 1924. Ampullarius and Rostrhamusat being discarded.To our knowledge, the only Georgetown, British Guiana. Nautilus 37: 73-77. predatorsof fresh-watermolluscs known to be MURPHY,g. C. 1955. Feeding habits of the Ever- regularly deterred by snail opercula are scio- glade Kite (Rostrhamussociabilis). Auk 72: 204- rnyzid fly larvae (seeBerg 1961). 205. AlthoughSnail Kites have been seentaking RUIz-TIBf•N, E., J. R. PALMER,& F. F. FERGUSON.1969. nonsnailprey under conditionsof extremefood Biological control of Biomphalariaglabrata by shortage(Sykes and Kale 1974), the observa- Marisa cornuarietisin irrigation ponds in Puerto tions reportedin this paper are the first of the Rico. Bull. World Health Org. 41: 329-333. speciesfeeding on a genus of snail other than SNYDER,N. F.R. 1967. An alarm reactionof aquat- ic gastropods to intraspecific extract. Cornell Pomacea. It is uncertain, however, whether or University Agricultural Experiment Station not kites are successfulenough with Marisa to Memoir 403. be able to exist on it alone in areas where Po- --, & H. A. SNYDER.1969. A comparativestudy macea does not occur. Marisa was introduced of mollusc-predation by Limpkins, Everglade into southern Florida over 20 yr ago (see Ed- Kites, and Boat-tailed Grackles. Living Bird 8: mondson 1959) and has fairly recently spread 177-223. out into regions where Snail Kites occur. It is --, & . 1971. Defensesof the Florida ap- not yet abundant there, but is widespread-- ple snail Pomaceapaludosa. Behaviour 40: 175- we have seen shellsalong the TarniarniTrail 215. and as far north as EvergladesHoliday Park SYKES,P. W., JR.,& H. W. KALEII. 1974. Everglade Kites feed on nonsnail prey. Auk 91: 818-820. near AlligatorAlley. To datewe have not found Moous, K. H., & T. J. MAN DIjK. 1973. How do any of its shells at kite feeding perches. As Snail Kites extract snails from their shells? Ardea studies of M. cornuarietis have shown the 61: 179-185. BEHAVIOR, MORPHOLOGY, AND SYSTEMATICS OF SIRYSTES SIBILATOR (TYRANNIDAE)

WESLEYE. LANYON1 AND JOHNW. FITZPATRICK2 •AmericanMuseum of Natural History, New York, New York 10024 USA, and •FieldMuseum of Natural History,Roosevelt Road at LakeShore Drive, Chicago,Illinois 60605 USA

ABSTRACT.--Fieldand laboratorystudies support recent hypotheses that Sirystessibilator is closelyallied with the myiarchine flycatchersrather than with Tyrannusor with the co- tingids. The first confirmednest of this tyrant (locatedin a natural tree cavity), its foraging behavior, and the derived characterstates of the nasal septum and of the syrinx argue convincinglyfor its placementin an assemblagewith ,Rhytipterna, and Casiornis. Data, still lacking, on the color of the eggs and nature of the nest lining are needed for Sirystes(and for Rhytipternaand Casiornis)before considerationcan be given to the generic relationshipswithin the myiarchine flycatchers.Received I February1982, accepted16 July 1982.

THE systematicposition of the Sirystes (Sir- appears to prefer primary, dry-ground forest ystessibilator), traditionally placed in its own to other, more inundated forest types, and it monotypic genus, has been enigmatic for de- shows an unusual propensity for foraging in cades. Though included in the family Cotin- tall, emergent trees adjacent to treefalls. gidae by Ridgway (1907), largely on the basis Sirystesforages in a fashion virtually iden- of the scutellationof the tarsal envelope, Sir- tical to the unusual style of Myiarchus, espe- ysteshas been considereda tyrant flycatcherby cially in its frequent use of outward and down- Hellmayr (1927) and subsequentauthors. Hell- ward hovergleaning(Fitzpatrick 1980, unpubl. mayr allied it with the Tyrannusgroup, but notes). Moreover, while searching the sur- Traylor (1977), influenced by Warter's (1965) rounding vegetation,Sirystes uses the peculiar study of the tyrannoid skull, placed it in the posture typically associatedonly with the ge- Myiarchus group. On the basis of syringeal nus Myiarchus.The head is held out horizon- characters,Ames (1971) was unable to place it tally, stretched away from the body in more with conviction in any of his "structural exaggeratedfashion than in other flycatchers. groups" of tyrannids. Perhaps for this reason, the throat and crown This paper was prompted by Fitzpatrick's feathersare often erected, giving the bird a large discoveryof the first confirmednest of Sirystes, headed, "throaty" silhouette. Again in con- to which description he has added behavioral junction with these habits, both Sirystesand observationsmade in recentyears. Lanyon has Myiarchuspeer about and slowlybob the head contributedan analysisof cranialand syringeal in more deliberatefashion than any other fly- morphology, part of a more general study of catcher (except in the unrelated genus Mio- the putative relativesof Myiarchus. nectes).Members of the genus Myiarchus are BEHAVIOR amongthe few tyrannidsthat habitually"pop" the bill loudly while being held in the hand Sirystesis a regular but uncommon inhabit- after removal from a mist-net. Although the ant at the Cocha Cashu BiologicalStation, lo- significanceof this behavior is obscureat best, cated in the Manu National Park, dept. Madre it may be worthy of note that a Sirystesre- de Dios, (11ø55'S, 71ø18'W, elev. 380 m). moved from a mist-net (also at Cocha Cashu) It foragesalmost exclusively in the high canopy repeatedlypopped its bill while being banded of the alluvial plain forestbordering the mean- and photographed. dering Manu River (Terborgh et al. in press). The typical daytime vocalizationgiven dur- Mean foraging heights for 21 independent ing active foraging is a clear, ringing "wher- sightings in 1976 and 1977 was 38 + 8.5 m in p•w." During more intense calling, intrapair a forest whose average canopy treetop is be- displaying, and dawn-singing,this frequently tween 40 and 45 m (unpubl. data). $irysteshere is lengthenedinto "wher-pe-pep-pep-pew-pew

98 The Auk 100: 98-104. January 1983 January1983] Sirystessibilator (Tyrannidae) 99

visits occurred, the intervals between visits av- eraging 8.6 min (shortest interval 1 min, lon- gest 16 min). Both birds visited the nest with food, frequentlycalling loudly togetherin the nest tree. They appearedto be foragingup to 200 m from the nesttree, althoughthey almost invariably continuedto sally after prey items high in the nesttree itself, even while carrying largeboluses of food. Twice, the birds returned simultaneouslyto the nest tree; one waited perchedon twigs outside the hole, while the other entered and exited. Nest visits varied from just a few secondsup to about 1 min. One fecal sac removal was observed. The nest site was determinedthrough tele- photolenses to be about 32 m abovethe ground. In comparisonwith the birds' body sizes the hole appearedto be at least15 cm wide, per- hapslargerß It wasat the baseof a nearlyper- pendicular "elbow" in the tree limb, with the holerunning roughly horizontally into themain branch. An old break had rotted out to create the cavity,and it left a 6- to 8-cmlong lip rim- ming the lower edgeof the hole. The depth of the hole,penetrating horizontally into the limb, could not be determined,nor could any fea- tures of nest construction or contents be seen. Fig. 1. View fromground of Sirystessibilator nest Calicophyllurntrees are known locallyas "Palo site (white arrow near center)in emergentCalico- Calato" ("naked pole") becauseof their ex- phyllumtree, well above nearby forest canopy (dark traordinarily smooth, seeminglybarkless ap- vegetationto lower right and upper left). Note un- pearance(Fig. 1). This makes the trunk and usuallyopen surroundings at nestheight amidst sev- limbs extremelyslippery and difficultto climb, eral old treefallopenings. presumably for predators as well as for hu- mans. Judgingfrom the frequencyof nestvisits and ß . . pew." On the morningof 2 October1981, the largesize of the food bolusesbrought in incessantrepetitions of this latter song,in an on 2 October, the nest must have contained area known even in previousyears to be fre- well-developednestlings. Fitzpatrick left the quented by a pair of Sirystes,attracted JWF's study site the following day, but N. Pierpont attention to two birds high in a 45-m Calico- reported nest visits by both birds on 3, 4, and phyllurnspruceanurn tree (Rubiaceae)adjacent 5 October.Thereafter, no furtheractivity was to a slightly overgrowntreefall clearing (Fig. observedat the nest. Early on 6 October, at 1). Bothbirds were bobbing and flittingexcit- least two Sirysteswere heard activelycalling edly from perch to perch with their slightly about 100m from the nest. They did not come elongatedcrown feathersraised into vague near the nest tree during a 20-rain observation "crests." One bird was carrying in its bill a period there (1450--1510) nor on two subse- largewad of dark material,apparently . quent days. Assumingthe nestlingsfledged Severalminutes later, the bird carryingprey successfully,they apparentlydid so on 5 or 6 disappearedinto a holealong a majorlimb high October. in the tree. It remainedout of sight about 1 The only previous, published information on min, then flew backout, sat briefly a few me- the nestinghabits of Sirystesis Wetmore's(1972: tersabove the hole,then flew off to forageelse- 436)reference to birdsin "occasionally where. From 0755 to 0941 a total of 12 such nest examining cavities in tree trunks." 100 LANYONAND FITZPATRICK [Auk, Vol. 100

4

Fig. 2. Photographs,taken through a dissectingmicroscope, of the ventral aspectof the nasal region of the skulls of some tyrannine flycatchers(anterior end of the skull to the left): l. Sirystessibilator (USNM 347723);2. Myiarchustuberculifer (AMNH 7175);3. M. cephalotes(AMNH 9248);4. Rhytipternaimmunda (AMNH 11617);5. Deltarhynchusfiammulatus (AMNH 11391);6. Atlila cinnamomeus(AMNH 11615).Arrows indicate the internal supporting rod present in all skulls exceptthat of Attila; the knife-like ventral edge of the ossifiednasal septurn, s, lacks a trabecularplate; ossificationof the nasal capsuleindudes the alinasal turbinals,t, in all skullsexcept that of Attila. All specimensmagnified 16x.

MORPHOLOGY ment of Sirystesin the Cotingidae, but Ridg- Sirystesdiffers from mostMyiarchus in pos- way admitted that "the bill [of Sirystes]is typ- sessinga completesuture along the outer edge ically tyrannine, resembling closely that of of the tarsus as well as the usual full inner su- Tyrannusand the stouterbilled Myiarchi." The ture (i.e. tarsusholaspidean, not exaspidean). striking black-and-white plumage pattern of This arrangementof the tarsalenvelope was Sirystes,with diffuse olive dorsalstreaks, also the principalbasis for Ridgway's(1907) place- differs from the familiar and remarkablyuni- January1983] Sirystessibilator (Tyrannidae) 101 form plumagepattern of most Myiarchus.The compare the number, shape, and location of rectricesand wing covertsof the juvenal plum- these internal cartilagesas seen in Sirystes,in age of Sirystesare fringed with Antique Brown three other mylarchine flycatchers(Myiarchus (Smithe 1975), though this color is lacking in ferox, M. oberi,and Rhytipternaholerythra), and the adult plumage.This is a pattern character- in two other tyrannines (Deltarhynchusfiam- istic of all speciesof Myiarchus,including those mulatus and Attila cinnamomeus). speciesin which Antique Brown is lacking in The large cartilagesare J-shapedin Sirystes the adults. and the other myiarchineflycatchers, and their Recentanatomical studies involving the skull anterior ends are connectedto the ventral seg- (Warter 1965)and the syrinx (Ames 1971)have ment of the tracheo-bronchialjunction, at some indicated the effectiveness of these structures distance from the dorsal ends of the A-2 semi- in determining relationships within the sub- rings. The large cartilagesdo not have this oscines.Two complexesin particular have al- characteristicshape in Deltarhynchusand At- ready proven useful in diagnosingMyiarchus tila, where they can be seen to connectto the flycatchers and their relatives: (1) the nasal dorsal segmentof the tracheo-bronchialjunc- septurn, and specificallythe degree to which tion and the dorsal ends of the A-2 semirings. there is ossificationand buttressing with in- The smaller,ventral pair of cartilagesis lacking ternal supporting structures,and (2) the num- in the syrinx of Deltarhynchus. ber, shape, and position of the internal carti- The method of dissection and terminology lages of the syrinx (Lanyon 1982). follow Ames (1971). The nasal capsuleof Sirystes(Fig. 2, number DISCUSSION 1) is typicallytyrannine (as defined by Traylor 1977), in that the nasal septum is well-devel- The behavioral features described above, oped (ossified)but lacksa trabecularplate along even down to the quality of the song,once con- the ventral edge. The nasal septum is but- fused no lesscapable a field ornithologistthan tressed by a well-developed internal support- the late Paul Schwartzwith regardto birds that ing rod, identified with an arrow in Fig. 2. This he had recordedbut not collectedalong the Rio derived state of the nasal septum, found in Sir- Masparro, Barinas, . The recordings ystes, is shared with Myiarchus(20 of the 22 had been depositedwith the Library of Natural species;two still unavailable),Rhytipterna (all Sounds at Cornell University's Laboratory of three species),Casiornis (both species),and Del- Ornithologyand filed as "Myiarchusferox," but tarhynchusand is not found in any of the re- Lanyonwrote Schwartz that they could not have maining genera placed in the Tyranninae by been made from any speciespresently regard- Traylor (Lanyon1982, unpubl. notes).Note the ed as Myiarchus.Six monthslater Schwartzre- lack of an internal supporting rod in the nasal plied: "The birds from the Rio Masparro I've septum of Attila (Fig. 2, number 6). now definitelyidentified as Sirystessibilator. ! Sirystesshares a suite of syringealcharacters feel embarrassed,but not terribly so. If you with Myiarchus(15 speciesexamined thus far), already know the bird or if you some day ex- Rhytipterna (all three species), and Casiornis perienceit I'm sure you'll know what I mean. (both speciesexamined) that sets these genera In size, shape, attitudes and actionsthis bird apartfrom all other tyrannids (Lanyon 1982,un- is extremelyMyiarchus-like rather than like the publ. notes).In thesefour generathere are two groupwith which it is presumedto be related pairs of cartilagesthat lie within the internal [the Tyrannusgroup] .... ! recently encoun- tympaniform membranes: (1) a large, dorsal tered the bird again down near Burgua, alerted pair, which are J- or L-shaped and are con- by the voice. Here again I was convincedthey nectedanteriorly to the ventral segmentof the wereMyiarchus for the samereasons, until with tracheo-bronchialjunction (with no connection playback I finally got them to where the sun to the dorsal ends of the A-2 semirings), and was at my backand they were againsta foliage (2) a much smaller, ventral pair, of variable background.... ! collected one just as a shape and located within the membranous clincher." Schwartz's field experience,though connectionsthat extend between the posterior anecdotaland unconvincingby itself, was of endsof the largecartilages and the ventral ends sufficientimportance to alert both of us to the of the B-2 bronchial semirings. In Fig. 3 we need for further information about this species. 102 LANYOI•AND FITZPATRICK [Auk, Vol. 100

1' 2 '

a Ic Ic

sc.•' sc

Ic

Ic

sc• sc •

Ic

sc

Fig. 3. Photographs,taken through a dissectingmicroscope, of the dorsalaspect of the syringesof some tyrannine flycatchers:1. Sirystessibilator (YPM 2791);2. Myiarchus[erox (AMNH 8112);3. M. oberi(AMNH 6677);4. Rhytipternaholerythra (AMNH 8260);5. Deltarhynchusfiammulatus (AMNH 8128);6. Attila cinna- momeus(AMNH 8126).Large cartilages, lc; smallcartilages, sc; dorsal end of A-2 semirings,a. All specimens magnified19x.

Cavity-nesting is a rare and almost certainly with Myiarchus,one would hypothesizethat derived condition in the Tyrannidae, being they too are cavity-nestersand have eggslike most prevalentin the genusMyiarchus and its thoseof Myiarchus.That Sirystesis now known close relatives (Lanyon 1978, 1982). Casiornis to be a cavity-nesterlends support to the hy- has been reported to be a cavity nester, but the pothesis(Traylor 1977, Lanyon 1982) that Sir- eggsand nest lining have not been described ystesis a dose relative of Myiarchus. (fide Helmut Sick pers. comm.). The nesting The nest lining of Myiarchus is unique habits of the three speciesof Rhytipterna are among the 12 genera of tyrants in which nest- still unknown, but, on the basis of derived ing in tree cavitieshas been reported,bearing morphologicaland behavioralcharacters shared in mind that we still lack complete data for January1983] Sirystessibilator (Tyrannidae) 103

Rhytipterna,Casiornis, and Sirystes.Invariably, Ames (1971) concludedthat "the syrinx of the lining consistsof quantities of fur and Sirystesis typically tyrannid in cartilagesand , in addition to vegetable fibers, and musculaturewith little similarity to the simple frequentlythere are fragmentsof shedreptilian syringes of the Cotingidae," but questioned skin or plastic and paper substitutes(Lanyon Hellmayr's (1927)placement of the specieswith 1978,unpubl. notes).These unique contentsof the Tyrannusgroup, becauseit lacksalmost all the nest lining form an importantpart of the of the syringeal features of that assemblage. genericdiagnosis of Myiarchus,for they help Lanyon agreeswith Ames on both of these to embrace three of the more "aberrant" species points but has difficulty reconcilingAmes' fail- formerly placed in monotypic genera: semiru- ure to recognizethe syrinx of Sirystesas typi- fus, validus, and magnirostris.Because no de- cally mylarchine.The single specimeninitially tails could be obtained regarding the nest lin- availableto Lanyon agreesin all respectswith ing or the colorpattern of the eggsin the single the syringes of Myiarchus, Rhytipterna, and nest record reported here, the generic relation- Casiornis;it could not with certainty be sepa- ship of Sirystesand Myiarchusremains uncer- rated from a seriesof syringestaken from those tain. Indeed, on its own, the cavity-nestingbe- threegenera. Ames did not stainhis specimens havior of Sirystescould be a simple case of for bone and cartilage and consequentlymay convergence,as it most certainlyis in the case have overlooked some of the critical mylar- of Colonia(subfamily Fluvicolinae). chine characters. He kindly loaned his two The scutellation of the tarsal envelope, once specimensof Sirystesto Lanyon and gave per- given considerableweight in the classification mission for them to be stained with alcian blue of the suboscines, has been shown to be more and alizarin red (after Dingerkus and Uhler variable within genera and families than even 1977).These two specimenswere found to share Ridgway (1907:328-329, 336) would admit, and the same derived complex of syringeal char- severalworkers have cautionedagainst placing acters that distinguishes the myiarchine as- undue relianceon this character(Lanyon 1967, semblagefrom all other tyrannids. Ames et al. 1968, Snow 1973, Zimmer unpubl. Although we believe the affinitiesof Sirystes notes).We are reminded that Myiarchusvalidus with the myiarchinegroup of tyranninesto be was at one time placed in a monotypic genus unequivocal now, we recommendcaution in and transferredto the Cotingidae becausethe any revisionary considerationof the generic tarsal envelope was nonexaspidean(Ridgway relationshipsamong Sirystes, Myiarchus, Rhy- 1906). Indeed, the suggestionby Warter (1965) tipterna,and Casiornis.Descriptions of the col- and Ames (1971)that Rhytipterna,Casiornis, and or and markingsof the eggsand of the nature Attila be transferredfrom the Cotingidae to the of the nestlining, still lackingfor all but Myiar- Tyrannidaehas been followed for Peters'Check- chus,are needed before we can adequately ad- list (Snow 1973,Traylor 1977),even though the dressthe problem of generic limits and devise tarsal scutellation of these genera is nonexas- a meaningful phylogeny for this group. pidean. Virtually every feature of the plumagecolor ACKNOWLEDGMENTS and pattern of Sirystescan be found within the Fitzpatrick wishes to thank the Peruvian ministery genusMyiarchus: the dark cap and contrasting of agriculture,Direccion General Forestal y de Fauna, pale back characterizes the Andean popula- for their permissionto work at CochaCashu Biolog- tions of tuberculifer,the pale-tipped rectrices ical Station. JohnTerborgh greatly assistedwith the are found in apicalis, and the gray and white logisticson most trips into and out of the Manu Na- underparts distinguish sagraeand antillarum. tional Park. Nina Pierpont kindly visited the Sirystes Even the white rump patch and extensivewhite nest site numerous times and contributed her ob- in the wings, so conspicuousin northern Sir- servations.M. A. Traylor, Jr. read and criticized sev- eral drafts of this manuscript. Peter L. Ames loaned ystes, virtually disappear in the southernmost us two syringesof Sirystes,catalogued in the collec- populations of the species. With full cogni- tion of the PeabodyMuseum of Yale University, and zance of the range of variation within the 22 George Watson loaned us two skeletonsof Sirystes speciesof Myiarchus,it becomesdifficult to ar- from the collection of the National Museum of Nat- gue againstmerger of Sirystesand Myiarchus ural History. GuidoDingerkus and CharlesW. Myers, on the basis of plumage color and pattern. of the American Museum of Natural History, gave 104 LANYONAND FITZPATRICK [Auk, Vol. 100 valuable advice for staining and photographing the 1982. Behavior, morphology, and system- anatomicalmaterial. We thank these colleaguesfor atics of the FlammulatedFlycatcher of Mexico. their help. Auk 99: 414-423. RtDGWA¾,R. 1906. Some observationsconcerning LITERATURE CITED the Americanfamilies of oligomyodianpasseres. Proc. Biol. Soc. Washington, 19: 7-16. AMES,P. L. 1971. The morphologyof the syrinx in 1907. The birds of North and Middle passerinebirds. Bull. Peabody Mus. Nat. Hist., Yale Univ. No. 37. America, part 4. Bull. U.S. Natl. Mus. No. 50. --,M.A. HEtMERDtNGER,& S. L. WARTER. 1968.The SmITHE,F.B. 1975. Naturalist'scolor guide. New York, Amer. Mus. Nat. Hist. anatomyand systematicposition of the antpipits SNOW, D. W. 1973. The classification of the Cotin- Conopophagaand Corythopis.Postilia No. 114. gidae (Aves). Mus. Comp. Zool., Breviora, No. DINGERKUS,G, & L. D. UHLER. 1977. Enzyme 409. clearing of alcian blue stained whole small ver- TERSORGH,J. W., J. w. FITZPATRICK,& L. EMMONS. tebrates for demonstrationof cartilage. Stain Technol. 52: 229-232. In press. Annotated checklist of bird and mam- FITZPATRICK,J. W. 1980. Foraging behavior of mal speciesof CochaCashu Biological Station, Manu National Park, Peru. Fieldiana. Neotropicaltyrant flycatchers. Condor 82: 43-57. TRAYLOR,M. A., JR. 1977. A classification of the HELLMAYR,C. E. 1927. Catalogueof birds of the Americas and the adjacent islands, part 5. Publ. tyrantflycatchers (Tyrannidae). Bull. Mus. Comp. Zool. 148: 129-184. Field Mus. Nat. Hist., Zool. Ser., vol. 13. WARTER,S. L. 1965. The cranial osteologyof the LANYON,W. E. 1967. Revision and probable evo- New World Tyrannoidea and its taxonomic im- lution of the Myiarchus flycatchersof the West Indies. Bull. Amer. Mus. Nat. Hist. 136: 329- plications.Unpublished Ph.D. dissertation,Bat- 370. on Rouge, Louisiana, Louisiana State Univ. 1978. Revisionof the Myiarchusflycatchers WETMORE,A. 1972. The birds of the Republic of of South America. Bull. Amer. Mus. Nat. Hist. Panama, part 3. Smithsonian Misc. Coil. 150: 1- 631. 161: 427-628. THE FUNCTION OF SINGING IN FEMALE BLACK-HEADED GROSBEAKS (PHE UCTICU$ MELANOCEPHAL US): FAMILY-GROUP MAINTENANCE

GARY RITCHISON 1 Departmentof Biology,UMC 53, Utah StateUniversity, Logan, Utah 84322USA

ABSTRACT.--Theresponses of young Black-headedGrosbeaks (Pheucticus melanocephalus) to playbackof the songsof parentsand strangerswere examined.I found that young gros- moved about more, called more, and were more often oriented toward speakers in responseto parental song(both male and female) than in responseto the songsof strangers. I suggest that such a responseis part of a system used by Black-headedGrosbeaks to maintain family groups after the young fledge. In this system, an adult with food, but unawareof the locationof its young, singsto elicit beggingfrom its young. The parent bird is then able to locateand feed its young. Received31 August1981, accepted26 May 1982.

AMONG passerines, song is typically the and Hooker 1969, Bertram 1970, Payne 1971). function of the male. There is also much evi- Concerninginstances of female singing other dence,however, of regular, occasional,and ar- than antiphonalsinging or duetting, Kern and tificially inducedsong by females.It has long King (1972)have suggesteda number of func- been recognized,for instance,that femalesof tions, including stimulating the breeding ac- many speciescan be induced to sing by injec- tivities of the male. Nottebohm (1975) sug- tion of testosterone (Baldwin et al. 1940, Kern gestedthat singingby femalesmay influence and King 1972). There are also many reports of a bird's socializationand choiceof partner and, speciesin which the female sings only in ex- in some cases, aid in territorial defense. ceptionalcases, e.g. the WesternMeadowlark Among the speciesof birds in which singing (Sturnellaneglecta; Lanyon 1957), Indigo Bun- by females has been reported is the Black- ting (Passerinacyanea; Nolan 1958), Song Spar- headed Grosbeak (Pheucticusmelanocephalus; row (Melospizamelodia; Van Tyne and Berger see Fig. 1). Weston (1947), in a general study 1976), EasternPhoebe (Sayornisphoebe; Smith of the breeding behavior of the grosbeak,re- 1969), (Sialia sialis; Morton et ported that femalessang "while incubatingor al. 1978), and White-crowned Sparrow (Zono- brooding, usuallyas the male comesto take his trichialeucophrys; Kern and King 1972).In oth- placeon the eggsor young. Severaltimes dur- er species,singing appears to be a regularfea- ing nest-building, the female uttered songsin ture of femalebehavior, i.e. many specieshave the vicinity of the nest and always in the pres- beenreported in which femalescommonly sing ence of the male. The female will also occa- in a variety of situations (Table 1). sionally sing while foraging in the peripheral Despite these many observations, the sig- foliageof trees,but only when the male is close nificance of singing in females remains ob- by." Armstrong(1963) statedthat singing by scure. Possible functions, however, have been females tends to be characteristic of cardueline postulated. Armstrong (1963) suggestedthat finches.Van Tyne and Berger(1976: 249), in a antiphonal singing is a "means whereby con- general discussionof female song, suggested tact, rapport, and the social bond are main- that the songs of the female Black-headed tained." Duetting is believed to be important Grosbeakare nearly as elaborateas thoseof the in the synchronizationof breeding behavior male. and in the reinforcementof the pair bond (e.g. Despite these observations, the function of Thorpe and North 1965, Thorpe 1966, Hooker female songin the Black-headedGrosbeak, as in other species,is unclear. The present study 1 Present address: Department of Biological Sci- is an attempt to ascertainthe function(s) of this ences, Eastern Kentucky University, Richmond, song. The initial year of this 2-yr study was Kentucky 40475 USA. devoted to extensive field observation and re-

105 The Auk 100: 105-116. January 1983 106 GARY RITCHISON [Auk, Vol. 100

TABLE1. Reportsof singing in femalesamong various arian species.

Situations in which Suggested female singsa function(s) a Species NB NR I B FY TD PB-M C F-GM Source

Greenshank Simms 1958 Tringa nebularia European Nightjar X Selous 1905 Caprimulguseuropaeus Gray-capped Flycatcher X Skutch 1953 Myiozetetesgranadensis LoggerheadShrike X Armstrong 1963 Lanius ludovicianus European Dipper X X Van Tyne and Berger Cinclus cinclus 1976 American Dipper X X Bakus 1959a, b Cinclus mexicanus House Wren X X X Armstrong 1955 Troglodytesaedon Wrentit X Armstrong 1963 Chamaeafasciata Blue Wren X Robinson 1949 Malurus cyaneus ParadiseFlycatcher X Moreau 1949 Terpsiphoneviridis Elepaio X X X X Conant 1977 Chasiempissandwichensis Willie Wagtail X Robinson 1949 Rhipiduraleucophrys White-crowned Sparrow X X Blanchard 1941, Kern Zonotrichialeucophrys and King 1972 Grasshopper Sparrow X X X Smith 1959 Ammodramus savannarum Variable Seedeater X Gross 1952 Sporaphilaaurita Cuban Grassquit X Baptista 1978 Tiaris canora Gray Catbird X Palmer 1949 Dumetella carolinensis Mockingbird X Armstrong 1963 Mimus polyglottos Brown Thrasher X X X Thomas 1952 Toxostomarufum European Robin X Lack 1939, 1943 Erithacus rubecula Eastem Bluebird X Thomas 1946 Sialia sialis RussetNightingale Thrush X X X Skutch 1958 Catharus occidentalis Gray-cheekedThrush X X Van Tyne and Berger Catharus minimus 1976 European Blackbird Messmer and Messmer Turdus merula 1956 American Robin X Armstrong 1963 Turdusmigratorius Brown Towhee X X Marshall 1960 Pipilofuscus January 1983] Singingin FemaleGrosbeaks 107

TABLE 1. Continued.

Situations in which Suggested female sings• function(s)a Species NB NR I B FY TD PB-M C F-GM Source Abert's Towhee X X Marshall 1960 Pipilo aberti Orange-billedSparrow X Skutch1954 Arremon aurantiirostris Rose-breasted Grosbeak X X X X Ivor 1944, Dunham Pheucticus ludovicianus 1964 X X X X X Laskey 1944 Cardinalis cardinalis Common Grackle X Wiley 1976 Quiscalusquiscula Greenfinch X Ferguson-Lees1943 Carduelis chloris American Goldfinch X Berger 1953 Carduelis tristis Lawrence's Goldfinch X Linsdale 1950 Carduelis lawrencei Red Crossbill X Lawrence 1949 Loxia curvirostra White-winged Crossbill X Bent 1968 Loxialeucoptera Magpie Lark X X Robinson 1949 Grallina cyanoleuca Butcherbird X Robinson 1949 Cracticussp. Black-backedMagpie X Robinson 1949 Gymnorhinatibicen • NB - nest-building; NR = nest-relief; I - incubation; B - brooding; FY - feeding young; TD - territorial defense; PB-M - pair-bond maintenance;C - courtship;F-GM - family-groupmaintenance.

cording. During this period the situations in singing. Such singing apparently serves a ter- which females sang were noted, and hypoth- ritorial functionand, in addition, probably en- eses concerningthe function(s) of this song ables the female to maintain contact with the were derived. During the secondyear of the male. Females infrequently sang while forag- study, these hypotheses were tested experi- ing near the male (Weston 1947, pers. obs.). mentally. Singing, alone, is not sufficient to maintain a territory, at least early in the breeding sea-

SYNOPSIS OF THE BLACK-HEADED son, and agonistic encountersinvolving chas- GROSBEAK BREEDING CYCLE ing and even physical contact occurred. Nearly all chasesinvolved males, although several fe- The first birds arrived in the study area (Mal- male-female chases were observed, and, on ibu-Guinavah Campground, in Cache Nation- three occasions, females were observed chas- al Forest 10 km east of Logan, Cache County, ing males. On one of these occasions,a female Utah) about the first week in May. Observa- was observedchasing a male, and, upon land- tions indicated that some birds were paired ing, she sang one loud song. On another oc- upon arrival. Such pairs may have been formed casion a female appeared to engagein a brief on the wintering grounds or during migration. singing duel with a neighboring male. During the early part of the breeding season, Following territory establishment, Black- paired birds foraged together within their ter- headed Grosbeaksbecome progressivelyless ritories. The female usually followed the male aggressive.This change in behavior was quan- as he moved through the territory feeding and tified in two ways: (1) male singing rates de- 108 GARY Rr?cmso2½ [Auk, Vol. 100

:] TIME (•) TIME(•)

TIME(•) TIME(•)

Fig. 1. Representativesonograms of the songsof male (above)and female(below) Black-headed Gros- beaks.

clined as the seasonprogressed, and (2) the Incubation.--Both sexesare surprisingly vo- frequencyof intraspecificagonistic encounters cal on and around the nest. Males frequently (i.e. chasesand/or actualphysical encounters) sangwhile incubating.At timesthis songap- decreased(Fig. 2). peared to be in responseto the singing of Nest building.--Thefemale normallybuilds neighboringmales, i.e. a male would be qui- the nest, and I observed three instances in etly incubatingwhen, upon hearinga neigh- which females sang while involved in such boring male singing,the incubatingbird be- construction.Twice, females sang while gath- gan to sing. At other timesthe male'ssinging ering nestmaterials, and, on anotheroccasion, appearedto be a signal to the female that he a female was observedsinging while sitting in was about to leave the nest. a partially constructednest. On eachof these Femalesrarely sing while incubating.On two occasions a male was within a few meters of occasionsincubating females sang in apparent the singing female. responseto the singingof neighboringmales. January1983] Singingin FemaleGrosbeaks 109

30'

14 -12 -10 8 6 4 -2 0 2 4 6 TIME IN DAYS Fig. 2. Distribution of intraspecificagonistic encounters (chases and/or actual physical encounters) among male and femaleBlack-headed Crosbeaks (Day 0 = first egg ]aid).

In most cases, however, females showed no re- hatching, the adults maintained the same sponseto the singingof neighboringmales. scheduleas when incubating. Both adults fed On many occasionsa male or female would and brooded the young, and their behavior approachthe nest and find its mate quietly in- when changingplaces on the nest was similar cubating. At these times, males and females to that during incubation,with one significant frequentlyuttered "chip" callsor sang.The in- difference.As the broodingperiod progressed, cubating bird, upon hearing its mate, would the femalesbegan to sing more frequently (Fig. then leave the nest. 3). Parentalcare.--During the first few dayspost- By the eighth day post-hatchingthe young

42 47

24 -

20

• = ,• Red[19781 •,•,•, RedBridge 119771 w•-.• Riverside o ...... • Gulnavahcs8 ." !1

o

z g8 /'-L :J'l /' / ! I1'

/ ,• _,_., ' ß i,1,.i lll,i i,-,i-, 1,1,-,i.,,,11 i. first youn9 young egg laid hatch leave the nest TIME IN DAYS

Fig. 3. Singing ratesof seIectedfemales during the 1977and 1978breeding seasons. 110 GARY RITCHISON [Auk, Vol. 100 were brooded infrequently. Adults approach- that both male and female Black-headed Gros- ing the nest to feed the young frequently vo- beakssing in the following situations:(1) as calized. These vocalizations were either "chip" they foragewith their matesbefore nesting, (2) callsor songs.Upon the arrival of the adult at after chasingneighboring males from their ter- the nest, the young grosbeaksimmediately be- ritory, (3) during nest building, (4) during in- gan calling. At times the young began calling cubation, (5) during change-oversat the nest upon hearing the calls or songsof their par- (both during incubation and brooding), (6) ents. when comingto the nest to feed the young, (7) The young may leave the nest as early as the when attempting to locate and feed fledged ninth day post-hatching,although departure at young. 10-14 days post-hatching was more common Theseobservations, along with thoseof Head (n = 21, •v= 11.5). After leaving the nest they (1902, 1904) and Weston (1947), suggestedthe scatteramong the shrubsnear the nest, perch- followinggeneralized functions of femalesong ing on low branches.During the first few days, in the Black-headed Grosbeak: (1) maintenance the young are rather quiet. As they are unable of thepair bond, (2)territorial defense, (3) fam- to fly at this time, they remain within a re- ily-group maintenance. strictedarea, and the adults generallyhave lit- Other reports (Ritchison in prep.) have re- tle trouble locating and feeding them (Weston vealed that female song in the Black-headed 1947, pers. obs.). If, however, a parent is un- Grosbeak does appear to be important in the able to locatea young bird, the parent will be- maintenanceof the pair bond, althoughit plays gin to utter "chip" calls and/or songs. Upon no role in territorial defense.The main objec- hearing their parents' vocalizations, young tive of the present study was to examine the grosbeaksrespond by uttering "phee-oo" and/ possiblerole of singingby femaleBlack-head- or "hunger-distress" calls. In this manner the ed Grosbeaksin family-group maintenance. parents and young are able to maintain con- tact. Maintaining contactbetween the parentsand MATERIALS AND METHODS their young obviouslybecomes more difficult Fieldworkwas conductedduring the breedingsea- after the young attain flight (approximately15 son of 1978at Malibu-Guinavah Campground.Eight days post-hatching). To maintain contact, younggrosbeaks from five differentnests were tested grosbeaksappear to use the same system de- in theseexperiments. All birds testedwere between scribed above. When parentshave food for the 10 and 17 days of age. young, but are unsure of the locationof the Apparatus.--Testing was conducted in an auto- mobile from which the birds had no contact (visual young, the parentsbegin to utter "chip" calls or vocal)with their parentsor other grosbeaks.The or, more frequently, songs. Upon hearing a test apparatusconsisted of a testing arenaand a port- parent, young grosbeaksutter "phee-oo" and/ abletape recorder(Nagra IIIB) connectedby cableto or "hunger-distress"calls. The parentthen flies a portable loudspeaker. to the young bird and feeds it. Occasionally, The arena was a box with floor dimensions of after hearinga parent sing, a young grosbeak 125 x 28.8 cm and walls 28.8 cm high. The floor and will fly to within a few meters or less of the long walls of the box were madeof wood while the adult and, if not fed immediately, will begin short walls consisted of cheesecloth.To provide a calling. referencegrid for the purposeof scoringa bird's po- At this stage(2-3 weekspost-hatching), fam- sition in the arena, the floor was marked out into 25 5-cm cells. In addition, the top of the apparatuswas ily groupsbegin to wander and no longerstay covered with a wire screen. within their territories. Because of this wan- During teststhe observersat in the front seatof dering, it is difficult to observespecific family the automobileand viewed the floor of the arena by groups over long periods, and, therefore, the means of a mirror suspendedat an angle above it. duration of such groups remains a question. For the playbackof recordedsongs the portableloud- Weston (1947) reported seeing young gros- speakerwas placedagainst the cheeseclothat one end or the other of the arena, and the tape recorder beaksbeing fed by adultsin earlyAugust, but was controlledby the observerfrom the front seat. he was unable to determine the actual length Playbackrecordings.--The songs of parents and of the dependent period. strangerswere recordedusing a Nagra IIIB tape re- Summarizingthe aboveobservations, I found corder with an Altec 633A microphone, which was January1983] Singingin FemaleGrosbeaks 111 housed in a 62-cm parabolicreflector. All recordings direct line from it to the end. In each 15-s interval were madeat a tape speedof 19 cm/s(71/2 ips). From the bird was thus scored as oriented toward the these recordingswere made 3-min test tapes, with speaker end, or as oriented toward the other end, or songsspaced at 15-s intervals. as unoriented. By counting up the intervalsin which Testing procedure.--Each bird was tested alone. orientation was one way and the intervals in which Initially, the young bird was placed at the central it was the other way I obtained the two scores.In a position on the floor grid of the arena, and its be- 3-min pre-test, test, or post-test, therefore, the max- haviorduring the next3 min wasobserved and scored imum possible score for one end, and for the two without playbackof any soundthrough the speaker. ends taken together, was 12. Such a period will be referredto as a "pre-test." At (ii) Position. As in the procedurefor orientation, the end of a pre-test, the young grosbeakwas re- two scoreswere obtained: one for the speaker end placed at the central position of the floor grid. For of the arena and one for the other end. The central the next 3 min songswere played, and again the position in the arena, from which a bird started out behavior of the bird was observed and scored. Such in eachpre-test, test, and post-test,carried a position a period with playbackwill be referredto as a "test." value of 0. The 5-cm transverse divisions of the floor Immediately after the test period, the young bird was carried position values, which increased from 1 to againreplaced at the centralposition of the floor grid. 12, reading from the central position to one side or Its behaviorduring the next 3 min was again ob- the other. At the end of each 15-s interval, a bird's served and scored.Such a period without playback positionon the floorgrid, determinedby the location after the test period will be referred to as a "post- of its feet, was notedand scored.Summing the scores test." for the two sides separatelygave the two position Eachbird was exposedto four typesof test in this scores.The maximumpossible score in a 3-min pre- way: two parental tests, in which the songsof its test, test, or post-test, for one side or for the two mother and father were played, and two "stranger" sidestaken together, was therefore144 (12 x 12). The tests, in which the songs of strange males and fe- maximum score would result if a bird spent all 12 maleswere played. The young birds experiencedeach intervals of a pre-test, test, or post-testwithin 5-cm of the four types of pre-test/test/post-testsequences of one or other of the end-walls of the arena. twice, once with the speaker at one end of the arena (iii) Position Change. A 15-s interval was scored and once with the speaker at the other end. The se- as positive for position changeif the bird's position quenceof presentationwas either parental-stranger- on the floor grid at the end of that interval was dif- parental-stranger or stranger-parental-stranger-pa- ferent from what it had been at the end of the pre- rental; as many birds experiencedthe one as the oth- ceding interval, regardlessof direction. The maxi- er. The sequencein which the endsof the arena were mum scorefor position changein a 3-min pre-test, used was also varied among birds, and the order of test, or post-test was therefore 12. ends was independentof the orderof test types. The (iv) Calling. The types and numbersof callsgiven complete sequenceof testing for a young grosbeak by a bird during the pre-test,test, and post-testpe- thus consistedof 8 pre-test/test/post-testruns, each riods were noted. 9 min in duration. I allowed 5 min between each run Playbackexperiments with free-living young.--In ad- for rewinding and changing test tapes on the tape dition to tests in the apparatus,a seriesof experi- recorder.The young bird remained in the arena dur- ments were performed with young grosbeaksin a ing this time. natural setting, i.e. while the birds were perched in For the tests the volume controlsof the tape re- a bush or small tree. Each experimentconsisted of corderwere adjustedso that modulometerreadings three 5-min segments(pre-test, test, post-test),and were similar for each tape used. No precise mea- throughout each test all sounds and nonvocal behav- surementsof soundintensity levelsin the arena were ior were noted. Each bird was tested twice with its made, however. mother'ssongs, and trials were at least 1 day apart. Scoring.--Forthe purposesof scoringthe behavior The speakerwas placed5-10 m from the young gros- of the young grosbeaks,each pre-test, test, and post- beaks in these experiments. test was divided into 12 15-s intervals. At the end of Playbackexperiments with adult females.--The re- each 15-s interval, a bird's position, orientation, and sponsesof femalesto playbackof the "phee-oo" calls the types and numbers of calls were noted. Details of their young were also examined in the field. The of the conventionsused for scoringthe behavior of proceduresfollowed in these tests were the same as the young grosbeaksare as follows: those used in the playback experimentswith the (i) Orientation. Two scores for orientation were young. Each experiment consisted of three 5-min accumulated:one for orientationtoward the speaker segments(pre-test, test, and post-test).Throughout end of the arena and one for orientation toward the each test all sounds and non-vocal behavior were other end. A bird was judged as orientedto one end noted. Eachfemale was tested twice with the "phee- or the other if it was pointed directly toward that oo" calls of one of her young. Different trials were end or in a direction within 20 ø to either side of the at least 1 day apart. 112 GARYRITCHISON [Auk, Vol. 100

TABLE2. Responsesof young grosbeaksto playbackof female songin the test apparatus.

Significancelevels t',c

PPT SPT PPT PT PAT Mean scores a'e VS. VS. VS. VS. VS. PPT SPT PT ST PAT SAT SPT ST PT ST SAT

Orientation to speaker 2.4 1.7 5.4 2.4 5.1 2.5 NS NS NS 0.05 NS Orientation to other end 1.3 1.9 1.3 2.0 1.2 2.8 NS NS NS NS NS Position score, speakerend 18.1 8.6 33.6 17.9 13.5 17.0 NS NS NS NS NS Position score, other end 31.0 37.1 28.8 32.4 32.0 25.0 NS NS NS NS NS Position change 2.2 4.5 2.4 2.0 0.6 1.2 NS NS 0.02 0.02 NS Number of "phee-oo" calls 17.4 15.7 54.2 27.0 37.6 26.8 NS 0.05 0.01 0.01 0.05 Number of "hunger- distress" calls 0.3 0.3 10.5 2.2 0 0 NS NS 0.01 0.02 NS

• The mean scoreswere derivedfrom two setsof experimentson eachof eight birds. The significancelevels are accordingto Wilcoxon matched-pairstests, two-tailed. • P < numbergiven: NS = not significant. •' PPT = parentalpre-test; SPT = strangerpre-test; PT = parentaltest; ST = strangertest; PAT = parentalpost-test; SAT = strangerpost-test.

RESULTS of response (Tables 2, 3). As in the parental tests,the young grosbeaks'first responseto the Responsesof younggrosbeaks to parentalsong songsof strangerswas often a seriesof "hun- in the test apparatus.--Theresults are summa- ger-distress"calls followed by nearly contin- rized in Tables2 and 3, togetherwith signifi- uous "phee-oo" calls. This vocal response to cance levels, given by Wilcoxon comparisons, the songs of strangers, however, was signifi- for the differencesamong all pre-tests, tests, cantly less pronouncedthan the responseto and post-tests.As these figures and compari- parental song (Tables 2, 3). In addition, sonsclearly show, the young grosbeaksmoved strangers'songs elicited no significant loco- about more, called more, and were more often motory responses.Finally, young grosbeaks oriented toward the speaker in responseto pa- showeda significantorientation away from the rental songthan to the songsof strangers.The speakerin responseto the playbackof the songs effectson a young grosbeak'sbehavior of play- of strangemales (Table 3). ing the parentalsongs were spectacularin most Playbackexperiments with free-livingyoung.- cases.Typically a bird's behavior in the pre- In the test apparatus, as well as under natural tests consistedof standing or sitting near the conditions, young grosbeaksresponded to pa- centerof the arenaand giving occasional"phee- rental songby uttering "phee-oo" callsand/or oo" calls. At the sound of the parental song, "hunger-distress"calls. A comparisonof the however, there was usually an immediate and responsesof young birds in the apparatusand sudden change:the young grosbeakraised its in a natural setting, however, revealeddiffer- head and started calling (often beginning by ences with respect to orientation and ap- uttering a seriesof "hunger-distress"calls fol- proach. Under natural conditionsyoung gros- lowed by nearly continuous"phee-oo" calls). beaks showed a significant tendency to In general, then, there was incessantcalling approachthe speaker(Table 4); birds in the test and locomotion (i.e. position change) in the apparatus,however, showed no suchtendency parental tests. (Tables2 and 3). Young grosbeaksin the ap- Responsesof younggrosbeaks to the songsof paratus did, however, show increased loco- strangersin the test apparatus.--Althoughthe motion. These results, although not predicted responsesof the young grosbeaksto playback initially, might be explainedas follows.In pre- of the songsof strangerswere lesspronounced, cocial species,such as the Laughing Gull (Lar- they were still significantin severalcategories us atricilla; Beer 1970a,b) and Ring-billed Gull January1983] Singingin FemaleGrosbeaks 113

TABrE3. Responsesof young grosbeaksto playback of male song in the test apparatus.

Significancelevels a'b

PPT SPT PPT PT PAT Mean scores a'b VS. VS. VS. VS. VS. PPT SPT PT ST PAT SAT SPT ST PT ST SAT

Orientation to speaker 2.8 2.4 3.2 1.5 2.3 2.4 NS NS NS 0.05 NS Orientation to other end 1.5 1.8 2.1 3.4 2.5 2.4 NS 0.02 NS NS NS Position score, speaker end 13.1 11.8 18.2 11.5 27.2 31.9 NS NS NS NS NS Position score, other end 12.3 23.6 37.0 36.4 19.8 36.6 NS NS NS NS NS Position change 0.6 0.7 5.2 1.7 1.7 1.1 NS NS 0.01 0.01 NS Number of "phee-oo" calls 17.9 23.0 62.9 44.1 37.4 28.7 NS 0.01 0.01 0.01 0.01 Number of "hunger- distress" calls 0 0.1 7.9 9.3 2.7 0 NS 0.02 0.02 NS NS

Mean scoresand significancelevels as in Table2. PPT = parentalpre-test; SPT = strangerpre-test; PT = parentaltest; ST = strangertest; PAT = parentalpost-test; SAT = strangerpost-test.

(L. delawarensis;Evans 1970b), chicks have been less and call or attempt to "fly" out of the ap- found to show orientation and approach re- paratus to locate and approach their unseen sponsesto the played-backcalls of their par- parent. Such reasoning might explain the ab- ents. Because precocial chicks normally ap- senceof significantorientation or approachbe- proach their parents by walking or running, havior by young grosbeaksin the test appa- the quickestroute to a "callingparent" (i.e. the ratus. speaker) is to orient toward and approach the Playbackexperiments with adultfemales.--The sound source. In the grosbeak,on the other responsesof females to the playback of the hand, young birds come in contactwith their begging calls ("phee-oo" calls)of their young parents by flying toward them or, if they are were significantin severalcategories, i.e. dis- not yet capableof flight, by remainingmotion- tanceof closestapproach, number of songs,and less and calling the adults toward them. In the syllablesper song (Table 5). Becauseonly the apparatus,therefore, young grosbeakswould calls of their own young were played to indi- not be expectedto walk or run toward the sound vidual females, these results do not constitute source(speaker), but instead(depending on the proof of individual vocal recognitionof young age of the young) would either remain motion- by females. As will be discussedlater, how-

TABLE4. Responsesof free-living young Black-headedGrosbeaks to playback of their mother's song.

Closest Number Number approach Number of of "hunger- Number (m) of flights "phee-oos" distress" of "chips" Responsesa Pre-test period (PTP) 7.00 0 0.20 0 0.10 Test period (P) 3.30 0.80 2.50 1.60 0.40 Post-testperiod (PP) 3.50 0.30 0 0 0.30 Significancelevels • PTP vs. P 0.05 0.05 0.05 NS 0.05 PTP vs. PP 0.05 NS NS NS 0.05 P vs. PP NS 0.05 0.05 NS NS

Values for responsesare averagesfor all tests.The significancelevels are accordingto paired t-tests (n = 6). p < numbergiven; NS = not significant. 114 GARY RITCHISON [Auk, Vol. 100

TABLE5. Responsesof femalesto playbackof the "phee-oo"calls of their young.

Closest approach Number Syllables Number Number Number of Number (m) of songs per song of "chips"of "wheets""distress"of flights Responsesa Pre-test period (PTP) 7.00 0 0 5.70 0 0 5.70 Test period (P) 3.00 1.70 5.60 23.20 10.40 0.20 7.20 Post-testperiod (PP) 4.80 3.80 4.70 13.90 1.20 0 3.20 Significancelevels b PTP vs. P 0.01 NS NS 0.05 NS NS NS PTP vs. PP 0.05 0.05 0.001 NS NS NS NS P vs. PP 0.05 0.05 0.05 NS NS NS 0.01

Valuesfor responseare averagesfor all tests.The significancelevels are accordingto pairedt-tests (n = 4). P <: numbergiven; NS = not significant.

ever, these results do lend support to the con- stated that a young Mockingbird (Mimuspoly- clusionthat singingby femalesis importantin glottos)recognized the voiceof the parentwho family-group maintenance. fed it and startedto beg on hearing it. Nestling European Blackbirds (Turdus rnerula) are re-

DISCUSSION ported to know their mother by her food call (Messmer and Messmer 1956), and fledged Recent studies involving the recognitionof young apparently recognize their fathers' parents by their young have concentratedon vocalizations(Thielcke-Poltz and Thielcke 1960). colonial species, e.g. Common Murres (Uria Young Ring Doves (Streptopeliarisoria) and aalge; Tschanz 1965, 1968), Black-billed Gulls Chiffchaffs (Phylloscopuscollybita) are also re- (Larus bulleri; Evans 1970a), Laughing Gulls ported to recognize the calls of their mother (Beer1970a, b), Ring-billed Gulls (Evans1970b), (Craig 1908,Gwinner 1961).Other authorshave and Black-LeggedKittiwakes (Rissatridactyla; reported observationssuggesting that altricial Cullen 1957). Evidenceof individual recogni- young may recognize parental song. For ex- tion betweenparents and young has been found ample, Saunders(1929) referred to young House in every colonial speciesin which it has been Wrens (Troglodytesaedon) being stimulatedto sought, with the exceptionof the Blackqegged opentheir bills by the males'song. Young Snow . The survival value of this recogni- Buntings(Plectrophenax nivalis) may alsobe able tion in colonial speciesseems quite apparent. to distinguishtheir father'ssong from the song Becausea young bird is surroundedby adults of other males (Armstrong1963). The present who are not its parents, who are unlikely to studyprovides clear evidence that young Black- feed it, and who may even attack it, individ- headed Grosbeaksare able to recognize the uals who beg only from their own parents songsof their parents. should conserveenergy and have a selective Grosbeakfamily groupsbegin to wander 2- advantage over indiscriminating young. The 3 weeksafter hatchingand no longerstay with- caseof the Black-leggedKittiwake may be re- in their breeding territories. Becausethe birds garded as an exception that proves this rule, are moving through thick vegetationin unfa- for in that speciesthe young remain confined miliar areas,maintaining contact becomes more to the nest until they fledge (Cullen 1957), so difficult. Under such conditions individual vo- that up to that time the youngdo not normally cal recognitionis essential.Without such rec- encounter adults other than their parents. ognition, the maintenance of family groups In contrast to the situation described for co- would probably be impossible.The need for lonial species,there is little information avail- the recognition of parental song by young able concerningthe recognitionof parents by grosbeaks,therefore, is apparent. Even with their young in noncolonial, altricial species. such recognition,however, it would certainly Only a few observationssuggesting the pos- be .possiblefor young grosbeaksto stray from sibility of suchrecognition have been report- the family 'groups. Given that possibility, a ed. For example, Michener and Michener (1935) positive responseto the songs of other adult January1983] Singingin FemaleGrosbeaks 115 grosbeakswould be advantageous.Such a re- BAKUS,G.J. 1959a. Observationsof the life history sponse to strange adults would presumably of the Dipper. Auk 76: 190-207. tend to enhancethe chancesof "adoption" by 1959b. Territoriality, movementsand pop- ulation density of the Dipper in Montana. Con- foster parents. This, in fact, appears to be the dor 61: 410-425. "strategy" used by young grosbeaks.Although BALDWIN, F. M., H. S. GOLDIN, & M. METFESSEL. fledglings respond more strongly to the songs 1940. Effects of testosterone on female Roller of their own parents,they alsoshow significant Canaries under complete song isolation. Proc. responsesto the songs of strange males and Soc. Exp. Biol. Med. 44: 373-375. females (Tables 2, 3). BAPTISTA,L. F. 1978. Territorial,courtship and duet The responsesof young grosbeaksto paren- songsof the Cuban Grassquit(Tiaris canora). J. tal song representbut one side of the parent- Ornithol. 119: 91-101. young relationship. The responsesof parents BEER,C. G. 1970a. On the responsesof Laughing to the vocalizations (or absence of vocaliza- Gull chicks to the calls of adults. I. Recognition of the voices of the parents. Anim. Behav. 18: tions) of their young are equally important. 652-660. Among older fledglings, contactwith parents 1970b. Individual recognition of voice in generallyresults from thesefledglings flying to the socialbehavior of birds. Adv. Study Behav. the parents in response to parental song. 3: 27-74. Youngerfledglings, however, as well as young BENT,A. C. 1968. Life histories of North American birds that have left the nest but are not yet Cardinals,grosbeaks, buntings, towhees, finch- capableof flight, maintain contactwith parents es, sparrows,and allies. U.S. Natl. Mus. Bull. by means of the mutual recognition system No. 237. discussedpreviously, i.e. an adult with a food BERGER,A. J. 1953. Female American Goldfinch item, but unaware of the location of its young, warbles while incubating. Wilson Bull. 65: 217- 218. will often begin singing to elicit begging BERTRAM,B. 1970. The vocalbehavior of the Indian ("phee-oo" and/or"hunger-distress" calls) from Hill Mynah. Anim. Behav. Monogr. 3: 81-192. its young. In this way a parent is able to locate BLANCHARD,B. D. 1941. The White-crowned Spar- its young. The responsesof females to the rows (Zonotrichialeucophrys) of the Pacificsea- playback of the "phee-oo" callsof their young board: environment and annual cycle. Univ. appearto verify the existenceof sucha system. Calif. Publ. Zool. 46: 1-178. Upon hearing the playback,females showed a CONANT,S. 1977. The breedingbiology of the Oahu significant approach response,as well as sig- Elepaio. Wilson Bull. 89: 193-210. nificant increasesin the number of flights and CRAIC,W. 1908. The voicesof pigeonsregarded as number of "chip" calls (Table 5). Such re- a means of social control. Amer. J. Sociol. 14: 66-100. sponses would, under natural conditions, en- CULLEN,E. 1957. Adaptations in the Kittiwake to able a parent to locatea young grosbeakquick- cliff-nesting. Ibis 99: 275-302. ly. On the other hand, once playback ended, DUNHAM, D. W. 1964. Behavior of the Rose-breast- females showed significant increasesin sing- ed Grosbeak, Pheucticus ludovicianus. Unpubl. ing rates. Under natural conditions this would Ph.D. Thesis. Ithaca, New York, Cornell Univ. elicit calling by young grosbeaksand, thus, al- EVANS,R. M. 1970a. Parental recognition and the low a parent to locate its offspring. "mew" call in Black-billed Gulls. Auk 87: 503- 513. ACKNOWLEDGMENTS 1970b. Imprinting and mobility in young I am grateful to Keith L. Dixon for his help in all Ring-billed Gulls, Larusdelawarensis. Anim. Be- phasesof this study. I also wish to thank E. C. Oaks, hav. Monogr. 3: 195-248. B. Gilbert, J. Gessaman, and I. Palmblad for their FERGUSoN-LEEs,I. J. 1943. Song of female Green- many helpful suggestions.This study was supported finch. Brit. Birds 39: 244. by grants from The Frank M. Chapman Fund of the GROSS,A. O. 1952. Nesting of Hicks' Seedeaterat American Museum of Natural History and Sigma Xi. Barro Colorado Island, Canal Zone. Auk 69: 433- 446. LITERATURE CITED GWINNER, E. 1961. Beobachtungen tiber die ARMSTRONG, E. A. 1955. The wren. London, Ox- Aufzucht und Jugendentwicklung des Weiden- ford Univ. Press. laubsiingers.J. Ornithol. 102: 1-23. 1963. A study of bird song. London, Ox- HEAD, A. 1902. A study of the Black-headedGros- ford Univ. Press. . Condor 4: 119-121. 116 GARYRITCHISON [Auk, Vol. 100

1904. Nesting habits of the Black-headed ROBINSON,A. 1949. The biologicalsignificance of Grosbeak. Condor 6: 35-37. bird song in Australia. Emu 48: 291-315. HOOKER,T., & B. I. HOOKER. 1969. Duetting. Pp. SAUNDERS,A. A. 1929. Bird song.New York State 185-205 in Bird vocalizations (R. A. Hinde, Ed.). Mus. Handbook 7: 1-121. London, CambridgeUniv. Press. SELOUS,E. 1905. Bird-life glimpses.London, Con- IVOR, H. R. 1944. Bird study and some semi-cap- stable and Co. tive birds: the Rose-breasted Grosbeak. Wilson S•MMS, E. 1958. The conversational calls of birds as Bull. 56: 91-104. revealedby new methodsof recording.Acta 11th KERN, M. D., & J. R. K•NG. 1972. Testosterone- Intern. Ornithol. Congr. 1954: 623-626. induced singing in female White-crowned Spar- SKUTCH, A. F. 1953. How the male bird discovers rows. Condor 74: 204-209. the nestling. Ibis 91: 1-37. LACK, D. 1939. The behaviour of the robin. Proc. 1954. Life histories of Central American Zool. Soc. London 109: 169-178. birds. Pacific Coast Avifauna 31: 1-448. 1943. The life of the robin. London, With- 1958. Life history of the Violet-headed erby. . Wilson Bull. 70: 5-19. LANYON,W. E. 1957. The comparativebiology of SMITH,R. L. 1959. Songsof the GrasshopperSpar- the meadowlarks (Sturnella) in Wisconsin. Publ. row. Wilson Bull. 71: 141-152. Nuttall Ornithol. Club 1: 1457. SMtTH,W.J. 1969. Displaysof Sayornisphoebe. Be- LASKEY,g. R. 1944. A study of the Cardinal in haviour 33: 283-322. Tennessee. Wilson Bull. 56: 27-44. TH•ELCKE-POLTZ, H., & G. THIELCKE. 1960. Aku- LAWRENCE, L. DE K. 1949. The Red Crossbill at stisches Lernen verscheiden alter schallisolierter Pimisi Bay, Ontario. Can. Field-Natur. 63: 147- Amseln und die Entwicklungerlernter Motive 160. ohn und mit k•nstlichem Einfluss von Testos- LINSDALE, J. M. 1950. Observations on the Law- teron. Z. Tierpsychol. 17: 211-244. rence Goldfinch. Condor 52: 255-259. THOMAS,R. H. 1946. A study of Eastern MARSHALL,g.J. 1960. Inter-relations of Abert and in Arkansas. Wilson Bull. 58: 143-183. Brown Towhee. Condor 62: 49-64. 1952. Crip, come home. New York, Harper MESSMER, E., & I. MESSMER. 1956. Die Entwick- and Brothers. lung der Lautausserungenund Einiger Verhal- THORPE,W. H. 1966. Vocal imitation in the Trop- tensweisen der Amsel unter Naturuchen Bedin- ical Bou-BouShrike as a meansof establishing gungenund nachEinzelaufzucht in Schalldichten and maintainingsocial bonds. Ibis 108:432-435. Raumen. Z. Tierpsychol. 13: 341-441. --, & M. E. W. NORTH. 1965. Origin and sig- MICHENER,H., &: J. R. MICHENER. 1935. Mocking- nificance of the power of vocal imitation with birds: their histories and individualities. Con- specialreference to the antiphonalsinging of dor 37: 97-140. birds. Nature 208: 219-222. MOREAU,R. E. 1949. The breeding of a Paradise TSCHANZ,B. 1965. Beobachtungenund Experi- Flycatcher.Ibis 91: 256-279. mente zur Entstehungder "pers6nlichen"Be- MORTON, E. S., M. S. GEITGEY, & S. MCGRATH. 1978. ziehund zwischen Jungvogelund Eltern bei On bluebird "responses to apparent female Trottellummen. Verb. Schweiz. Naturforsch. Ges. adultery." Amer. Natur. 112: 968-971. 1964: 211-216. NOLAN,V. 1958. Singingby femaleIndigo Bunting 1968. Trottellummen die Entstehung der and Rufous-sided Towhee. Wilson Bull. 70: 287- pers6nlichenBezichungen zwischen Jungvogel 288. und Eltern. Z. Tierpsychol.Suppl. 4. NOTTEBOHM,F. 1975. Vocal behavior in birds. Pp. VAN TYNE, J., & A. J. BERGER. 1976. Fundamentals 287-332 in Avian biology, vol. 5 (D. S. Farner of .New York, Wiley and Sons. and J. R. King, Eds.). New York, AcademicPress. WESTON,J. G., JR. 1947. Breedingbehavior of the PALMER,R. S. 1949. Maine birds. Bull. Mus. Comp. Black-headed Grosbeak. Condor 49: 54-73. Zool. 102: 1-656. W•LEY, R. H. 1976. Affiliation between the sexes PAYNE,R.B. 1971. Duetting and chorussinging in in common Grackles. II. Spatial and vocal coor- African birds. Ostrich Suppl. 9: 125-146. dination. Z. Tierpsychol.40: 244-264. NOTES ON THE BEHAVIOR OF THE COSTA RICAN SHARPBILL (OXYRUNCU$ CRISTATU$ FRATœR)

F. GARY STILES 1 AND BRET WHITNEY 2 •Escuelade Biologia,Universidad de CostaRica, CiudadUniversitaria "Rodrigo Facio," Costa Rica, and •602 Terrace Mountain Drive, Austin, Texas 78746 USA

ABSTRACT.--TheCosta Rican Sharpbill (Oxyruncuscristatus frater) is a chunky bird of stolid demeanor,with rapid and jerky movements.Males apparentlyform explodedleks of 3-4 birds, advertisingtheir lek territoriesby singing a high, thin, wiry, descendingtrill from conspicuousperches high in the canopy of precipitous,mid-elevation rain forest. The singing (and probablebreeding) season extends from late Februaryor March to late May or early June. The bright vermilion crest of the males is erected during intense aggressive interactionsbut not during singing. Probablecourtship is described,but we were unable to observecopulation. We alsodescribe a probableflight display of unknown significance. Sharpbillsemploy varied foragingtactics and take both fruit and animal prey. The sharp- pointed bill is used to pry open rolled leaves, undehisced fruits, and tufts of moss and epiphytes. The Sharpbills'bill and foraging appear to represent a unique specialization within the great tyrannine suboscineradiation of the New World tropics and a striking convergencewith the family Icteridae in particular.Received 18 January1982, accepted31 August 1982.

THESharpbill (Oxyruncus cristatus) is a wide- E. Eisenmann (cited in Ridgely 1976) ascribes ranging but extremely poorly known subos- to O. c. cristatusof "a long-drawn high cine, generally placed in its own monotypic thin whistle that gradually slides down in family. Five apparentlydisjunct races have been pitch." Also, we have recently receiveda short described over its broad distribution from Costa accountby T. H. Davis of the behavior of O. Rica to Paraguay,with the peculiaritythat the c. hypoglaucaof Suriname (in lift.), which will northernmost and southernmost forms are more be discussed below in connection with our ob- similar to each other than either is to the in- servations of O. c. frater. tervening forms (Chapman 1939). The north- In this paper we describethe song, singing ernmost race, O. c. frater Sclaterand Salvin, behavior, and probable socialsystem of O. c. has been recorded from Bijagua, in the Cor- frater, includinga flight displayand a possible dillera de Guanacaste of northern Costa Rica, courtshipvocalization. We also present obser- south to Veraguas,Panama, principally on the vations on foragingand other behaviors,some Caribbean slope but with one specimeneach of which appear to shed light on the species' from San Jos• and Santa Maria de Dota, on the chief morphological peculiarity, its sharp- Pacific slope (Slud 1964, Wetmore 1972). From pointed bill. Our observationsof Sharpbills specimen recordsand his two observationsof were made in Parque Nacional Braulio Carrillo, the bird in Costa Rica, Slud (1964) considered Provincia de San Jos•, Costa Rica, chiefly in it a bird of heavily forestedregions in the cool the vicinity of "La Montura," a house beside and very humid portionsof the subtropicalbelt, the road (under construction) from San Jos• to where it occurredin the canopyand in tall trees Guapiles, in the Atlantic lowlands. Except for at the forest border. He noted it foraging in a narrow strip of devastationalong the road dense foliage at medium heights in the trees, itself, the area is heavily forested;the terrain clinging upside-down while feeding on small is exceedinglyprecipitous (see Fig. 3). Most berries. We know of no other substantivepub- observations were made at an elevation of about lished accountsof O. c. frater in life, and there 1,100m, near the top of a steep-sidedridge that are very few of any of the otherraces. Wetmore separatesthe Rio Patria watershed on the west (1972) describesthe body proportions, ptery- from that of the Rio La Hondura to the east. lography, and stomachcontents of two O. c. The general area falls in the PremontaneRain brooksi he collected in Darien, Panama, while Forest life zone of Holdridge (1967). Between 117 TheAuk 100:117-125. January 1983 118 $T•LESAND WHITNEY [Auk, Vol. 100 late June 1980and early April 1982,Stiles made a total of 18 visits of 1-5 days' duration to La Montura, accumulatingsome 18 h of observa- tions on singingSharpbills plus numerousob- servationson foraging and other behaviors. Whitney watchedsinging Sharpbills for about 8 h on 17-18 March and 4 April 1982 at La Mon- tura and briefly observedother behaviors there and elsewhere on the same dates.

GENERAL APPEARANCE AND BEHAVIOR

In life, Sharpbillsappear chunky and short- legged,about the bulk of a kingbird (Tyrannus) but with a relativelysmaller head and bill and a shortertail (Fig. la). The upperpartsare olive Fig.1. Somepostures of theSharpbill, drawn from green, with two yellowish wingbars; the un- 35-mm slides.a. Normal perchingposture. b. and c. derpartsare pale yellow, the breastand sides Singing male. heavily spottedwith black. The feathersof the head and throat are flecked or scaled with blackish. Those of the sides of the crown are to the edge of the canopy.Even when perched tipped black (more broadly in males) and ob- in plain sight, Sharpbills are often difficult to scure a bright erectile median crownpatch, observe, owing to their inconspicuouscolors which in the male is bright vermilion, having and sluggishdemeanor. The Sharpbill'sstolid more and longerfeathers than the paler orange manner was most evident on one of the few patch of the female and forming a conspicuous occasionswhen Stiles saw a male's crest (part- crest when fully erected. The iris of singing ly) raised.This occurredwhen a maleflew from males appearsreddish, the bill and feet grey; his song perch to join a furious aggregationof there may be a sexualdifference in iris color, ,honeycreepers, warblers, and hum- becausea bird that visited a song perch of a mingbirds that were respondingby vigorous male and was tolerated by him in the vicinity mobbing behavior to Stiles' whistled imita- had an orangeiris and was almostcertainly a tions of the call of the Highland Pygmy- female (see below). The other notable sexual (Glaucidiumjardinii). Flying rapidly into a near- difference involves the outermost (tenth) pri- by tree, the Sharpbill hopped swiftly up to a mary: in males, but not females,the barbs of perch in densefoliage and sat for severalmin- the basal two-thirds of the outer web are stiff- utes peering incessantlyabout, partly raising enedand recurved,giving a saw-toothedeffect and lowering its crest severaltimes. Its slug- not unlike that of the outermostprimary of male gishnesscontrasted strikingly with the active Stelgidopteryxswallows. Notwithstanding these flitting of the other birds, its silencewith their differences, the sexes are too similar in ap- excited scolding. pearanceto be safelydistinguished in the field Sharpbills are solitary birds under most cir- under any but the most favorable circum- cumstances.Singing males only exceptionally stances. tolerateother individuals, probablyfemales (see Sharpbills usually perch erectly and rather below), near their song perches.We have nev- stolidly,often sitting in the samespot for min- er seen any lasting associationbetween two utes at a time (althoughlooking alertly about). Sharpbillsnor any indication that the singing When they do move, their movements are areas of males are also nesting areas. In our abrupt and jerky, with heavy hops and jumps experience,foraging Sharpbills are also seen in foliage or fast, direct flights betweentrees. singly,although they may associateloosely with The normal flight consistsof bursts of fast, mixed-speciesflocks, particularly those canopy shallowwingbeats alternating with short level flocks organized around parties of tanagers glides. The birds usually alight in dense fo- (Chlorospingusophthalmicus and/or Tangara liage, where they may sit quietly or progress spp.). Sharpbills seem to join flocks that re- by shorthops and flightsupward and outward main for some time in the same spot rather January1983] SharpbillBehavior 119

2

1

2;0 $œ½0ND$ Fig. 2. Sonagramof the songof the Sharpbill,recorded 22 March1981 on a Uher4000-L recorder. Son- agramprepared by J. W. Hardy.

than moving with the flock for considerable Monturais from late Februaryor early March periods.Davis (in litt.) reportsthat Sharpbills to late May or early June.In 1981,Stiles heard in Surinameregularly join and follow mixed- no Sharpbillssinging on 8-9 February,but they speciesforaging flocks. were in full songby 13-15 March. In 1982, he heardonly very sporadicsinging on 17-19Feb- SONG AND SINGING BEHAVIOR ruary,but by 17-18March Whitney found them in full song.Song was still frequenton 22-25 The most conspicuousvocalization of the May 1981but was not heardon 1-5 July(nor Sharpbillat La Monturais a high,thin, wiry, on 29 June-4July 1980). The singingseason is descendingtrill soundinglike "eeeeuuuurrr," almostcertainly synonymous with the breed- with a rathercicada-like quality. The sonagram ing season,as nearlyall smallcanopy-dwelling (Fig. 2) shows the trill to have a duration of birds (excepthummingbirds) at La Montura about2.5 s; it commencesat a frequencyof breed during these months (Stilesunpubl. about 3 kHz, descendsto about 2.5 kHz, and data). hasa fine structureof about65 syllables/s.Vo- In both1981 and 1982, the number of singing calizingbirds usually perch on deador thinly males at La Montura varied from 3 to 4 on dif- foliagedbranches projecting conspicuously out ferentdates; their territories were arranged in of the canopy,about 15-30 m aboveground a looseduster on both sidesof the ridgetop (Fig. 3). Eachbird seemsto use a smallnumber (Fig.4). Mostof our observationsof songand (ca. 3-5) of suchperches regularly, within a social behavior were of the male whose terri- radiusof 50-100m; the callingareas of adjacent tory includedthe ridgetopitself, hereafter called birdsdo not overlap,and in factthe birdsap- the "ridgetopmale" (althoughin the absence pearto defendthem (seebelow). The only vo- of banded birds we cannot be certain that the calizingbirds that we couldobserve dosely def- sameindividual was involvedin both years). initely had the vermilion crown-patchof the The arrangement of territories varied some- male Sharpbill; becausethe behavior of all call- whatbetween years, although many of thesame ing birds seemedsimilar, we concludethat all song percheswere used in both (evidently suchbirds were probablymales. The loudness, sometimesby different males). In 1982, a dead stereotypy,and persistentdelivery from con- treetopsome 40 m northeastand slightly below spicuous, defended perches of this vocaliza- the housewas a regular song perch of the tion convince us that it representsthe territo- ridgetopmale and provedexceptionally favor- rial advertisingsong of the species.A similar ablefor observation;most of the significantso- vocalizationwas heardfrom SurinameSharp- cial interactionsseen involved this perch and bills by Davis (in litt.), who reacheda similar its vicinity. (In 1981 this tree was alive, bore conclusionregarding its function. thickfoliage, and was not used as a songperch.) The singingseason of the Sharpbillat La No other singingSharpbills were locatedin the 120 STILESAND WHITNEY [Auk, Vol. 100

Fig. 3. Forestat La Montura, includingtwo songperches of the ridgetopmale (1981):the dead branch in the extremeupper left of the large (ca. 25-m) tree at center,and the uppermostdiagonal branch of the smaller(ca. 17-m),thinly foliagedtree at the extremeright. Phototaken looking southwest across the road into the Rio Patria drainage;note the steepnessof the terrain and the quantity of mossand epiphytesin the large tree at center.

La Montura area, despite repeatedvisual and was neverexposed during singing;indeed, the auditory searchesalong severalkilometers of feathersoften appearedso tightly appressedas roadside and study trails. to form a concavityin the centerof the crown. The singingposture of the perchedbird was During periodsof activesinging, the ridgetop characteristic: it stretched its neck forward, male gave an averageof about two songsper fluffing its head feathers and opening its bill minute (interval between songs15-58 s, mean wide (Fig. lb, c). The vermilion crown-patch 31 s for 17 successivesongs at about 1300 on January1983] SharpbillBehavior 121

.... 1981

Fig. 4. Approximate locationsof territories and song perchesof Sharpbillsat La Montura in 1981 a•'d 1982. Contour interval about 10 m; note extreme steepnessof terrain.

21 March 1981).A typical songbout contained perchesduring thesehours and foragedlargely about 10 songs(mean 10.7, range 3-21 for 11 within sight of them. Even at these times, bouts on 21 March and 16 April 1981);gener- however, he was often absent for 5-15 min at ally, the ridgetop male changedperches one or a stretch. A briefer and less intense period of more times during a songbout (mean number singing usually occurredat 0540-0700;during of consecutivesongs from a perch 3.8, range the rest of the day, singingwas highly sporad- 1-9 for 8 bouts on 16 April 1981). About 10% ic. Periods of heavy rain often disrupted this of the songswere given from perchesother than schedule;during the very rainy observation the regular song perches(in the case of the period of April 1981, the birds did not sing at ridgetop male, many songs were given from all on some days. two Clusiafeeding trees immediatelybehind The only instancein which markedly differ- the house). Often the ridgetop male was an- ent singingbehavior was observedoccurred at swered by, or sang in responseto, other sing- about 1530 on 4 April 1982, some 8 km north ing males on either side of the ridge. During of La Montura, at an elevation of about 700 m. the height of the singing season,the Sharpbills Whitney observeda lone Sharpbill perched on sang most consistently around midday (ca. a small dead limb near the top of a tall (ca. 20 1000-1400).During clearweather, the ridgetop m) tree at the edge of a clearing. This bird's male spent 40-80% of his time on his song song seemedslightly higher-pitched,thinner, 122 $T•LESAtaV Wn•TtaE¾ [Auk,Vol. 100

and shorterthan the songsof the La Montura flown eastdown the hill toward anothersing- birds but was easily recognizableas that of a ing male, Stiles saw two Sharpbills burst out Sharpbill.After giving a song,the bird sat si- of a tree some 75 m below the road, locked in lently on its perchfor 3-4 min, then suddenly combat.The fluttering,grappling birds dropped launchedinto a strange,rapid, flutteryflight about 10 m and separated; one flew off to the in which the wings quivered while the tail east,and the other,evidently the ridgetopmale, pumped sharply downward as the bird flew flew to a tree just below the road and perched, directlyacross the clearing.While in mid-flight, his crown-patchfully erectedinto a flaming the bird jerkedits head up slightly,opened its crest.He graduallylowered his crestand after bill wide, delivered one song, and continued a minute or so began to sing while hopping flying until lost from view behind a tree, after about in densefoliage. Finally, he flew to one which it couldnot be relocated.The songgiven of his regularsong perches and sangfor several in flight sounded exactly like that which the minutes, during which time he was answered bird had just given from a perch. No possible severaltimes by the bird to the east. On three mechanically produced sounds were heard other occasions,Stiles saw another Sharpbill during this peculiar flight, so the function of fly in and land near the song perch of the the serrationsof the male's outer primary re- ridgetop male, who without further ado mains unknown. We have never seen such a launched himselt at the intruder and chased it flight displayat La Montura and cannotdivine out of sight. Twice he remained out of contact its significanceat present. Whitney found no for several minutes, but after one such incident other Sharpbills in this area, nor had either on 18 February 1982 he returned within 30 s Stilesor Whitney seenor heard any in several with crestpartly raised,but quicklylowered it visits during the 1981 singing season.Unfor- and sang.All of theseinteractions were entire- tunately, a landslideblocked access to this area ly silent. in May 1982. We observed two possible instances of courtship.During the morning and early after- ACCRESSIVE AND COURTSHIP BEHAVIOR noon of 18 March 1982, Whitney frequently noted the presence of a secondSharpbill for- We rarely saw a male on his songperch in- agingin the treesadjacent to the ridgetopmale's teractingclosely with anotherSharpbill and es- songperch below the house. This secondbird timate that the ridgetop male on his song appearedto be ignored on most occasions,but percheswas quite alone for at least 95-98% of once the male flew to join the other bird in a the time. On several occasions Stiles saw this tree 15 m away. Here the two birds perched male give what appearedto be a stylizeddou- very closetogether on a mossylimb some 12 ble wing-stretch,snapping his wingshalf open m above the ground and out of Whitney's di- horizontally,holding the posefor a secondor rect line of sight. At least one of the birds re- so, then snappingthem shut.Such movements peatedly gave a soft, descendingchatter of 6- occurredmore often during a bout of singing 10 syllablesthat reminded Whitney of a vocal- back and forth with another male, but Stiles ization of Leptopogonamaurocephalus. After was unable to determine whether or not the several repetitions of the chatter, one bird left maleswere in visual contact.Certainly no oth- the tree and flew out of sight,followed 2 s later er Sharpbill was detected close to the male's by the other. Followingthis, the ridgetop male song perch on such occasions,and we do not did not return to his songperch for about 10 know the function (if any) of this apparent dis- min. On 5 April 1982, Stilestwice saw a Sharp- play. bill fly in and land on the ridgetopmale's song Nevertheless,two types of socialinteractions perch below the house while the male was ab- centeredabout the males' song percheswere sentnear midday. This individual differed from observed: aggressive interactions, evidently the ridgetopmale in that its iris was orangish involving two males, and what appearedto be and it appearedto have lessfluffy crown feath- femalevisits leading to courtship.At leastsome ers; it perchedsilently while looking actively of the departuresof the ridgetop male from his about. On the first occasion, this bird flew off territory were evidentlyfor the purposeof in- after about 5 min without the male having put teractingwith other males. At about 1130 on in an appearance.On the secondoccasion, at 21 March 1981, some 2 min after this male had 1245, the ridgetop male flew directly in and January1983] SharpbillBehavior 123

supplantedthis bird on the songperch but did sumablymales) separated by distancesof 1-2 not chase it. The second bird flew into a tree km and alsoconcluded that they were exploded some20 m away and appearedto forage.After leks. singing twice, the male flew into the same tree FORAGING BEHAVIOR and perched in dense foliage, apparently close to the other bird but out of sight. Stiles heard We observed several kinds of foraging be- severalbursts of soft, high-pitched, thin, de- havior by Sharpbills, both for fruits and for scendingchatters in the next 15-20s (evidently prey. On various occasionswe the samevocalization heard by Whitney); then watched a Sharpbill hop deliberately along a the birds apparentlyleft the tree heading east moss-ladenbranch well up in the canopyand or northeastout of Stiles' line of sight. Some 5 probe with the bill into tufts of moss or small min later the ridgetop male flew into a Clusia epiphytes,sometimes prying smallcushions of tree behind the house, approachingfrom the moss, etc. away from the branch to peer be- east;after foragingbriefly, he flew to his song neath them. The birds also sometimesclung to perch and sang.During the next 1.5 h, no other the sidesof branchesto peer beneath, like big, Sharpbill was seen in his territory. The male's slow-moving tanagers (Tangara, etc.), again apparenttolerance of thesebirds near his song apparently poking and probing into the moss perchesand the occurrenceof a distinct type tufts on the branch undersides.Occasionally, of vocalization when the two birds were in close Stiles has seen Sharpbills make short, rather contactstrongly suggest that thesevisitors were clumsy looking sallies, presumablyto pluck females and that the interaction observed was prey from nearbyfoliage; such sally-feeding is courtship. Unfortunately, both of the chatter- common in some South American populations ing interactionsoccurred in foliage too dense of the species(M.P. Fogden pers. comm.) On for us to seeprecisely what occurred,although 3 October 1981 Stiles, T. Moermond, and J. the interactioncertainly did not appear to be Denslow watched a Sharpbill foragefor over 10 aggressive. min. This bird was deliberately inserting its The nature of theseinteractions, plus the lack fine-pointed bill into the tips and edges of of any long-term association between two tightly rolled young Clusialeaves and appar- Sharpbills,leads us to concludethat the cluster ently openingup the leavesby gapingwith the of male territoriesfunctions as an explodedlek. bill, oftenhanging acrobatically in the process. Aggressive interactions around song perches Once it gobbled up somethingthat evidently between lek males in Phaethornissuperciliosus had been hidden in the rolled leaf. The bird and other lek species(Stiles and Wolf 1979 and also practiced this technique on rolled dead includedreferences) are strikinglysimilar to the leaves as we watched. On 20 December 1981 aggressionseen in Sharpbills. The only non- Stileswatched a similar bout of foragingat the aggressiveinteractions between Sharpbillsseen rolledyoung leavesof a stranglerfig (Ficus)well were the presumed courtshipsdescribed above, up in the canopy. The spiders and their egg which resemble male-female interactions seen casesfound by Wetmore in Sharpbill stomachs on the explodedleks of someCotingidae (Snow could well have been procuredby the poking 1972,1977). No sign of nestingactivity was ever and prying techniquesdescribed here. seen in the males' territories, and the fact that Sharpbillswere alsoobserved taking fruit on the maleswere apparentlycontinuously pres- numerousoccasions. Sometimes they took small ent throughout the supposedbreeding season berries (Ericaceae, Loranthaceae, etc.) as de- argues against any lasting pair bond being scribed by Slud (1964). The most frequent formed: in those tyrannine suboscineswhere fruit-foraging we saw, however, was at trees a pair bond is formed, the malesmay not build bearing arillate fruits, especiallyHampea ap- the nest or incubatebut they do help to feed pendiculata (Tiliaceae) and Clusia oblanceolata the young (Skutch 1960, 1969). Although the (Guttiferae). The ridgetop male regularly for- ridgetop male's territory contained several aged at one Hampea and three Clusia trees fruiting trees (more in 1981 than in 1982), we within 100m of his favorite songperch. In each saw no overt defense of the trees per se, and case,the bird hoppedswiftly but heavilyin the our data strongly suggestthat much feeding foliage,grasping the twig and hangingupside- occurred off territory. In Suriname, Davis (in down while inserting the point of its bill into litt.) found clustersof 3-4 singing birds (pre- the pods or husks of fruits that were just be- 124 STILESAND WI-IITNEY [Auk, Vol. 100

ginning to dehisce and in which the aril was members of the Cotingidae in their general not yet fully exposed.Prying open the pod, the body proportions and singing behavior and bird would swallow one or more of the arillate probably in their social system(cf. Snow 1970, seeds before leaving the fruit. The seeds are 1972, 1977; Wetmore 1972). Their spotted evidently regurgitated later, as Stiles found plumageis highly unusual but approachedby severalof the black, shiny Hampeaseeds under some other cotingidslike Pipreolaand Laniiso- the male's favorite song perch. A number of ma. On the other hand, the concealed reddish otherbirds eagerly ea• the arillateseeds of both crown-patch,modified primary, and flight dis- Hampea and Clusia, including toucanets, play somewhat recall certain tyrannids. More thrushes, tanagers,and flycatchers,such that information is required to evaluateother known when we observedthem, thesetrees had very or suspectedanatomical peculiarities of Sharp- few open fruits with uneatenseeds at any one bills properly, such as pterylography,jaw and time. The ability of the Sharpbillto exploitthese leg muscles,etc. The nesting habits of Sharp- fruits beforethey open fully may well give it bills remain undescribed, and knowledge of an advantagein potentialcompetition for them. their behavior is still very fragmentary.For the It would thus appear that the Sharpbill's present, it seemssafest to continue to recog- sharp bill serves the function of a wedge in nize the Oxyruncidae as a monotypic family prying open rolled leaves and dehiscing fruit until more data become available. and perhaps in prying up tufts of moss, etc. on branches.Strikingly similar pry-and-gapefor- ACKNOWLEDGMENTS aging behavior has been well documented in Stiles thanks R. G. Campos, I. Chac6n, C. G6mez, various similarly "sharp-billed" membersof the and V. Zeled6n for help in the field. J. W. Hardy family Icteridae (e.g. Skutch 1954) and in Ver- preparedthe sonagramof Oxyruncus,for which the mivora chrysopteraand V. peregrinaof the Pa- Cornell Laboratorysupplied the tapes and the West- rulidae (Slud 1964, Morton 1980, Tramer and ern Foundationof VertebrateZoology the recorder. Kemp 1980). In the forest canopy at La Mon- The Servicio de Parques Nacionales de Costa Rica, tura, severalother speciesforage in somewhat especiallyB. Madriz, J. M. Cartin, and G. Flores, similar ways, but we have seennone that du- providedthe necessarypermits and logisticalassis- plicatesthe range of pry-and-gapetactics em- tance for work in Parque Nacional Braulio Carrillo. ployed by Sharpbills. The icterids Cacicusuro- For financial assistance Stiles thanks CONICIT and pygialisand Zarhynchuswagleri regularly pry the Vicerrectoriade Investigaci6nof the Universidad de Costa Rica. Whitney's observationswere sup- into moss and epiphytes on branches and (Ca- ported in part by Victor Emanuel Nature Tours, cicus)rolled dead leaves, especiallyof palms, Inc. V. Zeled6n helped with typing the manuscript. but we have not seen them take fruit. The bar- We thankJ. Fitzpatrick,G. V. N. Powell,M.P. Fog- bet Eubuccobourcierii regularly probes rolled den, T. Moermond, and J. Denslow for helpful dis- dead leaves (cf. Slud 1964), but we have never cussionsand commentary.Finally, we appreciatethe seen it open rolled young leaves of Clusia or courtesyof T. H. Davis in sending us his unpub- Ficusor dehiscing fruits. The same applies to lished material on Oxyruncusin Suriname. the much smaller Vermivora warblers, which

are probablyunable to handle the tough, leath- LITERATURE CITED ery Clusia leaves or take fruits with seeds as large as those of Hampea in any case. C•APMAN, F. M. 1939. The riddle of Oxyruncus. Amer. Mus. Nat. Hist. Novitates No. 1047.

CONCLUSIONS REGARDING SHARPBILL HOLDRIDGE,L.R. 1967. Life zone ecology.San Jos•, Costa Rica, Tropical Sci. Ctr. RELATIONSHIPS MORTON,E.S. 1980. Adaptationto seasonalchanges The Sharpbill's sharp bill and associated by migrantbirds in the PanamfiCanal Zone. Pp. foraging behavior would appear to represent 437456 in Migrant birds in the Neotropics(A. Keastand E. S. Morton, Eds.).Washington, D.C., a unique specialization in the great cotin- Smithsonian Inst. Press. ga-manakin-flycatcher radiation of the Neo- RIDGEL¾,R. S. 1976. A guide to the birds of Pan- tropics. Other aspectsof its morphology and amfi. Princeton, New Jersey,Princeton Univ. biology are perhapsless unusual among its pu- Press. tative relativesin the Cotingidaeand - SKUTCH, A.F. 1954. Life histories of Central Amer- dae. Sharpbills resemble certain of the larger ican birds, vol. I. Pacific Coast Avifauna No. 31. January1983] SharpbillBehavior 125

1960. Life histories of Central American ST•LES,F. G., & L. L. WOLF. 1979. The ecologyand birds, vol. II. Pacific Coast Avifauna No. 34. evolution of behavior in the Long- 1969. Life histories of Central American tailed Hermit hummingbird. Ornithol. Monogr. birds, vol. III. Pacific Coast Avifauna No. 35. No. 27. SLUD, P. 1964. The birds of Costa Rica: distribu- TRAMER,E. J., & T. R. KEMP. 1980. Foraging ecol- tion and ecology. Bull. Amer. Mus. Nat. Hist. ogy of migrantand residentwarblers and vireos 128. in the highlands of Costa Rica. Pp. 285-296 in SNow, B. K. 1970. A field study of the Bearded Migrant birds in the Neotropics (A. Keast and Bellbird in Trinidad. Ibis 112: 299-329. E. S. Morton, Eds.). Washington,D.C., Smith- 1972. A study of the Calfbird Perissosc- sonJan Inst. Press. phalus tricolor. Ibis 114: 139-162. WETMORE,A. 1972. The birds of the Republic of 1977. Territorial behavior and courtship of Panama, vol. 3. Smithsonian Misc. Coil. No. 150. the male Three-wattled Bellbird. Auk 94: 623- 645.

The FrankM. ChapmanMemorial Fund givesgrants in aid of ornithologicalresearch and alsopostdoctoral fellowships.While thereis no restrictionon who may apply, the Committeeparticularly welcomes and favorsapplications from graduatestudents; projects in game managementand the medicalsciences are seldomfunded. Applications for projectsin 1983should be submittednot later than 15 February;pros- pectiveapplicants and advisorsshould note there will shortlybe a changein the Chapmanmeeting and deadlineschedule. Application forms may be obtainedfrom the FrankM. ChapmanMemorial Fund Com- mittee, The American Museum of Natural History, Central Park West at 79th St., New York, New York 10024. Chapmangrants during 1982, totalling $40,453 with alcyon);Carlos A. Delannoy, Breedingbiology and a mean of $499, were awarded to: Jonathan L. At- ecologyof the PuertoRican Sharp-shinned Hawk wood, Speciation in the Black-tailed Gnatcatcher (Accipiter striatus venator); Kim Craig Derrickson, (Polioptilamelanura) complex; James C. Bednarz, Analysisof the Mockingbird (Mimuspolyglottos) vo- Cooperativepolyandry in theHarris' Hawk; CraigW. cal repertoire for behavioral and situational corre- Benkman,Food availability, foraging efficiency, and latesand seasonaltrends; David F. DeSante,Stability the regulation of crossbills(Loxia) in eastern North and dynamicsof a subalpinebreeding bird commu- America;Thomas K. Bicak,Food resources and forag- nity; Robert JackDowsett, Conservationplan for the ing behaviorof Long-billedCurlews (Numenius ameri- relict evergreenforests of Malawi and investigation canus) in western Idaho; David Edward Blockstein, of bird populationstherein; Paul J. DuBowy, Optimal Reproductivebehavior and parental investment in foraging, resource partitioning, and community the Mourning Dove, Zenaidamacroura; Sharon Ann structure of western North American Anatini; Patrick Brady, Effect of habitat size on the breeding and J. Dugan, Investigation of the information-center wintering ecology of the (Seiurusauro- hypothesisin tree-nestingherons; Bonita C. Eliason, capillus);Gregory S. Butcher,Sexual differences in the Mating system,parental care, and individual repro- color and behavior of the Northern Oriole; William ductive strategiesin the Blackpoll Warbler, Den- J. Carmen, Juvenile dispersal, flocking behavior, droica striata; Keith William Emmerson, Bird com- and habitat use in the CaliforniaScrub Jay (Aphelo- munity of the laurel forest of Tenerife (Canary Is- comacoerulescens californica); John H. Carothers,For- lands);Robert C. Fleischer,Host choiceby individual agingefficiencies in a nectar-feedingguild of Hawai- female Brown-headed Cowbirds; Frank B. Gill and ian honeycreepersat a commonfood source;Ralph Douglas Wechsler, Evolution of avian mating sys- V. Cartar,Incubation behavior of the White-rumped tems: breeding biology of a promiscuoustropical Sandpiper (Calidrisfuscicollis); Michael D. Carter, flycatcher,Mionectes oleagina; Steven M. Goodman, Socialorganization and parasitichabits of breeding Avifaunal survey of the central Egyptian eastern Bronzed Cowbirds (Molothrus aeneus); Christine desert; Kathleen Diane Groschupf, Song repertoires Copenhaver,Experimental analysis of decisionmak- and singing behavior of Rufous-wingedand Cassin's ing in :the effectof resourcedistribu- sparrows:functional significance of diverse singing tion on territory defense;Robert L. Curry, Evolution strategiesin Aimophila sparrows; Lise A. Hanners, and ecologyof communalbreeding in Galapagos Development of social behavior in Laughing Gull Mockingbirds;William JamesDavis, Significanceof (Larusatricilla) chicks;Linda Heald, Behavioralplas- vocalizationsin the Belted Kingfisher (Megaceryle ticity in a Tyrannid flycatcher:effects of environmen-

(continuedon p. 138) EVIDENCE FOR A POLYPHYLETIC ORIGIN OF THE

STORRS L. OLSON NationalMuseum of Natural History, SmithsonianInstitution, Washington,D.C. 20560 USA

ABSTRACT.--Despitetwo recent anatomical studies to the contrary, the order Piciformes appears to be polyphyletic. The structure of the zygodactyl foot in the Galbulae is very distinct from that in the , and no unique sharedderived charactersof the tarsometatarsus have been demonstratedfor thesetwo taxa. The supposedlythree-headed origin of M. flexor hallucis longus sharedby the Galbulae and Pici is doubtfully homologousbetween the two groups, leaving only the Type VI deep flexor tendonsas defining the order Piciformes.This condition is probably a convergentsimilarity. Evidence is presentedsupporting a close relationship between the Galbulae and the suborder Coracii and between the Pici and the Passeriformes.There are fewer characterconflicts with this hypothesisthan with the hy- pothesis that the Piciformes are monophyletic. Problems concerning fossil taxa are also addressed.Received 24 September1981, accepted15 May 1982.

A MONOPHYLETICorigin of the Piciformes with outsidegroups in a manner indicating that appears to have gained support from the si- the zygodactyl condition in ,, multaneous appearance of two cladistic, ana- and Piciformes had arisen independently, tomicalpapers (Swierczewski and Raikow 1981, through convergence. Simpson and Cracraft 1981) that concur in the Although I certainly do not advocate a traditional conceptof the order--a conceptthat monophyleticorigin of zygodactylbirds, the has prevailed at least since the time of Gadow argumentsthat Simpson and Cracraft(1981) and (1893). I depart from this view in considering Swierczewskiand Raikow (1981)present against each of the two major subdivisionsof the Pic- such a hypothesis do not meet the require- iformes, the Galbulae (Bucconidae, Galbuli- ments of their cladisticmethodology. Simpson dae) and the Pici (Capitonidae, Ramphastidae, and Cracraft(1981: 484) concludeonly that "the Indicatoridae, Picidae), to be more closely re- relationships of cuckoos and parrots remain lated to another group than to each other. My among the most enigmaticwithin ornithology purpose here is (1) to show that the evidence .... "although "there is a generalacceptance for monophyly is weak, uncorroborated,and among avian systematists. . . that piciforms has in part been misrepresentedby Simpson are most closely related to coraciiformsor to and Cracraft (1981), and (2) to make prelimi- passeriformsand that cuckoosand parrots are nary suggestionsas to the probableclosest rel- not." They concedethat "this hypothesishas atives of the Galbulae and the Pici. yet to be tested cladistically "Swier- czewski and Raikow (1981: 469) state that: "The

THE WEAKNESS OF THE EVIDENCE FOR muscular component of the foot mechanism is PICIFORM MONOPHYLY quite different in [the Cuculidae,Psittacidae, and Piciformes] ... which supports the con- Zygodactyly.--Obligate zygodactyly, the tention (Bock and Miller 1959: 30) that those condition in which the fourth toe is perma- groups became zygodactyl independently." nently reversedand has an enlargedaccessory Both setsof authors have thus tacitly accepted articulatingprocess (the "sehnenhalter"),oc- differences between taxa as evidence of non- curs in cuckoos (Cuculidae, Cuculiformes), relationship,a procedureof which Cracrafthas parrots (Psittacidae,Psittaciformes), and in the been outspokenlycritical (see Olson 1982). Piciformes. This is obviously a derived con- The most complete and original work on the dition in birds that could be used to define nature of the zygodactyl foot is that of Stein- these taxa as a monophyletic group in a cla- bacher (1935), whose resultshave seldombeen distic sense, unless it were shown that each of accurately represented (a notable exception thesezygodactyl taxa sharesderived characters being Sibley and Ahlquist 1972). Steinbacher

126 The Auk 100: 126-133. January 1983 January1983] PolyphyleticOrigin of thePiciformes 127

A B C D

Fig. 1. Posterior(top row) and lateral (bottom row) views of the distal end of the tarsometatarsusin the four groups of birds with obligate zygodactyly(illustrations from Steinbacher1935). A, ,Centropus ateralbus(Cuculidae, Cuculiformes);B, , Galbularuficauda (Galbulidae, Galbulae, "Piciformes"); C, , Ramphastostoco (Ramphastidae,Pici, Piciformes);D, , Amazonaochrocephala (Psittacidae, Psittaciformes).Abbreviations: Sh = sehnenhalter,C II = trochleafor digit II, C III= trochleafor digit III, C IV = trochleafor digit IV, G IV = articulatingsurface for digit IV, Gsh = articulatingsurface of sehnen- halter, M 16 = depressionfor origin of M. extensorbrevis digiti IV, R = groovefor tendon of M. extensor brevisdigiti IV, M 4 = depressionfor tendonof M. flexorperforatus digiti IV. It canbe seenthat the form of zygodactylyin the Galbulae is completelydifferent from that in the Pici, and the two can in no way be regardedas homologous. The bestinterpretation of the evidencefrom the tarsometatarsusis that zygodactyly evolved independently in all four of these groups. showed that there were four distinct types of viding evidence "that osteological... char- morphologyof the tarsometatarsusin birds with acteristicsof zygodactyly are distinct for the obligatezygodactyly, with that in the Galbulae piciformsand different from cuckoosand par- being as different from that in the Pici as either rots," Simpsonand Cracraft(1981: 484) have of thesetwo is from parrotsor cuckoos(Fig. 1). dearly misrepresentedthe factsand Steinbach- In eachof thesefour groupsthere is a sehnen- er's interpretation of them. halter. Steinbacher (1935: 234) even identified Nowhere do Simpson and Cracraft, nor any a sehnenhalterin (Strigiformes),which other authors(e.g. Bockand Miller 1959),show are facultatively zygodactyl. Thus, the state- that there are derived characters of the tarso- ment by Simpsonand Cracraft(1981: 485) that metatarsusthat will distinguish the Galbulae "zygodactyly and the presenceof a sehnen- and Pici from parrotsand cuckoosand that will halter can be interpreted as derived characters establish the Piciformes as a monophyletic definingthe piciforms as monophyletic" is dis- group. In fact, the apparentlyless modified ingenuous.In citing Steinbacher(1935) as pro- trochlea IV and sehnenhalter in the Galbulae 128 $To•s L. O•,soN [Auk, Vol. 100

are actually more similar to the condition in expanded so that fibers also originated from cuckoos,whereas the larger,more discrete,and the proximal end of the fibula and from the distally projecting sehnenhalterin the Pici is median raphe of the adjacentM. flexor perfor- more like that in parrots(Fig. 1). There simply atus digiti III. [Incidentally, Fig. 6d in Simpson is no evidence in the structure of the tarso- and Cracraft (1981) is mislabelled--"Fpp3" metatarsus that will demonstrate a close rela- shouldread "Fp3."] In Trachyphonusthe fibers tionship between the Galbulae and the Pici. originatingon the fibula, and thoseoriginating Swierczewski and Raikow (1981) present six on the raphe of M. flexorperforatus digiti III, myological charactersto define their Clade B are slightly separatedfrom the main belly of (= Galbulae) and six additional myological the musclethat originatesin the popliteal fos- charactersto define their CladeG (= Pici). Thus, sa, thus making the muscle three-headed. these two taxa differ from each other in at least Clearly, M. flexor hallucis longus, in contrib- 12 myological charactersof the hind limb, as uting to the flexion of two toes in addition to well as having a completelydifferent structure the hallux, has become strengthenedby ex- of the tarsometatarsus associated with their re- panding the area of its origin to the two nearest spective forms of zygodactyly. May we not, available structures.This is directly correlated then, apply the same statement that Swier- with the Type VI arrangementof the flexorten- czewski and Raikow used against monophyly donsand is part of the samecharacter complex. of all zygodactylbirds to argue againstmono- I could not detect any separateheads of or- phyly of the Piciformes,namely that "the mus- igin in the bucconid ,however. In- cular componentof the foot systemis quite dif- deed, Swierczewski(1977: 57) statesthat in the ferent in thosegroups," a differencesupporting Galbulidae and Bucconidae the heads are the contentionthat they "becamezygodactyl "somewhatdifficult to separatefrom each oth- independently" (Swierczewski and Raikow er." He alsonotes that the "commonbelly ex- 1981: 469)? tends only about two-thirds the length of the Origin of M. flexor hallucislongus.--Swier- tibiotarsus" in the Galbulae, versus almost the czewski and Raikow state that M. flexor hal- entire lengthof the tibiotarsusin the Pici. Thus, lucis longus has three heads in the Piciformes, it seems far from certain that the nature of the which they interpretas a derivedcondition that origin of this muscle is homologousbetween supports monophyly. Simpson and Cracraft the Galbulae and the Pici or even that it can (1981:483) imply this conditionto be unique really be said to have three heads in the Gal- to the Piciformesby stating that "the muscle bulae. arises by one or two heads in other birds," Type VI flexor tendons.•adow (1893) de- whereas Swierczewski and Raikow discuss the fined the Piciformesby their possessionof the fact that M. flexorhallucis longus also has three Type VI configurationof the deep flexor ten- heads in most passerines.Because the lateral dons,whereby M. flexorhallucis longus, which head has a different relationshipto the tendon ordinarily has a direct tendinous connection of M. iliofibularis in Passeriformes, Swier- only with the hallux (digit I), flexes digits II czewski and Raikow (1981: 473) consider that and IV as well, and M. flexordigitorum longus the "condition in the two orders is therefore flexesonly digit III. This conditionmust have probably not homologous." At this point I arisen when digit IV was reversedand began would also question whether the condition in to function as a secondhallux (zygodactyly). the Galbulae and Pici has been established as That parrotsand cuckoosare zygodactyl but do being homologous. not have the Type VI flexortendons is evidence Before I was able to consult Swierczewski's that different evolutionarypathways can pro- (1977) unpublished dissertation for details, I duce similar functional results. dissected one specimen each of the barbet The Type VI tendon arrangementwas used Trachyphonusdarnaudii (Capitonidae: Pici) and originallyto define the Piciformes,and it is still the Hypnelus bicinctus(Bucconidae: the only characterthat canbe citedto unify the Galbulae) in order to assessthe configuration order. Given that the condition of origin of M. of M. flexor hallucis in each. In the normal avi- flexor hallucis longus is part of the same com- an condition, this muscle originates in the plex and is doubtfully homologousin the Gal- popliteal fossa of the femur. In the specimen bulae and Pici anyway, then it may fairly be of Hypnelusthat I examined,the origin was said that tte new studies of osteology and January19831 PolyphyleticOrigin o[ thePici[orraes 129

A

G

D E F

Fig.2. Lateraland ventral views of skullsof a roller,Coracias benghalensis (A,D); a puffbird, panamensis(B,E); and a barbet,Trachyphonus purpuratus (C,F). Note the greatoverall similarity between the roller() and the puffbird (Bucconidae), whereas neither dosely resembles the barbet (Capitonidae). Thearrows indicate the ventral extent of thepostorbital process, which is greatlydeveloped in the Coracii and Galbulae. Not to scale. myology of the Piciformeshave failed to reveal czewskiand Raikow(1981) amply demonstrate. a singlenew characterthat independentlycor- Of the 43 myologicalcharacters presented in roboratesthe hypothesisof monophyly. their Table1, 40% evolvedmore than oncejust If we consider that the Piciformes are not within the Piciformes.With the probability of monophyletic,then the Type VI conditionof convergencebeing so high, it is dearly im- the flexor tendons must have arisen more than practicalto justify an entire order of birds with once.This is not at all an unreasonablehy- a single myologicalcharacter. In another situ- pothesis.Convergence in myologicalcharac- ation, Bertnan and Raikow (1982: 55) found that ters occurswith great frequency, as Swier- only colies(Coliidae) and parrotshave a branch 130 STORRSL. OLSON [Auk, Vol. 100

of the M. extensordigitorum longus tendon Coracias (Coraciidae) in almost every aspect extending to the hallux; yet, they considered (Fig. 2)--bill shape, truncatepalatines, straight that "this character alone cannot demonstrate and narrow pterygoids, heavily ossifiednasal a common ancestry for the Coliiformes and septum, shape and position of the temporal Psittaciformes.... "Why, then, shoulda sin- fossae,the inflated ectethmoidplate, and the gle tendinal characterbe acceptedas demon- greatly exaggerated and ventrally produced strating the common ancestry of the Pici- post-orbital process.In all of these characters formes? the Bucconidaeare consistentlydifferent from the Pici. THE RELATIONSHIPS OF THE GALBULAE The major differencesbetween the skulls of AND THE ]VICI the Bucconidae and Coracias are in the reduced Here, I will briefly outline my reasons for lacrimal and the dorsal expansion of the pala- believing the Galbulaeto be closelyrelated to tines onto the parasphenoid rostrum in the the rollers, or Coracii of Maurer and Rai- Bucconidae. These differences are more like kow (1981), which includes the Coraciidae, thoseobserved between genera within a family Brachypteraciidae,and Leptosomidae.The Pici, than between different orders. They are less on the other hand, I believe are more closely profound than the differences observed be- related to the Passeriformes. These observa- tween the four families of Pici, for example. In tions are preliminary; a more completeassess- many respects,the skull of Coraciasdiffers less ment of the interrelationshipsof all the higher from that of the bucconidMalacoptila than it orders of land birds requires more evidence does from Eurystomus,the only other modern than is now available. For example, it is diffi- genus in the Coraciidae. cult to make myological comparisonsof the In the Galbulae, and in all of the Coracii, the Piciformesand Coraciiformesfrom the existing postorbitalprocess is greatlyenlarged and ex- literature because Swierczewski and Raikow tends straight ventrally as far as the jugal bar (1981)and Maurer and Raikow (1981)have pre- (Fig. 2). From this there is a very strong,short, sented only the evidence that supports their post-orbital ligament that attachesto a process classifications,while omitting the descriptive on the medial surfaceof the mandible just an- observations of actual dissections. Even with terior to the articulation (Fig. 3). The M. ad- access to the unpublished dissertations of ductor mandibulae complexis correspondingly Swierczewski (3.977) and Maurer (1977), cross narrowed, enabling it to passthrough the rel- comparisons are difficult, because a character atively small foramen formed by the enlarged that was deemed important in one group was postorbitalprocess (Fig. 3). On the other hand, often not consideredsignificant in the other, the postorbitalprocess is quite smallin the Pici, so that certaindescriptions may be inadequate most Passeriformes,the Trogonidae,and in the for comparison,necessitating the re-examina- remainder of the Coraciiformesexcept the Bu- tion of specimens. cerotidae and some of the Momotidae. In the Sibley and Ahlquist (1972) have previously last two instances, the postorbital process is suggesteda relationshipbetween the Galbulae well developedbut does not extend nearly as and the , but they particularly far ventrally as it does in the Coracii and the singled out the kingfishers (Alcedinidae) as Galbulae. In the one example of barbet possibleaffines. The kingfishers,however, be- (Trachyphonusdarnaudii) that I dissected, the long to the alcedinine group of Coraciiformes postorbitalligament was very weak and was that is characterizedby a derived morphology scarcelydifferentiated from the overlying fas- of the stapes (Feduccia 1975) and several de- ciae. rived myological characters(Maurer and Rai- In the elementsof the postcranialskeleton of kow 1981) that do not occur in the Galbulae. A the Galbulae, there is greater similarity to the lack of relationship between the Galbulae and Coracii than to the Pici. The coracoid in the the Alcedinidae does not, however, precludea Galbulae is almost identical to that in the roller relationship between the Galbulae and some group (Fig. 4) and is very different from that other section of the Coraciiformes. found in the Pici. The humeri and carpometa- The skull and mandible in the Galbulae, par- carpi in the Galbulae are also more similar to ticularly in the less-specializedfamily Buccon- those in the rollers than to the Pici (see below). idae, show a remarkable similarity to those of Those who have experience identifying lanuary1983] PolyphyleticOrigin of thePiciformes 131

pometacarpus in the Pici differs from that in m the Galbulaeand Coraciiformesin havinga broad and very well-developedintermetacar- pal tubercle,a conditionshared only with the Passedfformesamong the higher land birds. The humerus in the Pici differs from that in the Galbulae and Coraciiformes and resembles that in the Passedfformesin having the shaftshort and stout, the proximal end broader, and the deltoidcrest squared, rather than triangular or m rounded. Whereas we have seen that no characters in- dependentof the deep flexortendons could be foundto corroboratepiciform monophyly, most B othercharacters, whatever their "polarity" may be, are concordant when the Galbulae are al- lied with the Coracii and the Pici are allied P with the Passeriformes.Thus, the oil glandis covered with down and lacks a tufted orifice in the Galbulae and Coracii but is nude, with a tufted orifice, in the Pici and Passedfformes (Gadow1896). The caecaare well developed in G the Gabulae and Coracii but absent or rudi- mentary in the Pici and Passedformes(Gadow Fig. 3. Dissectedheads of a roller,Coracias gar- 1896; pers. obs. for Galbulae). The structure of rulus(A); a puffbird, pectoralis (B); and a the down in the Galbulae is like that of the barbet,Trachyphonus darnaudii (C). Note the marked Coracii, whereas that of the Pici is similar to similaritybetween the rollerand the puffbird in the that of the Passedfformes(Chandler 1916). Of verystrong ligament (1) from the enlarged postorbital process to the mandible and in the narrowed adduc- the six derived myologicalcharacters that tor mandibulaecomplex (m), whereasin the barbet Swierczewskiand Raikowuse to definethe Pici, thepostorbital process (p) is very small, the ligament four (characters1, 33, 39, and 41) are found in is vestigial,and the adductor mandibulae complex is most, some, or all passedfies.Mm. popliteus, of morenormal development. Not to scale. adductordigiti II, andextensor brevis digiti IV are absent in the Pici (characters33, 39, and 41) and in Passeriformes,whereas eachof these musclesis presentin both the Galbulae and the isolated bird bones are aware of the skeletal Coracii.If my interpretationof Maurer's(1977) similarities between the Pici and the Passeri- descriptionsis correct,four characters(15, 29, formes;a possiblerelationship between these 31, and 32) of the six that Swierczewski and two groupshas long beenrecognized (see Sib- Raikow (1981) use to define the Galbulae as ley and Ahlquist1972). Lowe (1946) considered monophyleticalso occur in the Coracii, where- the Pici to be but a suborder of the Passedf- as only two (14 and 42) appearto be unique formes and did not include or even mention (autapomorphous).When all of the data can be the Galbulae. analyzed, I am confident that there will be far The coracoif in the Pici and Passeriformes is fewer character conflicts when the Piciformes very slenderand elongate,with the sternalend are split apart, as I have proposedhere, than andhead narrow, the sterno-coracoidal process whenthey are maintained as monophyletic. reduced,and the procoracoidprocess usually vestigialor evenabsent (Fig. 4). This contrasts with the condition in the Galbulae and most COMMENTS ON FOSSIL TAXA Coraciiformes, in which the coracoif is com- paratively short, the head and sternal end Simpson and Cracra[t's (1981: 491) discus- expanded,and the sterno-coracoidaland pro- sionso[ the Primobucconidaeand Zygodacty- coracoif processeswell developed.The car- lidae are misleading.Their statementthat the 132 STORRSL. OLSON [Auk,Vol. 100

A B C D

Fig. 4. Ventral view of coracoidsto show the similarities between the ground roller Brachypteracias leptosornus(A), and a puffbird, Malacoptilapanarnensis (B). Thesediffer greatlyfrom the coracoidin the Pici (C, the ,Colaptes auratus), which is morelike that in passerines(D, a rhinocryptid,Pteroptochos rnegapodius).Not to scale.

Primobucconidae have not been shown to be tors that were not zygodactyl, and forms tran- a monophyletic group is irrelevant. Such a sitional between those ancestorsand the fully statement could truthfully be made about the zygodactyl modem Galbulae must have exist- vast majority of groups of birds or other or- ed. The Galbulae would have to be included at ganisms. Simpson and Cracraft present no in- some taxonomic level with birds that did not formation to suggestthat the primobucconids have the fourth trochlea as modified as do the are not monophyletic. Their contention that modem members of the suborder. Because "the bucconidsand galbulids are fully zygo- Simpson and Cracraft present no evidence to dactyl," but that the "Eocene genera [of pri- showthat the primobucconidsare moreclosely mobucconids]are not," hinges entirely on the related to some other group, there is no reason definition of "fully zygodactyl,"which they do not to follow Brodkorb (1970) and Feduccia and not provide. Two specimens of primobuccon- Martin (1976)in consideringthe Primobuccon- ids (the holotype of Neanis kistneri, Feduccia idae to be primitive members of the Galbulae 1973; and the holotype of Primobuccoolsoni, that share more similarities with the Bucconi- Feduccia and Martin 1976) are preserved with dae than with any otherextant family of birds. the outer toe completelyreversed, in the zygo- Simpsonand Cracraft(1981: 492) tentatively dactyl manner. Perhaps by "fully zygodactyl" suggest"placing the Zygodactylidaeas a basal Simpson and Cracraftmean that the primobuc- member of the Pici," but they presentno evi- conids do not have the fourth trochlea and seh- dence for this either. Ballmann (1969a, b) de- henhalter as modified as in other zygodactyl liberately did not put Zygodactylusin any ex- birds, but the primobucconidswere certainly isting order, because he considered that its functionallyzygodactyl. [Note also that Simp- affinities could not be determined from the tar- son and Cracraft (1981:491) misquote Feduccia sometatarsusand tibiotarsus, the only ele- and Martin (1976)--the description of the pri- ments yet known. For descriptivepurposes he mobucconid tarsometatarsus should read "a made comparisons not only with the Pici- distinct grooveseparating the posteriorportion formes but also with the Psittaciformes. Ball- of the trochlea" not "supporting" it.] mann (pers. comm.) has emphasizedverbally Simpson and Cracraft contendthat the more to me his belief that Zygodactylusis not pici- primitive tarsal morphology of the primobuc- form. If Zygodactylustells us anythingat this conids arguesagainst their being placed with- point, it is probably that the specializedzygo- in the Galbulae, but this follows only if one dactyl condition of the tarsometatarsus, in defines the Galbulae solely by the possession which the fourth trochleabecomes enlarged and of a completely modified fourth trochlea. At bears a sehnenhalter,has evolved yet another somepoint, the Galbulae obviouslyhad ances- time. January1983] PolyphyleticOrigin of the Piciformes 133

CONCLUSION CHANDLER,A. C. 1916. A study of the structureof feathers, with reference to their taxonomic sig- As I have indicated elsewhere (Olson 1981), nificance. Univ. California Publ. Zool. 13: 243- the higher level systematicsof birds has a very 446. poor foundation. The questionsof whether or FEDUCCIA,A. 1973. A new Eocenezygodactyl bird. not currently recognizedorders are monophy- J. Paleontol. 47: 501-503. letic and what the interrelationshipsof these 1975. Morphologyof the bony stapes(col- ordersmay be are still largelyunanswered. Al- umella) in the Passeriformesand related groups: though the studies of Swierczewski and Rai- evolutionary implications. Univ. Kansas Mus. Nat. Hist. Misc. Publ. 3: 1-34. kow (1981) and Maurer and Raikow (1981) are , • L. D. MARTIN. 1976. The Eocene zygo- useful in documentingthe monophylyof some dactyl birds of North America (Aves: Pici- of the subunits of Coraciiformes and Pici- formes). Smithsonian Contr. Zool. 27: 101-110. formes, at the level of ordinal and interordinal GADOW,H. 1893. Vogel. II. SystematischerTheil. systematicsthey are less successful.Despite Pp. 259-270in Klassenund Ordnungendes Thier- these workers' accumulation of much new data, Reichs, vol. 6(4) (H. G. Bronn, Ed.). Leipzig, C. they could recognizethe ordersPiciformes and F. Winter. Coraciiformes as monophyletic only by the ß 1896. Caeca, pp. 68-70; Oil-gland, pp. 653- configuration of the deep flexor tendons--the 654 in A dictionary of birds (A. Newton). Lon- same character that had been used to define don, Adam & Charles Black. these orders nine decades ago. Their studies LowE, P. R. 1946. On the systematicposition of the (Pici), honey-guides (Indica- were designed only so as to test hypotheses tor), hoopoes, and others. Ibis 88: 103-127. that had previouslybeen formulated,whereas MAURER,D. g. 1977. The appendicularmyology they seem to lack the means to generate alter- and relationships of the avian order Coracii- native hypotheses.Herewith, I have supplied formes. Unpublished Ph.D. dissertation, Pitts- one for the Piciformes, and there is every rea- burgh, Pennsylvania,Univ. Pittsburgh. son to expect alternative hypotheses of rela- , & g. J. RAIKOW. 1981. Appendicular myol- tionships for the Coraciiformesand for other ogy, phylogeny, and classificationof the avian orders based on single charactersor that are order Coraciiformes(including Trogoniformes). otherwisepoorly defined. Ann. Carnegie Mus. Nat. Hist. 50: 417-434. OriSON, S. L. 1981. The museum tradition in or-

ACKNOWLEDGMENTS nithology--a responseto Ricklefs.Auk 98: 193- 195. I am grateful to Helen F. James,David W. Stead- ß 1982. A critique of Cracraft's classification man, Richard P. Vari, and Richard L. Zusi for read- of birds. Auk 99: 733-739. ing and commenting on the manuscript. I am like- SIBLEY,C. G., & J. A. AHLQUIST. 1972. A compar- wise indebted to Victor E. Krantz for the photog- ative study of the egg-white proteins of non- raphy. passerinebirds. Peabody Mus. Nat. Hist. Yale Univ. Bullß 39: 1-276. LITERATURE CITED SIMPSON,S. F., & J. CRACRAFT.1981. The phylo- genetic relationships of the Piciformes (Class BALLMANN, P. 1969a. Les oiseaux mioc•nes de La Aves). Auk 98: 481-494. Grive-Saint-Alban (Is•re). Geobios 2: 157-204. STEINBACHER, Gß 1935. Funktionell-anatomische 1969b. Die V6gel aus der altburdigalen Untersuchungenan Vogelffissenmit Wendezeh- SpaltenffJllungvon Wintershof (West) bei Eich- en und Rfickzehen. J. Ornithol. 83: 214-282. st•tt in B'ayern.Zitteliana 1: 5-60. SWIERCZEWSKI,E.V. 1977ß The hindlimb myology BERMAI•,S. L., & g. J. RAIKOW. 1982. The hindlimb and phylogeneticrelationships of the avian or- musculature of the (Coliiformes). Auk 99: 41-57. der Piciformes. Unpublished Ph.D. dissertation, BOCK, W., & W. DEW. MILLER. 1959. The scan- Pittsburgh,Pennsylvania, Univ. Pittsburgh. , & g. J. RAIKOW. 1981. Hind limb mor- soffal foot of the woodpeckers, with comments phology,phylogeny, and classificationof the Pi- on the evolution of perchingand climbing feet ciformes. Auk 98: 466-480. in birds. Amer. Mus. Novitates 1931: 1-45. BRODKORB,P. 1970. An Eocenepuffbird from Wy- oming. Univ. Wyoming Contr. Geol. 9: 13-15. MONOPHYLY OF THE PICIFORMES: A REPLY TO OLSON

ROBERTJ. RAIKOW1 AND JOELCRACRAFT 2 •Departmentof BiologicalSciences, University of Pittsburgh,Pittsburgh, Pennsylvania 15260 USA and Carnegie Museumof Natural History, Pittsburgh,Pennsylvania 15213 USA, and 2Departmentof ,University of Illinoisat the MedicalCenter, Chicago, illinois 60680 USA and FieldMuseum of Natural History,Chicago, illinois 60605 USA

ABSTP•CT.--Olson(1983) questioned the hypothesisthat the order Piciformesis mono- phyletic and suggestedinstead that each piciform suborder is allied with a nonpiciform group. His attempt to discreditthe synapomorphiesjoining the Galbulae and Pici is refuted by correctionsof his interpretationsof previous work. The Piciformes share a complex derived hindlimb morphologyinvolving zygodactyly,a tripartite flexorhallucis longus, and Type VI flexortendons. Olson's argument for polyphylycombines inadequate data with the inappropriatetechnique of equatinggeneral overall similarity with affinity.His hypothesis is thereforerejected. Problems concerning fossil taxa are also discussed.Based on current information,we believethat a monophyleticorigin of the Piciformesremains the hypothesis of choice.Received 18 August1982, accepted18 October1982.

SWIERCZEWSKIand Raikow (1981)and Simp- more primitive than that of the Pici, which is son and Cracraft (1981) studied the phyloge- more "specialized," and thereforethat mono- netic relationshipsof the Piciformes.Both con- phyly requiresthe conditionin the Pici to have cluded that the order is monophyleticand that evolved from that in the Galbulae. Here, he it contains two monophyletic suborders, the confusessister-group relationship with ances- Galbulae (Bucconidae and Galbulidae) and Pici tor-descendentrelationship. We did not pro- (Capitonidae, Ramphastidae, Indicatoridae, pose that the Pici evolved from the Galbulae and Picidae). Olson (1983) criticizesthe hy- (higher taxa cannotbe ancestors).Olson asks pothesisof monophyly. First, he questionsthe "... why would such a transformationtake argumentssupporting this view. He then pre- place?Once a group of birds has becomeper- sents an alternative hypothesis that the Gal- manently zygodactyl,is it possibleto become bulae are most closely related to the Coracii more zygodactyl?"This question manifestsa and the Pici to the Passeriformes.Finally, he confusion. One group is not "more zygodac- criticizes Simpson and Cracraft'sdiscussion of tyl" than the other;both are describedby this certain fossil birds. After carefullyexamining term. It may be expected,however, that dif- Olson's critique, we believe that monophyly ferenceswill accumulatein separatelyevolving remains the more strongly supportedhypoth- lineagesafter they have split; it is this process esis. The possiblevalue of these discussions of evolutionary divergence or charactertrans- goesbeyond the questionof a singlebranching formation that gives rise to the hierarchical point in a phylogeny,as it providesan oppor- structureof organicdiversity. We suggestthat tunity to compareand contrasttwo profoundly the zygodactylconditions of the Galbulae and different approachesto systematicanalysis. Pici are homologousbecause other characters (see below) corroboratethe unity of the Pici- formes. PICIFORM MONOPHYLY M. flexorhallucis longus.--Swierczewski and Z!/godact!/l!/.--Olsonsuggests that zygodac- Raikow (1981)reported that the flexor hallucis tyly aroseindependently in the Pici and Gal- longusmuscle (FHL) shows a derived condi- bulae becauseSteinbacher (1935) found differ- tion in its origin by a certainpattern of three ences in their tarsometatarsi. He does not heads. In most birds there are one or some- consider the possibility that the differences times two femoral heads, but the Piciformes aroseafter the two groupsdiverged from a zy- have as well an extensiveorigin from the fib- godactylouscommon ancestor. Olson consid- ula. Olson dissectedone capitonid (Trachy- ers that the tarsometatarsus of the Galbulae is phonus;Pici) and one bucconid(Hypnelus; Gai- 134 The Auk I00: 134-138.January 1983 January1983] Monophylyof thePiciformes 135

FIlL MED each of two species.In contrast,Swierczewski INT LAT dissectedbetween one and four specimenseach of 45 species, and concluded unequivocally (1977: 58) that "Three heads of origin were foundin allspecies studied herein." Nevertheless, one of us (R.J.R.) has, in re- sponseto Olson'scomments, dissected the limb of Colaptes auratus (Picidae), tamiata (Bucconidae), and Galbula dea (Galbulidae). Like Olson, Raikow confirmed in Colaptesthe three heads previously described(Fig. 1). In Hypnelus, Olson "... could not detect any separateheads of origin . . ." and questioned whether "... the nature of the origin of this muscle is homologousbetween the Galbulae and the Pici or even that it can really be said to have three heads in the Galbulae." The dis- section of Galbula and Bucco has fully con- firmedour previousreport that the threeheads, including the large fibular head, are presentin the pattern described (Fig. 1). This corrobo- ratesthe hypothesisof piciformmonophyly and refutes Olson's claim that the condition is not found in both piciform suborders. Type VI flexor tendons.--Did this condition Fig. 1. Lateralviews of the regionof the knee and upper crus in three speciesof piciform birds to show arise oncein the Piciformesor separatelyin the Galbulae and Pici? Olson states that it is the modeof originof M. flexorhallucis longus (FHL). In all Piciformesthe musclearises by three separate "probably a convergentsimilarity," but offers heads, medial (MED), intermediate (INT), and lateral no evidence. He suggeststhat the Galbulae (LAT), all of which lie medial to the tendon of M. would be difficult to identify with their feet cut iliofibularis (IL. FIB.). The representativeforms il- off, but fortunately all of our specimenspos- lustrated are Colaptesauratus (Picidae; Pici), Bucco sessedfeet. (Actually, the relevant structures tamiata (Bucconidae; Galbulae), and Galbula dea are legs, not feet, as the FHL belly lies in the (Galbulidae;Galbulae). The drawingswere madewith crus.) Olson feels that such characters do not a camera lucida and dissectingmicroscope and are not to scale. justify ordinalrank, but we do not agree that taxonomic rank is related to the corporeal lo- cation of the relevant characters. Here, Olson bulae). In Trachyphonushe confirmedthe report confusesphylogeny with taxonomy;the ques- of Swierczewski and Raikow (1981), but in tion at issue is whether the group is mono- Hypnelus he did not. Olson quotes Swier- phyletic, not what rank it shouldhave in a clas- czewski (1977: 57) as saying that in the Gal- sification. Perhaps the FHL should rank the bulae the headsare "somewhatdifficult to sep- group at the generic level, becauseits origin arate from each other." Swierczewski's full lies at the knee joint (Juncturagenus). statement, however, conveysa different mean- ing: "In the Galbulidae and Bucconidae, the ANALYSIS OF OLSON'S HYPOTHESIS medial head arises semitendinous from the in- tercondyloidregion of the femur, and the in- Inadequatedata.--Olson comparescharacters termedlate head arises fleshy from the proxi- in only some of the relevant groups, so that mocaudal surface of the external femoral comprehensivecomparisons cannot be made. condyle;the headsbeing somewhatdifficult to For example, he notes that "in plumage pat- separatefrom each other." This statement re- tern, the ground roller Brachypteraciaslepto- fers to only two of the three heads;it doesnot somusis quite similar to certain of the Buccon- refer to the fibular head, which is the signifi- idae such as Malacoptila." But what are the cant structure. Olson dissected one specimen patterns in the other Coracii, the other Buc- 136 RAIKOWA•r) CRACRAFX [Auk, Vol. 100 conidae, and the Galbulidae?Why are they not ic similarity. He claimsthat we acceptthe idea discussed?Convergence in plumagepatterns is that "... differences between taxa [are] evi- common, which is why comprehensivecom- dence of nonrelationship .... "but this is in- parisons are needed to distinguish between correct.We mean only that the absenceof sim- synapomorphiesand spuriousresemblances. ilarities fails to corroboratea hypothesis of Again, in discussingthe skull Olson argues monophyly, not that the presenceof differ- for similarities between the Galbulae and Cor- encesrefutes such a hypothesis."Nonrelation- acii, but restricts his comparison almost en- ship" has no intrinsic meaning;we are search- tirely to the Bucconidaeand Coraciaswhen list- ing for patternsof commonancestry identified ing severalfeatures. "In all of thesecharacters," by synapomorphy. he notes, "the Bucconidaeare consistentlydif- The problem with phenetic comparisonsis ferent from the Pici." But what about the Gal- that charactersare not analyzed so as to iden- bulidae and other Coracii? tify the level within the phylogenetichierarchy This patternof selective,almost casual choice at which they define taxa. To illustrate this of data characterizes Olson's entire presenta- point, we will consider one example from a tion. It is also difficult to assess his data, be- recentstudy of the relationshipsof Pedionomus cause they are presented in a narrative form torquatus.This speciesis placed in the mono- within the text but are not tabulated in a way typic family Pedionomidae, and the problem that would allow one to determine the state of is to find its closest relatives. Olson and Stead- each character in all the relevant taxa. A tab- man (1981) rejected the previous hypotheses ulation of this sort would have revealed that that it is related to the Turnicidae () broad conclusions are often based on little data; or the Galliformes and concluded that it be- e.g. Olson dissectedonly one barbet to deter- longs in the . Olson and mine the form of the postorbital ligament in Steadman(1981: 3) note that a hallux is present the large and diverse suborderPici. in Pedionomus and most Charadriiformes but Incorrectdetermination of polarity.--Although is absent in the Turnicidae. They consider (p. generallyeschewing cladistic methods, Olson 21) that this supportsthe removal of Pediono- does make one attempt to show the derived mus "from the vicinity of the Turnicidae and nature of a condition, the complexof the post- its placement in the Charadriiformes." Actual- orbital process,postorbital ligament, and ad- ly, it does nothing of the kind. The hallux ductor mandibulae. He considers this derived evolved in vertebrates as part of the transfor- becauseit "... does not occurin Archaeopter- mation of the pelvic fin into a limb. As such, yx, in presumablyprimitive land birds suchas it is a derived character at the level of the Te- Opisthocomusand the Cuculiformes,or else- trapoda. Inasmuch as birds form a tetrapod where in the higher groups of land birds." subgroup, the presenceof a hallux is a primi- Several problems invalidate this conclusion. tive state and reveals nothing about the rela- First, there is no reason to assume that a char- tionships of any avian speciesto any other. It acter is primitive becauseit occursin "primi- tells us onething only about Pedionomus,name- tive" taxa. All taxa, including fossil forms, are ly that it is a member of the Tetrapoda. It def- mosaicsof primitive and derived characters. initely refutes any hypothesisthat Pedionomus Second,while we concedethe general primi- is afish, but that is all it does. tivenessof ,we questionthe ex- This example illustrates the principle that tent to which the ligament and musclecan be charactercomparisons convey maximum infor- reconstructed in it. Furthermore, on what basis mation about phylogenetic relationships only are the Cuculiformes and Opisthocomus"pre- when one determines the specific points at sumablyprimitive"? What are "higher groups" which they are relevant within the nested sys- of land birds? Why is the comparisonlimited tem of cladesforming a phylogeny.Simplistic to the nonmonophyleticnontaxon of landbirds? phenetic comparisonsfail to provide this in- Olson's analysisfails to show that the condi- formation. Olson'sassessments of "similarity" tion is derived, or even that it is generallychar- cannot be interpreted, becausewe do not know acteristicof the groupsinvolved. which of them represent shared primitive Use of pheneticsimilarity.--The basic prob- charactersand which derived for any group of lem with Olson's study is that he tries to de- species.The necessityfor suchanalysis may be termine phylogeneticrelationships by phenet- emphasized by pointing out the distinction January1983] Monophylyof thePiciformes 137

between similarity and phylogeneticrelation- ical unit. Presentmorphological evidence sug- ship, namely that they do not necessarilyco- geststhat the Primobucconidaeare a grade,not incide. When an evolving lineage splits, one a clade.If so, then the family hasno ontological daughter lineage may evolve faster than the status,and Olson should insteadbe defending other, so that a form may be less similar to a the usefulnessof discussingthe evolutionof a genealogicallycloser relative than to a more fictitious taxon. distant one. Lungfishes are more similar to Simpson and Cracraft (1981:491) were very goldfishesthan they areto goldfinches,but they specific in their criticism; none of the taxa of are more closelyrelated to goldfinchesthan to the Primobucconidae is known to have an en- goldfishes.Crocodiles are more similar to tur- largedsehnenhalter. For that reason,they con- tles than to turtle doves, but they are more cludedthat there is no justificationfor saying closely related to turtle doves than to turtles. that the primobucconidsare piciforms as cur- Such statements are based on the idea that re- rently accepted,let alone closelyrelated to the lationship is defined by recency of common Bucconidae. That conclusion, if true, also fal- ancestry and are correct within the context of sifies Olson'sunsupported speculations about specificphylogenetic hypotheses, in this case the independentorigin of zygodactylyin the those of Wiley (1979). Genealogicalrelation- Galbulae.From our standpoint,the latter group ships are postulatedby the recognitionof pat- was already zygodactylouswhen it arose, be- ternsof monophyly,which are definedby syn- causethe origin of zygodactylytook place at a apomorphy. For extended discussions of this higher hierarchicallevel. This is why Simpson principle, see Eldredge and Cracraft (1980), and Cracraft (1981:491) suggestedthat one or Wiley (1981), and Raikow (1982). more taxa currently included in the Primo- bucconidaemight be the sister group of the DISCUSSION Or Fossil TAXA Piciformes. Olson'scomments about the Zygodactylidae The sectionof Olson's critique "Comments containfurther inaccuracies.Simpson and Cra~ on Fossil Taxa" adds nothing new about the craft (1981: 492) did provide evidence for in- systematicsof the taxa placed in the Primo- cluding this fossiltaxon in the Pici, namely an bucconidaeand Zygodactylidaebut doesillus- trate the limitations of his systematicmeth- advancedform of the sehnenhalter,but they also stressedthe very tentative nature of this odology. Simpson and Cracraft (1981) pointed placement.Furthermore, Simpson and Cracraft out that no evidenceexists to support the mon- did not voice any major disagreementwith ophyly of the Primobucconidae;Olson in con- trast, believes that this observation is "irrele- Ballman's(1969a, b) interpretations,as implied by Olson. Their disagreementwas with the vant." But inasmuch as the "primobucconids" placement of the Zygodactylidaein the Gal- have played a pivotal role in the speculations bulae (Brodkorb1971). Simpson and Cracraft of Olson and his colleaguesabout the history of the North American bird fauna, it does not specificallystated that Ballmanmay be correct seem "irrelevant" to us to ask whether the Pri- in believing that Zygodactylusis not a pici- form. mobucconidaehave any objective reality as a natural group. Indeed, if the primobucconids CONCLUSIONS are not monophyletic, then Olson and his col- leaguesare constructingevolutionary scenarios Rather than reply to every individual point based upon an imaginary taxon. As a conse- raised by Olson, we have instead addressed quence, one might expect that Olson would the basic differencesin systematicphilosophy want to demonstratethat monophyly. His cri- that exemplifycurrent controversy in biology. tique, however, is a weak attempt to defend Olson's attempt to question the synapomor- the conclusions of Feduccia and Martin (1976) phies linking the Galbulae and Pici stands re- and lacksany relevant empirical evidence. futed; the Piciformes are characterizedby a Olson's attempt to shift the burden of proof complexderived morphology of the hind limb onto Simpson and Cracraft is altogether spu- involving zygodactyly,tripartite flexor hallucis rious. We suggestthat the burden of proof lies longus, and Type VI tendons.Olson supports with workers who erect a taxon without posi- his hypothesiswith a potpourriof casualphe- tive evidence for its existenceas a geneaolog- netic similarities analyzed by inappropriate 138 RAIKOWAND CR•CR•VT [Auk, Vol. 100 methods, and his argument is little more than OLSON,S. L. 1983. Evidence for a polyphyletic or- an opinion. On the basis of present under- igin of the Piciformes.Auk 100:126-133. standing,we concludethat piciform monophy- ß & D. W. STEADMAN. 1981. The relation- ly remains the preferredhypothesis. ships of the Pedionomidae(Aves: Charadri- iformes). Smithsonian Contri. Zool. 337. ACKNOWLEDGMENTS RAIKOW,R. J. 1982. Monophyly of the Passeri- formes:test of a phylogenetichypothesis. Auk We thank Kenneth C. Parkes and D. Scott Wood, 99: 431-445. CarnegieMuseum of Natural History, for providing SIMPSON,S. F., & J. CI•ACI•AFT. 1981. The phylo- specimensused in dissection.Parkes, Wood, and genetic relationships of the Piciformes (Class Mary C. McKitrick providedhelpful criticismsof the Aves). Auk 98: 481-494. manuscript. Our research is supported by N.S.F. STEINBACFIERßG. 1935. Functionell-anatomische grants DEB-8010898(R.J.R.) and DEB-7921492(J.C.). Untersuchungenan Vogelftissenmit Wendzeh-

LITERATURE CITED en und Rticksehen.J. Omithol. 83: 214-282. SWIERCZEWSKI,E. V. 1977. The hindlimb myology BALLMAN, P. 1969a. Les oiseaux mioc•nes de La and phylogeneticrelationships of the avian or- Grive-Saint-Alban (Isbre). Geobios 2: 157-204. der Piciformes.Unpublished Ph.D. dissertation, 1969b. Die V6gel aus der altburdigalen Univ. Pittsburgh,Pittsburghß Pennsylvania. Spaltenffillungvon Wintershof (West) bei Eich- ß & R. J. RAIKOW. 1981. Hind limb mor- st•ittin Bayera. Zitteliana 1: 5-60. phology, phylogeny, and classificationof the BRODKORB,P. 1971. Catalogueof fossil birds: part Piciformes. Auk 98: 466-480. 4 (Columbiformes through Piciformes). Bull. WILEY, E. O. 1979. Ventral gill arch musclesand Florida State Mus. 15: 163-266. the interrelationshipsof Gnathostomes,with a ELDREDGEßN., & J. CRACRAFT.1980. Phylogenetic new classificationof the vertebrata. Zool. J. Lin- patternsand the evolutionaryprocess. New York, naean Soc. 67: 149-179. Columbia Univ. Press. 1981. Phylogenetics:the theoryand prac- FEDUCCIA,A., & L. D. MARTIN. 1976. The Eocene tice of phylogeneticsystematics. New YorkßJohn zygodactylbirds of North America (Aves: Pici- Wiley & Sons. formes). Smithsonian Contri. Zool. 27: 101-110.

(continuedfrom p. 125) tal variability; Gary Richard Hepp, Cost-benefit Ani (Crotophagaani); Michael P. Lombardo,Auxiliary analysisof pair-bond formationin wintering Green- birds at Tree Swallow (Iridoprocnebicolor) nests; winged Teal (Anas creccacarolinensis); Marc Herre- David B. McDonaldß Role of resource distribution mans, Intra- and inter-island variation in the Bul- and behavioralinteractions in a lek systemwith male- buls ( Microscelismadagascariensis/ M. crassirostris) on male cooperation;G. M. O. Maloiy, Environmental Grand Comoroand Moheli; Wendy L. Hillß Behav- physiology of two East African land birds: Kori ioralflexibility of reproductivebehavior in theAmer- Bustardand SecretaryBird; JeffreyS. Marks, Natal ican ; Mark Alan Holmgren, Geographicand philoparryand breeding-areafidelity in Long-eared songvariation in westernField Sparrows;William G. Owls; John L. Maron, Intraspecific competition Hoppes,Avian frugivoryand seeddispersal in north amongwintering Sanderlings;John Marzluff, Corre- temperateforests; Anne E. Houde, Nest site selec- latesof reproductivesuccess in the Pition Jay, Gym- tion by Common and Roseateterns; Laurie Ann norhinuscyanocephalus; Donald B. Miles, Patternsof Hunter, Cooperativebreeding behavior of Purple morphologyand ecologyin a SonoranDesert avian Gallinules; Kristine Johnsonß Mate choice in the community;Patrick J. MockßComparative breeding Pition Jay, Gymnorhinuscyanocephalus; Nancy Joste, ecologyand energeticsof storm-petrelspecies in Baja Electrophoreticanalysis of paternity in the Acorn CaliforniaßMexico; Michael L. Morrison, Geographic Woodpecker;Robert S. KennedyßSystematic review variation in morphologyand singingbehavior of the of Philippine birds for "A Checklist of Philippine Black-throatedGray Warbler in Oregon; CharlesA. Birds"; RichardLee and SusanKay KnightßValue of Munnß Socialorganization and ecologyof canopy flockingby Bald Eaglesin relation to foragingand flocks in Amazonian Peru; Gerald J. Niemi, Ecomor- predation: an experimentalstudy; RachelN. Levinß phologyof breeding birds in peatlandhabitats of Adaptive significanceof antiphonalsong in Thryo- North Americaand northernEurope; Manuel Nores, thorus nigricapillus;Stewart T. Levinson, Bird/fruit Studyof specimensof a new speciesof antbirdfrom interactions in northern Florida; Robert K. Loftin, the east of Brazil; StephenNowicki, Physiologyand Communalnesting behavior of the Smooth-billed physicsof harmonically-complexbird calls;Allan R. (continuedon p. 148) WHY DO BROWN-HEADED COWBIRDS PERFORM THE HEAD-DOWN DISPLAY?

THOMASW. SCOTTl' 2 AND JUDITHM. GRUMSTRUP-SCOTT:•'4 •VeterinaryScience Department, The Pennsylvania State University, University Park, Pennsylvania16802 USA, and :•Departmentof Biology,The Pennsylvania State University, University Park, Pennsylvania16802 USA

ABSTRACT.--Fourhypotheses for the function of the head-down display performed by Brown-headedCowbirds were testedwith observationaldata from free-rangingand captive cowbirds.Free-ranging cowbirds performed 284 interspecificand four intraspecificdisplays during 59.2 daylighthours while roostingin mixed-speciesflocks adjacent to feeding areas. The most commonrecipients of displays,female Red-winged Blackbirds and HouseSpar- rows,preened cowbirds during 25 displays.Cowbirds that had just beenpreened displayed more often than those that had not recently been preened. Captive cowbirds displayed intraspecifically475 times during 13.3 h, and dominant captive birds displayed more often than their subordinates.The following hypothesiswas proposedto explain the display's function: the head-downdisplay of Brown-headedCowbirds is an appeasingagonistic be- havior, the displayor is most often dominant to the recipient, and subsequentdisplaying is stimulatedby interspecificpreening. The display may function in: (a) obtaining food, (b) minimizing roosting energetics,and/or (c) establishingflock order. Received26 November 1979, resubmitted5 January1982, accepted30 August1982.

BROWN-HEADEDCOWBIRDS (Molothrus ater) nations of the display in the form of four hy- perform an unusual head-down display (Dar- pothesesand to evaluatethose hypotheses us- ley 1968)during which they directthe backsof ing data from free-ranging and captive their heads and necks toward other birds (Se- cowbirds. lander and La Rue 1961). The head-down dis- Hypothesis1.--The display is an adaptation play, also called the preening invitation dis- for brood parasitism, which functions in re- play (Selanderand La Rue 1961),is performed ducing host species'aggressive tendencies to- both intra- and interspecifically(Selander and ward a cowbird so that the cowbird can remain La Rue 1961; Stevenson 1969; Rothstein 1977, on the host'sterritory and/or approach the host's 1980;Scott 1977). Interspecificdisplays may in- nest with lessopposition (Selanderand La Rue clude, in addition to the above posture, the 1961). preeningof the displayingcowbird's feathers Hypothesis2.--"The display is an aggres- by the display recipient (Selanderand La Rue sively motivated gesturethat cowbirds use in 1961, Scott 1977). Severalother speciesof cow- a variety of contextsto assessthe fighting po- birds perform a display similar, but not iden- tential of other birds .... to establish domi- tical, to that of the Brown-headed Cowbird nance," or to integrate social units (Rothstein (Selander 1964). Hereafter "display" will refer 1980). to the head-down display and "cowbird" will Hypothesis3.--The display is an exampleof refer to the Brown-headed Cowbird. behavioralmimicry. Cowbirds deceivedisplay The function of the head-down display has recipients: the recipient interprets the display been explained by a number of authors in as appeasing,while the cowbird'sintent is to sometimescontrasting ways. The purpose of threaten (Rothstein 1980). this study, therefore,was to summarizeexpla- Hypothesis4.--The displayis rare in nature, and, because there is a lack of recorded obser- 2 Present address: Yale Arbovirus Research Unit, vations, the display has little "biologicalsig- Department of Epidemiologyand Public Health, Yale nificance" (Dow 1968). UniversitySchool of Medicine,60 CollegeStreet, P.O. Box 3333, New Haven, Connecticut 06510 USA. STUDY AREA AND METHODS 4 Presentaddress: Child Study Center, Yale Uni- versity, 333 Cedar Street, New Haven, Connecticut Free-rangingcowbirds.--The first field study was 06510 USA. conductednear BowlingGreen, Ohio, from 6 March

139 The Auk 100:139-148. January 1983 140 SCOTTAND GRUMSTRuP-SCOTT [Auk, Vol. 100 until 28 April 1976. The site, a rectangularstrip of allowed to habituate to the observation cage for 5 land of approximately3 ha, was bisectedby a paved days. Observationsbegan on the sixth day. road and adjacent to the east bank of the Portage The entire group was observedfor 20 rain, and a River (North Branch) in Wood County. Observations record was kept of the total supplants,retreats, and were recorded from a car parked 10-25 m from the head-down displays for 5 days; the group was then river. That portion of the study site west of the road removedand the succeedinggroup placed in the cage and adjacentto the river was a floodplain dominated to begin its habituationperiod. The order of group by deciduoustrees, while the portion eastof the road observation was randomized. was a cultivatedfield consistingof cornstubble. Descriptionof behaviors.--"Supplanting"is an ag- The secondfield study was conductedin StateCol- onistic behavior that occurswhen one bird displaces lege, Pennsylvania,from 6 Januaryuntil 12 January another (Darley 1968: 20-21). Variation in supplant- 1981. The site included a row of forsythia bushes ing can arise from the speedof approachand the (Forsythiasp.) approximately8 m x 18 m x 2 m. A sequenceor amplitude of behaviors. gravelroad curvedaround the bushesand passed "Retreating"includes "any pattern,walking, hop- within 2 m of their north end. Observations were ping, or flying, used by a submissivebird to leave" recordedfrom a carparked 10-20 m from the bushes. (Darley 1968: 21). When a two-bird interaction re- The area surroundingthe busheswas mowed grass suited in one bird displacingthe other, there was to the east and cultivated fields of corn stubble to the always at least one supplantand one retreat. west. The head-down display was first describedby Se- Behavioral interactions were observed with the aid landerand La Rue (1961).While displaying,the cow- of 7-15 x 35-ram binoculars during 59.2 h of obser- bird's "head is bowed" so that "the bill is directed vation between 0800 and 1830. vertically downward or in toward the cowbird's Captivecowbirds.--A total of 45 cowbirds(24 AHY body." The orientationof the cowbird'sbody to the males, 8 HY males, and 13 females) was maintained recipient can either be directly toward or at an angle in captivity during the winter and spring of 1976. away from the recipient. Commonly, the cowbird HY males(males that hatchedduring 1975)were dis- will display as it approachesthe recipient, ceasing tinguishedfrom AHY males(males that hatched be- its approachwhen it is about 2 cm away from the fore 1975)by an incompletepost-juvenile molt of the recipient. Occasionally,the cowbird's approachcon- humeral tract (Baird 1958, Selander and Giller 1960). tinues until it touchesthe recipientbird's breastwith All 45 birds were caughtin northwesternOhio dur- the top of its head. Like supplanting and retreating, ing the late summerand fall of 1975with modified head-down displays varied in duration, amplitude, crow traps. and speed of the displayor'sapproach. Captive birds were separatedby sex, housed in Data analysis.--Captivebirds were rankedby so- 1.84 m x 2.63 m x 2.31 m holding cagesin a barn cial statusbased on their supplant and retreat records nearBowling Green, and exposedto the naturalpho- (Scott 1977). The bird that supplanted its group toperiod via windows. The observationcage, of the members most often and retreated least was ranked samesize as the holdingcages, had onewindow and as the alpha bird. five perches.Food (50-50% mixtureof crackedcorn Displaysby free-rangingbirds were comparedby and chick starter) and water were constantlyavail- sex, recipient species, and supplants. Displays by able. captivebirds were comparedby socialstatus. In Ta- Canvas covered the north wall of the observation bles 2 and 3 we analyzed"display incidents"as de- cage,providing visual but not auditoryisolation of fined by Rothstein (1977: 17). the study group from the othercaptive birds in the Statisticalanalyses followed methods presented by barn. Observationswere made from an opening in Zar (1974). Sample sizes greater than or equal to 10 a burlap enclosedwalkway so that the study group and less than or equal to 25 were analyzed using a as well as the rest of the birds in the barn could not binomial distribution. Samplesgreater than 25 were see the observer. evaluatedusing a two-tailedX 2 test corrected for con- Captivecowbirds were observedin four different tinuity. A samplesize of lessthan 10 wasconsidered groups of six birds each: females,HY males, AHY too small for meaningfulstatistical analysis. males,and mixedgroups (two females,two HY males, RESULTS and two AHY males). Femalescould not be separated confidentlyby age and thereforewere not divided Free-rangingcowbirds.--None of the recipi- into age groups.Each group type was observedfor ents of displaysby free-ranging cowbirds re- a 5-day period during 11 January-22February and during 26 March-30 April. Observationswere re- sponded by displaying in return to the initia- corded between 0805 and 1838. tor. Recipientsresponded either by retreating, All studygroups were treatedidentically. Birds in peckingat the displayor,supplanting the dis- eachgroup were banded with U.S. Fishand Wildlife playor, or seeming to ignore the displayor. Service bands and unique colored leg bands, then None of the recipients preened conspecifics. January1983] CowbirdHead-down Display 141

TABLE1. Head-down displaysperformed by free-rangingBrown-headed Cowbirds in northwesternOhio and central Pennsylvania.

Number of displays where recipient Sexof dis- Number preened Supplantsby playing cowbird Recipient species of displays cowbird Cowbird Recipient A. Ohio observations(54 h during 33 visits, 0820-1830,6 March-28 April 1976) FEMALE • House Sparrow 6 5 1 c• House Sparrow 2 2 9 Red-winged Blackbird 54 12 21 4 9 Rusty Blackbird 2 1 9 Brown-headed Cowbird 1 1 MALE C•House Sparrow 1 1 9 Red-wingedBlackbird 10 6 OHIO TOTAL 4 species 76 12 36 6 B. Pennsylvaniaobservations (5.2 h during 5 visits, 1125-1610,6 January-12January 1981) FEMALE 5• House Sparrow 23 2 12 c• House Sparrow 30 9 MALE 5• House Sparrow 52 37 c• House Sparrow 104 11 74 ? Brown-headed Cowbird 2 1 c• Brown-headed Cowbird 1 PENNSYLVANIA TOTAL 2 species 212 13 133 7 GRANDTOTAL 4 species 288 25 169 13

Free-ranging cowbirds performed 288 head- display appeared to be more clumped than down displaysduring 38 visits to the study evenly spread out in time. If display rate is sites that totalled 59.2 h of observation (Table considered, based upon the number of min- 1). Cowbirdspresented displays to four species: utes that cowbirds were present, a display was female Red-winged Blackbirds (Agelaiusphoe- observedevery 42.6 min in Ohio and every 1.5 niceus), a female Rusty Blackbird (Euphagus min in Pennsylvania. carolinus),male and female House Sparrows Female cowbirds in Ohio displayed more (Passerdomesticus), and male and female cow- often than males (65 times vs. 11 times, P •< birds (Table 1). 0.001; Table 1), though maximum estimatesof All but two recordeddisplays were observed males (n = 338) and females (n = 333) ob- while cowbirdswere roostingduring daylight servedwere nearly equal. Male cowbirds dis- hours in treesor shrubsin mixed-speciesflocks played more often than femalesin Pennsylva- adjacentto the feeding areasof corn stubble or nia (159 times vs. 53 times, P •< 0.01; Table 1); stockyard.The remaining two displayswere maximum estimates of males (n = 125) and fe- performedby cowbirdsin Pennsylvaniawhile males (n = 46) showed that males predomi- on snow-covered ground. nated. Some displayswere seen in rapid succes- Eighty-six percent of female cowbird dis- sion, separatedby about 1-5 s. The birds per- playsand 91% of malecowbird displays in Ohio forming thesesuccessive displays quickly fol- were directed toward female Red-winged lowed and repeatedly displayed toward the Blackbirds (P •< 0.001 in both cases; Table 1). recipient if the recipient remained in place or Other passerinessuch as male Red-wings, Eu- hopped or flew a short distanceaway (•<1 m). ropean Starlings (Sturnusvulgaris), Common If the recipientbird flew a longerdistance away, Grackles (Quiscalusquiscula), Northern Cardi- the cowbird either displayedtoward another nals (Cardinaliscardinalis), and Song Sparrows bird or stoppeddisplaying altogether.As a re- (Melospizamelodia) were commonly observed sultof thesesuccessive displays, the head-down at the Ohio site and appeared to be available 142 Scott ANDGRUMSTRUP-SCOTT [Auk, Vol. 100

TABLE2. Head-down displaysperformed by captive cowbirdsduring 1976. Each group was observedfor five 20-min periods.Matrices are arrangedwith the alpha bird in the top left comerand progressively subordinatebirds to the right and below the alphabird. Becauseno displayswere observedin the group of six AHY c• (16 April-20 April), no matrix is includedfor that group. Recipientsare listed horizontally, displayorsvertically.

6 AHY c• (16 Feb-20 Feb); mutual displays = 14 6 HY c• (23 Jan-27 Jan);mutual displays = 23 C A E B D F E D B E A C F Y C -- 2 0 i 4 4 11 D -- 0 0 0 0 0 0 A 0 -- 0 0 0 0 0 B 3 -- 0 0 17 5 25 E 0 0 -- 0 0 14 14 E 0 1 -- 3 0 0 4 B 0 0 0 -- 0 0 0 A 0 0 0 -- 0 2 2 D 0 0 0 0 -- 16 16 C 0 0 0 0 -- 0 0 F 2 0 0 0 6 -- 8 F 0 0 0 0 0 -- 0 E = 49 E = 31 5 52(11 Jan-15 Jan); mutual displays = 3 2 AHY d, 2 HY d, 2 52(4 Feb-8 Feb); mutual displays = 83 E C A D B E HYd HYd AHYd AHYd 9 9 E -- 0 2 0 0 2 H G H G G F E C 0 -- i 0 I 2 H -- 1 0 0 0 4 5 A 0 1 -- i 0 2 G 0 -- 0 0 3 16 19 D 0 0 0 -- 0 0 H 0 0 -- 9 0 37 46 B 0 0 0 0 -- 0 G 0 0 0 -- 0 1 1 G 0 0 0 0 -- 13 13 F 0 2 6 0 0 -- 8 E=6 >• = 92 6 HY c• (6 Apr-10 Apr); mutual displays = 7 6 52(26 Apr-30 Apr); mutual displays 48 D H G B C F E G E C D A B E D -- 0 0 2 6 i 9 G -- 0 12 0 0 0 12 H 0 -- 0 0 0 0 0 E 0 -- 11 0 2 0 13 G 0 0 -- 0 0 2 2 C 1 7 -- 1 1 1 11 B 0 0 0 -- 0 0 0 D 0 2 0 -- 1 2 5 C 3 0 0 0 -- 0 3 A 5 0 3 10 -- 6 24 F 0 0 0 0 0 -- 0 B 1 1 1 3 14 -- 20 E= 14 E = 85 AHY c•, 2 HY c•, 2 52(4 Feb-8 Feb); mutual displays= 7 AHYd HYd HYd AHYd 52 52 N F C O C E N -- 0 0 0 0 0 0 F 0 -- 0 0 4 0 4 C 0 0 -- 0 0 i 1 O 0 0 0 -- 0 0 0 C 0 0 0 0 -- 7 7 E 0 0 0 0 1 -- Z = 13

in time and spacefor cowbirds to display to- days of observation;estimates of the number ward if they had chosento do so. of birds presentper visit varied from 18 to more In Pennsylvania,where only cowbirds and than 150 for House Sparrows and from 5 to 75 House Sparrows were observed, House Spar- for cowbirds. Male cowbirds displayed more rows were recipientsof 209 of 212 displays(Ta- often toward male House Sparrows(P •< 0.001) ble 1). Male and female House Sparrows were than toward female House Sparrows(Table 1). present in approximately a 1:1 ratio. House Female Red-wings in Ohio and male and fe- Sparrows outnumbered cowbirds during all male House Sparrowsin Pennsylvaniawere the January1983] CowbirdHead-down Display 143

TABLE3. Summaryof two-tailedX 2 analysisof head- recipients supplanted a displaying cowbird, it down displays performed by dominant versus was done most often by pecking. Eight of 17 subordinatecaptive cowbirds. pecks by recipients toward displaying females Dominance position were associatedwith the retreat of the display- of birds that ing female. performed display Captivecowbirds.--A total of 475 intraspecific most often • displaysby captivebirds was recordedduring March- 13.3 h of observation (Table 2). Head-down Jan-Feb April displays were observed in all groups during Group type, age, obser- obser- both observation periods except AHY males and sexof displaying vation vation from 16 April until 20 April; they were not ob- cowbird period period Total served performing any displays. SegregatedAHY c•c• ------Captive cowbirds, in all groups and during SegregatedHY c•c• top top top both observation periods, exhibited a peck Segregatedc• c• top top top Segregated• • sm• -- -- dominancehierarchy, and femaleswere sub- Segregatedc• c• + 9 9 top -- top ordinate to males (see Scott 1977:25-26 for Mixed c•c• + • • top -- top dominance data). Using the term dominant to Total top -- top refer to the top three birds in the dominance • Top denotes that the top three birds (top two for females Jan-Feb hierarchy (top two for females Jan-Feb segre- segregatedgroup) in the dominancehierarchy performed the display more often (P • 0.05) than subordinates. gatedflock) and subordinateto refer to the bot- • Sample size was <10 and thereforenot tested. tom three birds in the hierarchy, we found that when there were differences in display fre- quencies,dominant birds alwaysperformed the only birds observedresponding to the head- display more often (P • 0.05) than subordi- down display by preening the feathersof the nates (Table 3). Differences in favor of domi- displayingcowbird. Thesebirds preenedmale nant birds were observed more often during and female cowbirds on the cowbirds' capital the first observation period than during the tracts and/or the backs of their necksduring 25 second. of 288 displays(9%; Table 1). Preening seemedto be a strongstimulus for DISCUSSION subsequentdisplaying at both sites, as a cow- bird that had been preened seemed more Hypothesis/.--Our study, as well as other quickly and more persistently to pursue and evidence,supports the appeasementportion of display toward the bird that had preened it. In hypothesis1. We observedrecipients preening Pennsylvania,we testedthis prediction by re- free-ranging cowbirds during 25 of 288 dis- cording displaysperformed by birds that had plays (Table 1). Interspecificpreening was not just been preenedversus thosebirds that had observedin any other context.Similarly, it was not been preened but that had displayed with- not uncommonfor captivebirds to contacteach in the previous5 min. The birds that had been other during a mutual display. Thus, display- preened displayed far more often than those ing birds, both free-rangingand captive, were that had not been preened (P • 0.001); two able to get closeenough to recipientsto make birds that had just been preened displayed 29 physicalcontact for periodsof time up to 2 min, times during the 3.5 min that we could follow somethingthat was not observedin any other them, while three birds that had not been context. preened displayed 33 times during 15.5 min. Similarly, Stevenson(1969) concluded after Displaying cowbirds in both Ohio and observing2,530 displaysby captivecowbirds Pennsylvaniasupplanted recipients more often that the display allows birds to approach re- (P • 0.001) than they retreatedfrom recipients cipientsmore closelythan if they were not dis- (Table 1). Following a supplant, displaying playing. Resultsfrom our study support Ste- cowbirds were often observed to continue their venson's conclusion. approach toward the recipient, which had In addition, Selander and La Rue's (1961) moved a short distanceaway. Frequently,this study showed that during interspecificdis- sequenceof behaviorswas repeated,resulting plays,recipients most often initially responded in a seriesof supplantsby the displayor.When by peckingat the displayingcowbirds but, af- 144 SCOTTAND GRUMSTRUP-SCOTT [Auk, Vol. 100 ter additional exposure to the display, gradu- seasonconcomitantly with increasinghormone ally began to peck less and preen more. Allo- production. Selander and La Rue (1961), how- preening was consideredless aggressivethan ever, did not observe variation in the rate at pecking. Robertsonand Norman (1976)in their which their captive male cowbirds displayed field study with mounted birds also showed during mid-April following bilateral castration that the display appeases;potential hostswere in January. less aggressiveto mounts in the display pos- Displays performed by other speciesof cow- ture than mounts in the "normal" posture. Fi- birds suggest that the display may not have nally, Lowther and Rothstein(1980) suggested evolved in a direct relationship with brood that displaysperformed by cowbirdsless than parasitism. Selander (1964) suggeststhat the 1.5 months of age may function in appease- Bay-winged Cowbird's (Molothrus badius) ment. "preening invitation display" is an evolution- We found no evidence during our field ob- ary precursorto the head-down display of the servationsthat the display and brood parasit- more evolutionarily specializedBrown-headed ism are directly associated.The possibility that Cowbird. The Bay-winged Cowbird is not a it is infrequently performed in that context, (Friedmann 1929). Hence, the however, cannotbe ruled out by this study. display may have evolved before the habit of Displays performed by free-ranging birds and brood parasitism. observed during this study were directed to- Hypothesis2.--Two aspectsof hypothesis2 ward speciesthat are uncommon hosts of the were supportedby this study. Patternsof dis- cowbird (Friedmann 1929, Hicks 1934, Scott playing were observed that support the dis- unpubl. data). Friedmann (1963) and Rothstein play's involvement in integrating social units (1980) also state that the display is seldomdi- (Rothstein 1977). For example, birds in the top rected toward common hosts. half of the dominancehierarchy displayed more Observations and conclusionsby other au- often than birds in the bottom half (P • 0.05; thors similarly fail to support the parasitism Table 3). In addition, when differences in dis- componentof hypothesis1. Mayfield(1961) and play rates existedin comparisonsbetween seg- Friedmann (1963) state that at the time when regated groups and the correspondingage or female cowbirds lay their eggs, small passer- sexclass of mixed groups,birds displayedmore ines are not apt to be present at their nests. often (P • 0.01) when both sexeswere present Hence, female cowbirds in the processof egg than when just one sex was present (Table 2). laying would not commonly interact with po- The only exception to this trend was that fe- tential hosts.Hann (1937,1941), Mayfield (1961), males in the segregatedflock during March- and Norman and Robertson (1975) observed fe- April displayedmore often than femalesin the male cowbirds"inspecting" potential host nests mixed flock. It also appearedfrom the study of before cowbird egg laying took place.No cow- captive birds that a displayormay be assessing bird behavior was reported that could be in- the agonistic tendencies of the recipient, es- terpreted as a head-down display. Though pecially when cowbirds first interact with an Prescott (1947) observed an instance when a unfamiliar individual (Rothstein 1980). This female cowbird approacheda Red-eyed Vireo appeared to be true following the introduction (Vireo olivaceus)sitting on its nest and Hann of House Sparrowsand additional cowbirds to (1937) observed the same with Ovenbirds (Se- the observationcage (Scottunpubl. data). iurusaurocapillus), the cowbirdssupplanted the We do not agree with that portion of hy- vireo and Ovenbirds without performing a be- pothesis2 that describesthe display as a threat. havior that couldbe interpretedas a head-down Rothstein concluded "that the display's true display. In fact, the cowbird observedby Pres- messageis threat" (Rothstein 1980: 157), be- cott supplantedthe vireo with a "distinct peck- cause dominant captive birds display more ing motion." often than their subordinates and because some Frequencyof the display does not appear to displaying captive birds alternatelydisplayed be influencedby hormone productionin males and pecked. We take issue with this conclusion (Selanderand La Rue 1961). Assuming that the for two reasons.First, it was not disproven that display is directly involved in brood parasit- dominants may display toward subordinates ism, one could predict that the frequencyof the in order to appeasethem. That dominant birds display would increase during the breeding display most often to subordinates does not January1983] CowbirdHead-down Display 145 clearly indicate that the display is aggressive. the head-downdisplay of a dominantmale as Dominant birds, free-rangingor captive, may was observed in our study. Furthermore, it is display most often toward subordinatesin or- not clear why such a complicatedthreat be- der to get and stay closeto subordinates,not havior would be selectedfor when other threats, to threaten them. For example, among captive like bill pointing or pecking, would seem to birds dominant males may display toward displaceand dominatea subordinaterecipient subordinatefemales in order to stay close to more quickly than would a head-down dis- them. We, as well as Stevenson (1969), inter- play. pret no clear intent by displayingcowbirds to Rothstein (1980) explains mutual displays, displacerecipients. On the contrary,based on like those above, as a "test of wills" or, alter- the frequentobservation that displayingfree- natively, as mutual threatsby individualsof rangingand captivecowbirds persistently pur- similar socialstatus. Although mutual displays sued fleeing recipients,it seemedto us that could be mutual threats by birds of similar sta- displaying cowbirds "wanted" recipients to tus, in this study subordinate females and remain in closeproximity. We discussedunder dominantmales commonly displayed mutual- hypothesis1 additional reasonswhy the dis- ly. Our observations,therefore, show that the play is appeasing.Second, the conclusionthat prediction that mutually displayingbirds are pecking and displaying have the same mean- of similar social status frequently is not true. ing for cowbirdssimply becausetheir perfor- Hypothesis3.--The displaycannot be behav- mances are closely associatedin time is not ioral mimicry, becausethere is no model for necessarilyvalid. cowbirdsto mimic. Suggestingthat the display Rothstein's (1980) results show that the dis- is an exampleof behavioralmimicry implies play is performedby captivebirds in a situa- that there is selectivepressure on cowbirdsto tion of conflict.Observations during this study mimic intraspecific behaviors of recipient agree. Such a context may explain why some species.No field-verified model behavior was cowbirds alternately displayed and pecked. proposedwith the mimicry theory, nor have Becausewe interpreted the display to be non- we seenduring hundredsof hours in the field aggressive,appeasing, and performedin a sit- a behavior performed intraspecificallyby re- uation of conflict, we prefer to refer to it as a cipient speciesthat might be interpretedas a complexagonistic behavior, i.e. "any behavior model for the display. Furthermore,the mim- associatedwith conflictor fightingbetween two icry theory doesnot explainwhat specificse- individuals," (Scott 1956: 214-215) rather than lectiveadvantage(s) a displayorgains by mim- as an aggressivebehavior. icking the behaviorof anotherspecies. Rothstein (1977, 1980) defined the situation As an alternative to the behavioral mimicry of conflict as the violation of another bird's in- theory, we proposea simpler, more straight- dividual distance. The only thing that we forwardexplanation: both displayorand recip- thoughtcould be threateningabout the display ient interpretthe displayas appeasement.Even was the displayor'sclose proximity to the re- if additional investigation substantiatesthe cipient, not the displayor's posture or move- more complex deception/aggressionhypothe- ments. sis (Rothstein 1980) and it is shown that "re- Rothstein's(1980) experiments also show that ceivers do not respond to the true message the display is most often performedby birds being sent" (Rothstein1980: 173), the display that are dominantto recipients.Observations should not be labeled as behavioral mimicry. during this study supportthat conclusion. Not all forms of deceptionconstitute mimicry. Our results show that among captive birds Hypothesis4.--Hypothesis 4 is refutedby our males were dominant to females (Table 2) and observationsof free-rangingcowbirds. We ob- that females, acrossobservation periods and served the display 288 times during 59.2 h of while in mixed groups, respondedto 66 dis- field observation (Table 1). Furthermore, the playingmales by displayingin return58 times. widespreadoccurrence of the display is sup- If cowbirdsdisplay in order to dominate other ported by the geographical variation in its birds and the display is a threat, we would not sightings.We have observedthe display in expectthat a subordinatefemale, who seldom Connecticut,Maryland (unpubl. data), Ohio, threatensmales with recognizedcowbird threat and Pennsylvania. Lowther and Rothstein behaviors(Scott 1977), would commonlyreturn (1980)reported it in Kansas;Rothstein (1980) 146 Scott ANDGRUMSTRUP-SCOTT [Auk,Vol. 100

saw it in California, Connecticut, and New speculatebriefly about the feeding and roost- York; Dow (1968) observed it in Tennessee; and ing functionsof the head-down display. Join- Darley (1968) saw it in Ontario. Selander and ing a foraginggroup may be advantageousfor La Rue (1961)have sightingsfrom Texas,Rhode a cowbird in obtaining food (Ricklefs1973: 226). Island, New York, Oklahoma, and Florida. The head-down display may facilitate a cow- If more observations were undertaken of bird's joining a foraging group by appeasing cowbirdsas the birds roostduring the day in flockmembers or by helping to assessagonistic mixed-speciesflocks adjacent to feeding areas, tendencies of flock members. In addition, the it would becomeapparent that the display is display may help a cowbird position itself in performed often by free-ranging cowbirds. roosting trees where overnight energy expen- Rothstein (1977, 1980) previously reported diture for thermoregulationwould be mini- cowbirds displaying in this context;observa- mized. Lustick and Kelty (1979) have shown tions during this study supportthe contextand that location in roosting trees affectsthermo- provide additional details. regulation by Starlings and blackbirds. In a Incorporatingthose aspects of hypotheses1- study of a winter roost in Texas (Johnsonet al. 4 that were not rejected, we proposea consol- 1980),researchers found that cowbirdmortality idated hypothesisfor the functionof the head- was positively associatedwith freezing tem- down display as follows: peraturesand "depletion of winter food sup- plies." Becausefeeding and roostingare often The head-downdisplay is an appeasing,ag- flock activities for cowbirds (pers. obs.), the onistic behavior that reducesagonistic be- display may placate other bird species with haviors of the recipient toward the display- whom cowbirdsflock, therebyfacilitating the ing cowbird. The displayor is generally cowbirds' inclusion in the flock. dominant to the recipient, and preening is a Rothstein (1980) suggestedthe flock-order stimulus for subsequentdisplaying by the functionfor the head-downdisplay. It haslong preened cowbird. The display functions in: been known that disorganizationof bird flocks (a) obtaining food, (b) minimizing roosting can result in negative effects on individuals energetics, and/or (c) establishing flock or- der. within the flock (Guhl and Alee 1944). Social disorganizationcould occur, especiallywhen For the display recipient, preening could be cowbirdsare firstcaptured and placedin a cage a "conflict behavior performed" toward cow- with unfamiliar individuals, because the birds birds who have violated the recipient's indi- are unnaturallyovercrowded in their cage(Em- vidual distance (Rothstein 1971, 1980). For len 1950). Rothstein (1977) has stated that for cowbirds,preening appearsto be important to captive cowbirds the number of intra- and in- the display'sappeasement function. Pennsyl- terspecificdisplays decreased rapidly after birds vania cowbirdsthat had just been preeneddis- had been maintained in intact groups for 2-5 played more often than thosethat had not been months. Perhaps the display is observed less preened (P • 0.001). Moreover, cowbirdsmay often among birds that have been togetherfor display toward female Red-wings or House a long time becausesocial order has been es- Sparrowsmost often, becausethe probability tablishedand groupagonistic interactions have, of being preenedis greatestwith thosebirds. therefore, becomeless frequent. In this captive Field observationsduring this study suggest setting, an appeasementdisplay could benefit that preening is a positive responseto the dis- a dominant bird by reducingagonisitic behav- playor that might signal that the agonisticten- iors from conspecificsand allowing more time dencies of the recipient have been reduced. In for behaviors that are more productive than this way, the displayor may be assessingag- fighting (like feeding), by allowing males to onistic tendencies,as Rothstein (1980)has pro- remain closeto females,or by providing cow- posed. birds with an assessementof the agonisticten- The hypothesized display functions are denciesof other flock members(Rothstein 1980). speculative. Although we will discuss them The flock-orderfunction of the display for separately, they are not necessarilymutually free-rangingbirds may operateas follows.Free- exclusiveand may or may not operatein con- ranging cowbirdsmight displaymost often in- cert. terspecificallytoward species with whom they Selander and La Rue (1961) were the first to flock becauseinterspecific flocking is impor- January1983] CowbirdHead-down Display 147 tant to their survival (e.g. for food gathering, GUHL, A.M., & W. C. ALLEE. 1944. Some mea- roosting-sitelocation, or predatoravoidance) surableeffects of socialorganization in flocks of and the head-down display reduces the dis- hens. PhysiolßZool. 17: 320-347. organizingeffects of a cowbirdjoining a mixed- HANN, H. W. 1937. Life history of the Oven-bird in southernMichigan. Wilson Bull. 49: 145-237. species flock. Because interspecific relation- 1941. The cowbird at the nest. Wilson Bull. ships are more fragile and more easilybroken 53: 211-221. off than intraspecificones (Scott 1975), the dis- HICRS, L. E. 1934. A summary of cowbird host play may have evolved into a behavior that speciesin Ohio. Auk 51: 385-386. fostersimportant interspecific relationships. JOHNSON, D. M., G. L. STEWART, M. CORLEY, R. To the bestof our knowledgeit is not known GHRIST, J. HAGNER, A. KETTERER, B. MC- whether or not particular free-ranging cow- DONNELL, W. NEWSOM, E. OWEN, &; P. SAMU- birdsdisplay most often or whetheror not those ELS. 1980. Brown-headed Cowbird (Molothrus ater) mortality in an urban winter roost. Auk 97: birdsare intraspecifically dominant, a trendthat 299-320. has been observed among captive cowbirds LOWTHER, P. E., & S. I. ROTI-ISTEIN. 1980. Head- (Table 3). If this trend were verified for free- down or "preening invitation" displaysinvolv- ranging cowbirds,our hypothesiswould pre- ing juvenile Brown-headed Cowbirds. Condor dict that dominant, free-rangingbirds display 82: 459-460. most often becausethe displayallows them to LUSTICK,S., & M. KELTY. 1979. Effectsof the roost remain close to birds with which they flock, site on the energeticsof blackbirds and Star- which in turn leads them to feeding or roosting lings. Pp. 260-272 in Proceedings8th Bird Con- sites, helps in avoiding predators,and/or pro- trol Seminar, Bowling Green State University, vides assessmentof the agonistictendencies of Bowling Green, Ohio. flock members. MAYFIELD,H. 1961. Vestiges of proprietory inter- est in nestsby the Brown-headedCowbird par- ACKNOWLEDGMENTS asitizing the Kirtland's Warbler. Auk 78: 162- 166. The following people contributedto this study NORMAN, R. F., & R. J. ROBERTSON. 1975. Nest- through discussion and/or written comments on searchingbehavior in the Brown-headed Cow- manuscripts.Ellen D. Ketterson,William B. Jackson, bird. Auk 92: 610-611. StephenI. Rothstein,and StephenH. Vesseycom- PRESCOTT,K.W. 1947. Unusual behavior of a cow- mented on earlier manuscripts,while H. B. Graves, bird and a ScarletTanager at a Red-eyed Vireo Peter E. Lowther, David L. Pearson,and an anony- nest. Wilson Bull. 59: 210. mous reviewer commented on a more recent manu- RICRLEFS,R. E. 1973. Ecology.Newton, Massachu- script.This researchwas fundedin part by the Vet- setts, Chiron Press. erinaryScience Department at The PennsylvaniaState ROBERTSON,R. J., & R. F. NORMAN. 1976. Behav- University.The manuscriptwas approvedas paper ioral defensesto brood parasitismby potential no. 6434 in The Journal Series of the Pennsylvania hosts of the Brown-headed Cowbird. Condor 78: Agricultural Experiment Station. 166-173. ROTHSTEIN,S. I. 1971. A reanalysis of the inter- LITERATURE CITED specific invitation to preening display as per- BAIRD,J. 1958. The post-juvenilemolt of the male formed by the Brown-headed Cowbird (Molo- Brown-headed Cowbird (Molothrus ater). Bird- thrus ater). Amer. Zool. 11: 638. Banding 29: 224-228. ß 1977. The preeninginvitation or head-down DARLEY,J. A. 1968. The social organization of display of parasiticcowbirds: I. Evidencefor in- breeding Brown-headedCowbirds. Unpub- traspecificoccurrence. Condor 79: 13-23. lished Ph.D. dissertation. London, Ontario, 1980. The preeninginvitation or head-down Univ. Western Ontario. display of parasitic cowbirds. II. Experimental Dow, D.D. 1968. Allopreeninginvitation display analysisand evidence for behaviouralmimicry. of a Brown-headed Cowbird to Cardinals under Behaviour 75: 148-184. natural conditions. Wilson Bull. 80: 494-495. SCOTT,J.P. 1956. The analysis of social organiza- EMLEN,J.T. 1950. Flockingbehavior in birds. Auk tion in animals. Ecology 37: 213-221. 69: 160-170. 1975. Violence and the disaggregatedso- FRIEDMANN,H. 1929. The cowbirds, a study in the ciety. AggressiveBehav. 1: 235-260. biologyof socialparasitismß Springfield, Illinois, SCOTT,T. W. 1977. Agonistic behavior and social CharlesC. Thomas, publisher. organization of the Brown-headed Cowbird 1963. Host relations of the parasitic cow- (Molothrusater). Unpublished MS thesis. Bowl- birds. U.S. Natl. Mus. Bull. 233. ing Green, Ohio, Bowling Green StateUniv. 148 SCOTTAND GRUMSTRUP-$COTT [Auk, Vol. 100

$ELANDER,R. K. 1964. Behavior of captive South preening invitation display of parasitic cow- American cowbirds. Auk 81: 394-402. birds. Auk 78: 473-504. --, & D. R. GILLER. 1960. First-year STEVENSON,M. 1969. Agonistic behavior in the of the Brown-headedCowbird and Red-winged cowbirdMolothrus ater. UnpublishedPh.D. dis- Blackbird. Condor 62: 202-214. sertation. Manhattan, Kansas, Kansas State Univ. --, & C. J. LA RUE, JR. 1961. Interspecific ZAR, J. H. 1974. Biostatisticalanalysis. Englewood Cliffs, New Jersey,Prentice Hall Inc.

(continued from p. 138) Phillips, Examinationof specimensat the American Potential effects of food abundance and social envi- Museum of Natural History; Nina Pierpont, Foraging ronment on migratory distancein Dark-eyed Juncos behavior, competition,and relative abundanceof (Juncohyemalis hyemalis); Kimberly Titus, Nest-site woodcreepers(Dendrocolaptidae) at an Amazonian habitat selectionpatterns of woodlandhawks in three forest site; Melinda Pruett-Jones,Vocal copying in temperate forest regions; Jill M. Trainer, Song dia- Macgregor's Bowerbird (Amblyornismacgregoriae); lects, communication, and social organization in StephenPruett-Jones, Ecology and evolutionof social Chestnut-headedOropendolas (Zarhynchuswagleri) organizationin Parotialawesii, Lawes' Six-wired Bird and Yellow-rumped Caciques(Cacicus cela); Peter B. of Paradise;Mark D. Reynolds,Mobbing behaviorof Turchin, Determination of paternity in a polyandrous nestingBlack-billed Magpies; Mark F. Riegner, For- bird, Jacanaspinosa; Katherine A. Voss-Roberts, Pa- aging ecologyof the Yellow-crownedNight Heron; rental attentiveness and embryonic tolerance of Emily FrancescaRissman, Hormones and socialbe- thermal stressin Yellow-headed Blackbirds (Xantho- havior in the Spotted Sandpiper, Actitis macularia; cephalusxanthocephalus); Thomas C. Will, Behavioral Gary Ritchison, Singing behavior of the Cardinal, ecology of species replacement: Blue-winged and Cardinaliscardinalis, with emphasis on the function Golden-wingedwarblers in Michigan; PamelaL. Wil- of singing by females;Nicholas L. Rodenhouse,In- liams, Comparisonof colonialand non-colonialnest- dividual and regional variability in the foraging be- ing by northern orioles in central coastalCalifornia; havior of forestpasserines; Stephen R. Sabo,Mixed- Kathy A. Winnett-Murray, Behavioralecology of four speciesflocks of Hawaiian honeycreepers;Bette J. Central American wrens; Barbara L. Woods, Mate- Schardien,Behavioral ecology of a southernpopula- guarding and female reproductivesynchrony in Barn tion of Killdeer; Debra H. Schlenoff, The "startle" and Cliff swallows; David A. Yokel, Mating system responseof wild-caughtbirds to artificial prey; Hel- of the Brown-headed Cowbird in two different en- mut Sick, Study of specimensin the AmericanMu- vironments; Robert M. Zink, Allozymic correlates seum of Natural History and identification of birds of dominancebehavior in Dark-eyedJuncos: a reap- from Brazil; Thomas Allen Stevens, Food hoarding praisal. by Great Grey and LesserGrey Shrikes, Laniusexcu- Correction to 1981 list: MarthLeah Chaiken re- bior and L. minor, a quantitative and experimental ceived a grant for research on the functions and study; Stuart D. Strahl, Ecologicalcorrelates of the development of variability in some calls of the socialsystem and foraging behavior of the hoatzin European Starling ( Sturnus vulgaris ). (Opiathocomushoazin) in Venezuela;Scott B. Terrill, TIME AND ENERGY BUDGETS OF THE MOCKINGBIRD (MIMUS POLYGLOTTOS) DURING THE BREEDING SEASON

DOUGLAS W. BIEDENWEG 1 Departmentof Zoology,University of Californiaat Davis, Davis, California92065 USA

ABSTRACT.--Atime and energy budget of male and female Mockingbirds(Mimus poly- glottos)in six phasesof their breedingcycle was completedusing the equivalenttemperature modelto predict perchingmetabolic rates. Recorded time-budget behaviors included perch- ing, bipedal locomotion, flight, and the time spent in the sun and shade. Over the total study period, malesand femalesspent an averageof 92.3% and 92.8% of their activeday perched,2.4% and 1.6% in bipedallocomotion, and 5.3% and 5.6% in flight, respectively. BMR plus thermostaticrequirements averaged, seasonally, 78.1% of DEEtotin males and 76.1% in females.The choiceof shadedor exposedmicrohabitats made very little difference in energy expenditure. DEEtotvaried from 96.9 kJ in maleswith incubating femalesto 113.0 kJin malescaring for fledglings,and from 75.1 kJin incubatingfemales to 113.2kJ in females caringfor fledglings.Unmated, pre-nest-building,and incubatingbirds spent significantly lessenergy than did birds caringfor nestlings(P • 0.05) or birds caringfor fledglings(P • 0.05). With the exceptionof the nest-buildingphase, when femalesare producingeggs, and the incubation phase, when females gain insulation from the nest, male and female DEE•ot closelyparalleled each other throughout the season.The large amount of predation that occurredat the studysite suggeststhat birds may allocatetime to perchingto minimize the probability of predation on themselvesor their offspring. Received26 February1982, accepted 4 October 1982.

A VIABLEapproach to ecologicaland evolu- enter the Hutchinsonian "ecological theater" tionary problemsis available through the use (Hutchinson 1965)armed with time and energy of time- and energy-budgetanalysis. Because concepts,we come out with a more thorough time and energy are resourcescommon to all understandingof the proximal elementslink- organisms, this approach can serve both to ing organismsto their environments.We be- quantify the diversity of interactionsbetween come sensitive to the energeticpatchiness of an organismand its environment(a functional an organism'senvironment and may visualize problem) and to decipher evolutionarytrends it as a seriesof microhabitatsin which biolog- and their causes(Porter and Gates 1969, King ical and physicalfactors interact to determine 1974). This is the key to the vitality of the ap- whether an organism experiencesnet energy proach: functional and evolutionary problems gains or losses.We are granteda clearerun- can be addressedsimultaneously (Mayr 1961). derstandingof the reasonsfor the distribution As more comparativetime and energy data are of organismswithin habitats. gathered and analyzed, ecologicaland phylo- The habits of the Mockingbird (Mimuspoly- genetic correlatesof life history elements (e.g. glottos)make it an ideal speciesof which to energy invested per egg, energy invested in analyzetime and energybudgets. Individuals offspring, time and energy devoted to off- nest in semi-open habitats, which facilitates spring by males and females)will emerge, in- behavioral observations;they are highly ter- creasingour power to discriminateecological ritorial; and their home rangesare sufficiently problems and the modes in which natural se- small that they can be covered rapidly by a lection attempts to solve them. But this is not person on foot. In this report I present time all that the technique does for us. When we and energy budgets for Mockingbirds during a singlebreeding season in Davis, California. The major goalswere to discernthe dynamics • Presentaddress: Department of Zoology,Wash- of time and energy expenditureof males and ington StateUniversity, Pullman, Washington 99164 females throughout the reproductive cycle, to USA. investigatethe role that microhabitatselection

149 The Auk 100: 149-160. January1983 150 DOUGLASW. BIEr)EtaWEG [Auk, Vol. 100 plays in determining energy expenditure,and ranges were traced on scalephotographs (1" = 100') to comparethe time budgetsof this Mocking- obtained from the Department of Architects and En- bird population with those collectedby Utter gineersat the Universityof Californiaat Davis. Areas (1971)on Mockingbirdsin New Jersey. were then calculatedfrom thesephotographs.

METHODS CALCULATION OF ENERGY EXPENDITURE The activity patterns of Mockingbirds were stud- An energybudget for an animal under steady-state ied in riparian and urban habitats near the veteri- conditions with its environment, where conduction nary medicalcomplex of the University of California is not important, can be written at Davis, Yolo County, California. The study lasted from 2 February 1978 to 6 June 1978. The habitat of MR - ?•E= [pcv/(r,,+ re)](Tb-- Te) (1) most birds consistedof fallow fields with patchesof asphalt parking lots and expansesof lawn. Oaks where MR is metabolic rate (W.m •), kE is latent ( Quercusilex and/or Quercusagrifolia ) were scattered heat loss (W'm-2), r•, is whole-bodythermal resis- over most birds' home ranges.Pyracanthaberry (Pyr- tance(s.m •), re is equivalentresistance, Tt, is body acanthasp.) was commonlyavailable, and its fruits temperature(øC), T• is equivalenttemperature, and were used by birds in all phases of the breeding the productof the density (p) and the specificheat cycle. of air (cv)is a constant(1,200 J. m-3. K-• at 20øC)(Ma- Five Mockingbirdswere mist-netted,weighed, and honey1976, Campbell 1977). This equationwas used color-bandedin January1978. The sexesof the birds to predict the metabolicrates of perchingMocking- were determined by subsequentobservations of be- birds during their activity phase whenever environ- havior (singing, displays,etc.). The mean weight of mental conditions were such that MR exceeded the the birds was 54.8 g (range, 48.9-59.3 g). The only metabolic rate predicted by the Aschoff and Pohl positively identified female in the sample also (1970) (A/P) diurnal equation for passerines,multi- weighed 54.8 g, so I assumethat there is no weight plied by a factor of 1.2 [the factor 1.2 correctsfor dimorphism. illuminationunder diurnal conditions (Mahoney and The duration of three types of activities and the King 1977)]. Such conditions existed when 35øC • time spent in the sun and in the shade were quan- T•. • 14øC. tified using a panel of four stopwatches.Data were To computea metabolicrate with equation1, it is primarily recordedin hourly periods. At the end of necessary to measure or calculate values for several each hourly period, air temperature and the per- bird and environmental parameters. Environmental centage of cloud cover were recorded, and wind- parametersinclude short-wave (SW) radiation,long- speeds were measured in the sun and shade in mi- wave (LW) radiation, windspeed, and ambient tem- crohabitats similar to the ones that the bird had perature. A bird's SW radiation environment was occupied. Windspeed was measured with a Ther- determined from measured values of total SW radia- monetics series 100 hot wire anemometer. Data on tion using the techniqueof Campbell (1977:table 5.1, total short-wave radiation per hour were obtained equations5.7 and 5.8; 1981: equations 1.5, 1.6, 1.9). from the meteorologicalstation at the University of The LW radiation environment was calculated from California at Davis, located 400 m from the study equation 11 in Unsworth and Monteith (1975) and site. equation 5.13 in Campbell (1977). Bird parameters Activity categoriesincluded perching, flying, and incorporatedin equation 1 include the characteristic bipedal locomotion. Bipedal locomotion (BPM) was dimension (d), the SW absorptivitiesof the back and restricted to the actual moments when locomotion belly, the amountof area exposedto differenttypes occurred.The BPM categorydoes not representtime of SW radiation, various resistancesto heat transfer, spentforaging. The open habitat allowed me to mon- and evaporative water loss. The characteristic di- itor the birds' behaviors continuously. In the rare mension was measured,and the absorptivitiesof the instancesthat birds were lost during observationpe- birds' backsand bellies (0.6 and 0.51, respectively) riods, the data for that period were excluded unless were estimatedby comparingMockingbird plumage visual contactwas regained in lessthan 2 min. to bird plumagesthat were similarly coloredand the The birds' seasonalcycle was divided into six data- absorptivities of which were known. The ratio of the collectionphases: unmated, pre-nest-building, nest- area exposedto perpendicularSW radiation over the building, incubation,nestling care, and fledglingcare. total area of the bird was estimatedas 0.21 (Mugaas Data were collectedthroughout the day and several 1976). It was also assumedthat the upper half of the times at night. Male and female power-consumption bird's total area was exposedto scatteredSW radia- data were pooled in each phase for statistical anal- tion, while the lower half was exposedto reflected ysis. SW radiation (Mahoney 1976). Maximum total resis- Home rangeswere estimatedfrom behavioral ob- tance was calculated from rt = 283mø'2•, where m is servations several times during the study. Home the bird's mass(g) (Robinsonet al. 1976).The max- January1983] MockingbirdTime-Energy Budgets 151 imum value that re could attain was 158 s. m • or 24% were collectedat night, when temperaturesdropped of the maximum total resistance(Mahoney and King below the TNZ, Kendeigh's(1969) equation, H (kcals/ 1977).Maximum body resistance,r•, max, during the day) = 4.769mø'• [where m is mass (g)], was used day was calculatedto equal 423 s.m • by equations to calculate a metabolic rate at 0øC, and metabolism given in Robinson et al. (1976). Using this value as was computed based on the assumptionthat it de- a peak for rb and the shapeof Fig. 2 in Mahoney and creaseslinearly as temperatureincreases from 0 to King (1977), I derived body resistancesat other 22.5øC. A single temperature (the mean nighttime equivalent temperatures.Evaporative water loss (E) temperature)was used to calculatehourly energy ex- was calculatedfrom data from the Pyrrhuloxia(Car- penditure. dinalis sinuatus) and the Northern Cardinal (C. car- Incubatingbirds.--The modelingprocedure for in- dinalis)(Hinds and Calder 1973). They give an equa- cubating birds was the same as for nonincubating tion for water loss between 20øC and 32.5øC: E(g/ birds exceptfor periods when birds were on the nest. day)= 0.48mø'•45, where m is body mass (g). This Becausebirds nested in heavy vegetationwhere little equation was used to calculateevaporative water loss or no light penetrated,SW radiationfell to 0 W. m • in the range of 20øC-32.5øC.Below 20øC,the graph under these conditions. The model assumed that 10% was drawn to 0.5% body mass evaporated per hour of the incubating bird's surfacearea was coveredby at 13ø C (in accordancewith their data) and then was the broodpatch, that 40% of its surfacewas exposed, further extrapolatedto 7ø C, the lowest T•. that oc- and that 50% of its surface area was in the nest (ex- curred during my study. At 45øCthe data of Hinds cluding the brood patch). The resistancefor a 1.9- and Calder give unreasonablylow resultsfor a Mock- cm-wide egg (Bent 1948) was set at 41.9 s.m •; the ingbird's total metabolic rate. For this reason, I as- brood patch resistancewas set equal to 45 s'm •; sumed that evaporative water loss is equivalent to body resistancewas approximatedas 423 s' m-• (max. metabolic heat production at an equivalent temper- night); and nest resistancewas set equal to 300 s. ature equal to the birds' body temperature (42øC), m-• (valuesestimated in the manner of Walsbergand and I set both equal to 105 W.m -2 at this tempera- King 1978b).The characteristicdimension for the nest ture. For equivalent temperaturesgreater than 35øC, with the bird was approximatedas 15 cm. All other this approximationyields the following equation: E facets of this model were similar to the analysis by (W.m -2) = 10.2 T•.(øC)- 324.2. Walsbergand King (1978b).In this phaseof the cycle, Calculationof hourly metabolicrates.--Two equiv- the calculated metabolic rates for birds on the nest alent temperatureswere determinedfor each obser- during the day were compared to the Aschoff and vation period: one for conditionsof full sunlightand Pohl (1970)resting equation. If predictedmetabolic one for conditions in the shade. A metabolic rate was rates on the nest during the day dropped below this determined for each equivalent temperature. The value, they were set equal to it. During the night, metabolic rate was then recalculatedto represent the incubating birds' metabolic rates were set equal to proportion of each hour spent in the sun and shade. 82% of a nonincubating bird's metabolicrate under The costs of bipedal locomotion and flight were similar temperature conditions (Walsberg and King consideredto be independent of environmentalcon- 1978b). ditions. The cost of bipedal locomotion (BPM) was Eggformation.--Egg formation was assumedto take assumed to be 1.5 times the cost of perching in the 5 days, and the cost of forming i g of egg, at an thermal neutral zone, and the cost of flight was cal- efficiencyof 70%, was estimated as 6.28 kJ (King culated as equal to 9.4 times the metabolic rate cal- 1973).Because a singleMockingbird egg weighs 3.75- culated from the Aschoff and Pohl (1970) equation 4.25 g (Hanna 1924), a total cost of 25.1 kJ per egg for passerinesresting at night (Hart and Berger1972). was assumed.Because clutch size near Davis ranged The final cost of an hourly period for a nonrepro- from 3 to 4 eggs(n = 8), the total coststo form 3 eggs ducing bird was the sum of the costof eachbehavior in 7 days and 4 eggsin 8 days were both calculated. (J'm 2.h-l), times the fraction of the hour spent in I chose the mean of these two values, 468 J'h •, to that behavior, multiplied by the surfacearea of the add to the hourly expenditureof egg-producingfe- bird (mS). Plumage surfaceareas were determined males. Nest-building femaleswere the only group of from SA = 7.81(m)ø'6•7 (Walsberg and King 1978a), birds to have this costof production added to hourly where m is bird mass(g) and $A is surfacearea (cm2). expenditure. Nighttime energy expenditure.--Hourly nighttime Convertinghourly data to daily energyexpenditure.- energyexpenditure in the thermalneutral zone (TNZ) To convert energy expenditureper hour to daily en- (22.5-35øC)was calculatedfrom 114.8M ø'726(M in kg) ergy expenditure (DEE,,t), a mean value of hourly (Aschoffand Pohl 1970).The rangeof environmental expenditurewas calculatedfor eachbird in that par- temperaturesthat delimits the TNZ was calculated ticular phase, and then a mean of these means was for the lower critical temperature from T•.c = T•, - calculated. This value was used for a bird during 6.98mø'•, wherem is mass(g) (Calderand King 1974). waking hours (from 15 min before civil twilight in The upper critical temperaturewas assumedto equal the morning to 15 min after civil twilight in the eve- 35øC. Becauseno windspeed or LW radiation data ning). DEEt,,•includes the costsof basalmetabolism, 152 DOUGLASW. BIEt)ENWEG [Auk, Vol. 100

TABLE 1. Sensitivity of DEEtot to errors in model TABLE2. Sensitivity of DEEtotto errors in model parameters at T•, = 10øC, T• = 12øC.a parametersat T(, = 25øC,T, = 27øCand To = 30øC, T• = 32øC.a Radiation (W.m -2) and wind- Variableband speed(m-s -t) Per- centage % change 50, 0.5 880, 0.5 50, 5 880, 5 Esti- deviation Mass mated from +10% +5% c +5% +5% +5% Variableb and DEEtot original -10% -5% -5% -5% -5% percentagechange (kJ.h •) estimate Perchingd Mass + 10% c 117.2 + 6% a +25% +21% +21% +21% +21% Perchingcost + 25%• 124.3 +13% -25% -21% -21% -21% -21% Flight cost + 25%• 115.9 +5% Surface area + 25% 116.4 +5% Flighte Nighttime metabolism+ 10% 113.6 +3% +25% +4% +4% +4% +4% • Under all radiation (W'm •) and windspeed(m.s -•) conditions -25% -4% -4% -4% -4% specified,the resultswere identical. • Variables tested were identical to those at % - 10øC. All other Surface area variablesled to lessthan a 2% changein DEEtot. +25% +7% +2% +8% +8% •' Changesin the minus directionmirrored changes in the positive direction for all variables in the table. -25% -6% -5% -9% -6% a The standardDEEtot under all conditionsof radiation and wind is 110.4 kJ. rb ' Costs include BMR. +25% -2% -1% -5% -2% -25% +9% +2% +11% +11% Night metabolism +10% +4% +4% +4% +4% -10% -4% -4% -4% -4% tion multiplied by 1.2 was comparedto that produced by the Te model, and the higher value was used in • Valuesgiven are the percentageincrease or decreasein DEEtotdue to the changein the parameterindicated, under the total short-wave calculatinga perchingmetabolic rate), someoutputs radiationand windspeadconditions specified. were asymmetricalin the analyses(e.g. r0 and sur- b Other variablestested and their deviationsincluded % _+IøC; % +- 10øC;To -+ 2øC;albedo, Av/At, An/At, d, E, re, and the costof BPM, face area at T, = 10; T• = 12øC).The percentagethat all varied by 25%; and back and belly absorptivitiesñ 10%. All vari- input parameterswere varied representsprobable ations led to less than a 2% changein DEEtot. maximal errors in these values. •' The standardgEEtot'S for the conditions of total short-wave radia- tion and wind given are 50, 0.5:108.2 kJ;880, 0.5:107.1 kJ;50, 5:112.3 For nonincubatingbirds at T•, = 10, 25, and 30øC, kJ; and 880, 5:108.8 kJ. the model is most sensitiveto changesin the values a Costsinclude BMR plus TR and encompassthe total 24-h period. ' Costs include BMR. of the cost of perching, the mass of the bird, the surfacearea of the bird, and the cost of flight. At T•, = 10øC, the model is also highly sensitive to changesin body resistance.Variation of other vari- thermoregulation,production, and activity. DEEtot ables, to the degreeindicated in Table 1, led to less values could be in error if there were trends in en- than a 2% change in DEEtotunder all conditions ergyexpenditure throughout the activityperiod, but specified. no daily trendswere apparentwhen the data were Errorsin the costof perchingcan have a large im- analyzedseparately in four activity periods:before pact on DEE due to the large Ëaction of time occu- 0900, 0900-1200, 1200-1700, and after 1700. pied by this activity (Ettingerand King 1980).A 25% error in the value of perching metabolicrate at T, = SENSITIVITY ANALYSIS 25 or 30øCleads to a 13% changein DEEtotunder all radiation and wind conditions, while at T, = 10øC a Becauseit was necessaryto estimate many param- similar error leads to a 21% error in DEEtot. A 10% eters in the DEE model, the effects on outcomes of change in mass with %, = 25 or 30øCcaused a 6% the model were analyzedby varying eachparameter increase in DEEtot, while for To = 10øCthe increases by 10-25%. Thesesensitivity analyses were run un- and decreaseswere 5%. Errors in DEEtotresulting der four conditions of radiation and wind, and at from 25% errorsin surface-areaapproximations pro- three different combinationsof air and ground tem- duced a changein DEEtot-< 9% (Tables 1, 2). Flight peratures(Tables 1, 2). The DEE valuesproduced by occupieda small percentageof the day but was an the model were compared to a matching standard, energeticallyexpensive behavior. A 25% changein which had identical SW radiation and windspeed the costof flight causeda 5% changein DEE•otat 25 values. Because the model used to calculate DEE var- and 30øCand a 4% changein DEEt,,•at T, = 10øC.A ied with environmental conditions (i.e. the meta- 25% decrease in rb at 10øC caused an 11% increase bolic rate calculatedfrom the Aschoff and Pohl equa- in DEEtotat high windspeeds and a 9% increase in January1983] MockingbirdTime-Energy Budgets 153

DEEtotunder low windspeeds and low radiation. In general, energy devoted to maintenance was utilized only when T•'s were below 14øC,and was obligatory, but, when ambient tempera- this occurredcommonly only during the unmated and ture was below 14øC,birds could modify this pre-nest-buildingphases (less than 7% of all other expenditurethrough the differentialutilization phasescombined had unshaded Te's below 14øC). of microhabitats. T,'s below 14øC occurred Therefore,these are the only phasesof the study that could be much affectedby errors in during 26 of 40 samplingbouts involving non- mated males, in 7 out of 29 bouts with males The major value of such sensitivity analysesis to indicate which variables need to be measured with in the pre-nest-buildingphase, and in 4 out of the greatestaccuracy under specifiedenvironmental 28 bouts dealing with pre-nest-building fe- conditions. To set confidence limits on model out- males. puts of DEEtotwhen all errorsare combinedis more To investigate the role that microhabitat difficult. The probability of all errors being simul- choiceplayed in determining TR requirements taneouslyadditive is scant, however (somewill can- during eachof thesephases, I determinedthe cel othersof oppositesign), so a probable95% con- mean fraction of time spent in the shade and fidence limit should lie in the range of +5% of the sun for each phase during periods when T•, DEE predictedby the model. was less than 14øC. A metabolic rate for each bird in the sun and in the shade for each of RESULTS AND DISCUSSION thesehourly periodswas then calculated.The BASAL AND THERMOSTATIC EXPENDITURE lower of these two values was used to deter- mine a minimum rate of energy expenditure, The percentageof the total daily energy ex- and this was compared to the actual rate of penditure devoted to basal and thermostatic energy expenditure the bird experienced. An requirements (BMR and TR, respectively) averagevalue of potential energy savingswas rangedfrom 67%, in femalescaring for fledg- then obtained for each phase. Unmated and lings, to 86%, in incubatingfemales. BMR plus pre-nest-buildingmales could have saved 119.2 TR averages,seasonally, 78.1% of DEEtotin and 53.7 J' day-l, respectively,by utilizing the malesand 76.1% in females.Only 12% (range physical environment in such a way as to 5.5-19%) of DEEtotis consumedby thermoreg- minimize energyexpenditure, while pre-nest- ulation in males and 9.6% (range 5.5-15.3%) building females could have saved 1,110 by thermoregulationin females.Approximate- J. dayL ly 99% of the thermoregulatorycosts accrued The ecologicalsignificance of thesepotential at night, as daytime temperaturesrarely fell energy savings can be investigatedby con- below 14øC.A comparisonof thesedata on the verting theseenergy equivalents into food ro- percentageof power consumptiondue to BMR sources.Pyracantha berries were commonly plus TR with extant data showsmy values to utilized on many territoriesduring thesephas- be the highestyet known. Kendeigh'srange of es and have a mean energycontent of 1.24 kJ' estimatesfor House Sparrow (Passerdomesti- berry • (n = 2;2?= 18.4kJ'g •, s = 0.20).Thus, cus) (67-80%) BMR plus TR comesclosest to under these environmental circumstances mi- the Mockingbird values (Kendeigh1972). crohabitatchoice for energy savingsis of little The large fractionof DEEtotaccounted for by importance.Nevertheless, the insightsinto the basal and thermoregulatorypower consump- within-habitat distribution of endotherms tion in this model, relative to previousmodels, gainedby breakingthe climatespace of Porter probablystems primarily from the limits of our and Gates (1969) into compartments corre- knowledgeon the energeticcosts of behaviors. sponding to regionswithin the thermoneutral This model assumesthat, under a given set of zone and outside the thermoneutral zone are environmental conditions, all daytime activi- potentiallygreat, as a similarapproach (that of ties in nonincubatingbirds, otherthan bipedal measuringfunctional-response rates for a wide locomotionand flight, cost a constantamount. varietyof traitsat differenttemperatures, iden- Previousmodels (e.g. Mugaas 1976,Walsberg 1978, Ettinger and King 1980) have assumed tifying peak responsetemperatures for each that nonflight activities (e.g. perching, sing- trait, and searchingfor temperaturesthat max- ing, foraging)other than BPM are more costly, imize the number of traits whose peaks over- which decreases the fraction of DEEt.t due to lap) has provento be for ectotherms(Brett 1971, BMR plus TR. Huey and Stevenson1979). 154 DOUGLASW. B•EDEt•WEG [Auk, Vol. 100

VARIATIONS OF ACTIVITY AND DEE BETWEEN AND WITHIN REPRODUCTIVE PHASES Orderingthe dataaccording to reproductive phaseallows the detectionof energeticcorre- lationsamong reproductive, other biotic (e.g. predation and food resources),and abiotic fac- tors. Those phasesimposing the strictestlim- itationson time budgetsand requiringthe greatestmobilization ofresources can thus be revealed. Using phases of the breeding cycle as standardsfor data collectionalso allows for precisecomparative studies within a taxon that mayreveal life-history or behavioraladapta- tions, allow them to be correlated with habitat changes,and suggest causes for adaptation (Lack 1968). Unmatedphase.--During thisphase males spentrelatively little time in flight (Table3). Foraging was restricted to fruits available withinthe territories, and males generally hopped or ran from bush to bush. Aggressive displaystoward robins (Turdus migratorius) were common.Because of the low tendencyto fly, the smallhome range sizes, and the short active daylength(Table 4), DEEpest[DEEpest in- cludesthe costs ofproduction.and activity alone (EttingerandKing 1980)] forthe males reached a seasonallow during the nonmatedphase. Pre-nest-buildingphase.--Birds firstformed a pairbond on 10 January. Quantitative obser- vations of this phase were made between 19 Februaryand 26 May. Severalbirds moved into thisphase more than once during this study. Male home-rangesizes were of intermediate values,and time spentin flightslightly in- creasedover that of unmated males, primarily dueto theexploration of surrounding areas (Table 3, 5). Increasing temperaturescompen- satedfor increasing active daylength and the increasedpercentage of time spentin flight, so that males'DEEtot remained constant between thesetwo phases(Tables 3, 4). The females' rangein energyexpenditure was more tightly clustered than that of the males (4.0-4.8 kJ'h -• to3.9-6.4 kJ.h -•, respectively).Females did not exploresurrounding areas during this phase. Nest-buildingbirds.--Nest building required 2 daysfor the three pairs of birds observed. Nestconstruction began on 20 February. Time- budgetobservations in this phasewere made on a singlemated pair. The malespent more time in BPM during this phase than in any other phase of the study (Table 3), mostly in January1983] MockingbirdTime-Energy Budgets 155

TABLE4. Calculationof mean DEEtotfor each phaseof the study.

Night- Daytime Day time Night expendi- hours expendi- hours ture avail- ture avail- Date Phaseof cycle (J.h- D able (J.h •) able DEEtot 10 February Unmated 4,422 12.0 3,930b 12.0 100.2 15 April Pre-nest-building(c•) 4,594 14.6 3,532 9.4 100.2 15 April Pre-nest-building( $ ) 4,364 14.6 3,532 9.4 96.9 1 May Nestbuilding (c•) 4,652 15.2 3,142 8.8 98.4 1 May Nestbuilding (•) 5,131•' 15.2 3,610c 8.8 109.8 5 May Incubation (c•) 4,381 15.3 3,434 8.7 96.9 5 May Incubation ( $ ) 3,307 15.3 2,816 8.7 75.1 21 May Nestlings (c•) 5,149 15.8 3,142 8.2 107.2 21 May Nestlings ( $ ) 5,138 15.8 3,142 8.2 107.0 24 May Fledglings(c•) 5,356 15.9 3,434 8.1 113.0 12 May•l Fledglings( $ ) 5,538 15.6 3,284 8.4 113.2 Activity begins15 min beforecivil twilight in the morningand ends15 min after civil twilight in the evening. Nighttimetemperatures were as follows: 10 February, 5.5øC; 15 April, 10.0øC;1 and21 May, 14.4øC;5 and24 May, 11.1øC;12 May, 12.8øC. Includes468 J.h • production. Femaledata collected for thisphase were centered around 12 May;male data around 24 May.

collectingtwigs. He contributed4-6 twigs per of their flight time to predator defense, how- hour, comparedto the female's range of 0-1 ever, while femalesgenerally used flight as a twigs per hour. Femalesincreased power con- means to reach a broader range of foraging sumption substantiallyduring this phase due habitats. to egg formation (Table 3). Except for this ex- The ranges in the mean values of daytime pense,the female'stime andenergy budget was energy expenditurewere large for both males essentially identical to the male's. and females (4.0-8.2 kJ-h • and 3.9-7.9 kJ'h -•, Incubationphase.--Incubation lasted approx- respectively). Males varied greatly in the imately 13 days in the one nest observed in amount of time devoted to feeding offspring detail. Similarly, Laskey(1962) found incuba- and predator defense. Observations of male 1 tion periods from 12 to 12.5 days in 18 nests. suggestthat he fed nestlingsmore frequently Male Mockingbirds neither incubatednor fed (4 times in 1 h) than did male 3 (who fed nest- the female at the nest. Both male and female lings but 7 times in 18 h of observation).Male power consumption reached a seasonal low 3 spent a large amount of energy fending off during the incubation phase (Table 3). Females potential predators. He attacked crows, (Cor- actuallyflew for a slightly greaterpercentage vus brachyrhynchos),American Kestrels (Falco of time in this phasethan they did in the pre- sparverius),dogs, and a Northern Harrier (Cir- nest-buildingphase, but increasingtempera- cus cyaneus)during the observationperiods. tures, combined with the large amount of time The variation in female values was caused by spent on the well-insulated nest (mean for ac- differences in foraging ranges. Female 2 uti- tive portion of the day = 78.2%), led to an lized a large home range and spent a mean of overall decline in DEE,,•. This decline is in ac- 11.2% of her active day in flight, whereas fe- cordancewith the theoryof Walsbergand King male I spenta mean of only 3.9% of her active (1978b) that the incubation phase is a period day in flight while utilizing a smaller home of reducedpower consumptionbecause of shifts range.For females,an increaseof time in flight within the entiretime/energy budget. did not reflect a greaternumber of feedingsat Birdswith nestlings.--The nestling period lasts the nest. Female I fed her young 6 times in 1 12 days. All nestscontained two young. Av- h. Female 2 fed her young in 11 of 12 h of erage male and female time/energybudgets observation, with a mean of 3.6 items (fruits during this phasewere virtually identical(Ta- or insects).Females apparently feed young at ble 3). Males and females substantially in- greater rates than males during the nestling creasedthe percentageof the active day spent phase (this study, Utter 1971). in flight. Males devoted a greaterpercentage Fledglingphase.--Only pair 1 raised an off- 156 DOUGLAS W. BIEDENWEG [Auk, Vol. 100

TABLE5. Home-range sizes.

Bird Phaseof cycle Date(s) Area (ha) Male 1 Pre-nest-building, nest-building,nestlings 8 February-ll April 0.58 Pre-nest-building 12 April 1.63 Fledglings 10 May 1.66• Fledglings 26 May 1.33 Male 2 Pre-nest-building 21 April 1.96 Nestlings 20 May-26 May 2.92 Nestlings 27 May 5.28 Nestlings 3 June 7.06 Male 3 Pre-nest-building 22 April 4.73 Nestlings 22 May-26 May 7.76 Pre-nest-building 27 May 11.00 Male 4 Unmated 18 February-19February 0.42 Male 5 Unmated 3 February-7 February 0.11 Also includesa 415-m flypath to a small grove of trees.

spring that survived. This bird left the nest on pie sizes. Statistical comparisonsof DEE,..t 10 May and was still being fed by the parents yielded similar results. when the study was terminatedon 16 June. The patternof interphasicvariation in DEE•o• Males and femalesreached peaks in seasonal almost certainly does not reflect an ultimate values of DEEtotand DEE,,,t during this phase patternof time and energyexpenditure evolved (Table3). The relativelylong activedaylengths by thesebirds. More likely, it is producedby and the increasein the percentageof the active variationsin a particularset of proximalfactors day devoted to flight combined to producethis in the birds' environments.Major proximal effect (Tables 3, 4). Both males and females fed factorsthat have the potentialof varyingbe- the young. The male of pair 1 fed the young tween phasesare socialstructure (territory at- very little early in the phase (1 time in 7 h). tainment and defense, pair-bond formation), During the same time span, the female fed the daylength,the physicalenvironment, the de- two offspring a mean of 6.1 fruits or insectsper mandsof eggsor young,predator pressure, and period (n = 18). Ten days after nest departure, resource abundance. the male bird startedconstructing another nest Unmated males showed extreme territorial and simultaneouslyslowly began to take over aggressionin January and February when the feedingof the fledgedoffspring. By the time breedingterritories were being obtained.Pair- incubationwas initiated by the female,the male bondformation did not increaseenergy expen- had taken over the feeding of the fledgling diture over that of territorial defense alone completely, providing a mean of 6.5 fruits or (compare unmated and pre-nest-building insects per hour (n = 9). males; Table 4), and even the additional de- mand of building a nest while holding a mate INTERPHASlC COMPARISONS AND TRENDS and a territory did not increasethe amount of energydevoted to activity. Unmated and incubatingbirds spent signif- Increasingdaylengths allow more young to icantly less energy than birds caring for nest- be produced,but it is the productionof young lings (P < 0.05) or birds caring for fledglings thatcauses the energyrequirements to change, (P < 0.01) (t-test, Steele and Torrie 1960). In- not increasesin daylength. A comparisonof cubatingbirds alsospent significantlyless en- BMR plus TR costsbetween phasesshows that ergy than pre-nest-building birds (P < 0.05). the physicalenvironment's impact on variation Pre-nest-buildingbirds spentsignificantly less between phases was small (Table 3). The en- energy than either birds caring for nestlings ergy demandsof production(eggs, then main- (P < 0.05) or birds caring for fledglings (P < tenanceand growth of young) increaseas the 0.05). No statistical tests were attempted on phasesprogress (Ricklefs 1974). This causesthe nest-building birds because of the small sam- total energydemands of parentsand youngto January1983] MockingbirdTime-Energy Budgets 157

TABLE6. Total energy expenditure(DEEtot) and energy expenditurein activitiesabove the basal and ther- mostaticrequirements (DEE,,.t) during the time period required to raise young to a stateof free existence. In the mid-fledgling phase, femalesbegin a secondbrood, while males care for fledglingsof the first brood.

kJ' phase

Male Female

Phase Days D EEtot D EE,,,.t D EEto D EE,,,.t Pre-nest-building 5.5 551 113 595 155• Nest-building 2.0 197 36 220 59 Incubation 12.0 1,163 217 901 124 Nestlings 12.0 1,286 377 1,284 378 Fledglings 37 c•c•, 13 2 2 4,181 1,310 1,481 478 Nestling/eggformation 7.5 -- -- 933 360• Incubation 12.0 -- -- 901 124 Nestling 4.5 -- -- 482 142 Total, kJ 7,378 2,053 6,797 1,820 Mean kJ.day -• 108 30 99 27 • In the model,egg formation occurs during the last5.5 daysof the pre-nest-buildingphase, so 11.2 kJ'day • havebeen added to valuesfrom Table 3 when appropriate. increasethroughout the season.Further energy and/orquality did not increasecommensurate- demands are elicited by the increase in the ly with energydemands. number of potential predators,which occurred A large percentageof time was devoted to with the start of the incubationphase and ex- perchingduring all phases.This seemsto be a tended to the fledglingperiod (eggs,nestlings, regularfeature of avian lifestyles(Ettinger and and fledglings are assumedto have higher King 1980).At leasta portion of this perching probabilitiesof being preyed upon than adults). time is not involved with foraging.Ettinger and As adults moved into these phases their de- King (1980)have suggestedthat Wilson's(1975) votion of time and energy toward potential "principle of stringency"may accountfor the predators increased (Utter 1971, this study). evolution of this portion, which they have Data collectedshow that unmatedand pre-nest- termed the "loafing" component.In the pop- building birds ocassionallyattack Scrub Jays ulation that I studied, and for passerines in (Aphelocomacoerulescens), but do not attack general,this is probably not the case.It is im- American Crows; nest-building and incubat- probable that birds would decreasetheir fit- ing birds attackjays and crowsbut not catsor nessesby foragingfor greaterperiods of time dogs; birds with nestlings display agonistic when conditions were favorable because later behaviorstoward crows,cats, dogs, and hawks; they might have to contend with an unpre- and birds with fledglingsassaulted crows, cats, dictable event such as cold weather, rain, or and dogs during the time that their fledglings snow that would reducethe food supply. Birds were alive. The extraenergy demands imposed have, no doubt, evolved life-history traits in by predators differ from the increased de- responseto this type of environmentalunpre- mands due to developmentof young in that dictability, but a behaviorsuch as time devot- they can vary from habitat to habitat or even ed to foragingshould be more proximallyflex- territory to territory. ible than a life-history trait (Warner 1980). When food abundance increases, extra en- Predation pressureis probably a primary se- ergy demands do not necessarilylead to in- lective force determining the "loafing" com- creases in energy expenditure in obtaining ponent. Utter (1971)found that alert perching food. Energy expenditure did increaseas the by Mockingbird males reached a maximum phasesprogressed, however, largelyas a result during the nestlingphase and was at a seasonal of increasesin flight time due to predatorsand minimum when the birds were unmated. He home-range expansion (Tables 3, 6). Home- too felt that this "perching" or "loafing" com- range expansionallowed birds to forage in ad- ponent was ultimately determinedby preda- ditional areas and suggeststhat food supply tion pressure.Furthermore, predation on pas- 158 Dot•GL;•s W. B•EDE•WEG [Auk, Vol. 100

serine eggs and nestlings is common in many rough measures:(1) qualitative statementscan temperate-zonespecies (36.2% of eggslaid are be made about each component and a balance prevented from developing into adults due to sheetdrawn up, and (2) DEE,,t, becauseit con- predation; Ricklefs 1969), and it is a common tains all production and activity costs,can serve assumption,although difficult to prove, that as an index of parental investment. organismsoccupied with tasks (e.g. mating, In Mockingbirds,territorial defenseis a male- foraging)experience higher predation rates than biased behavior involving primarily the de- those perched or sitting. In this study, the votion of energy resourcesand risk; egg pro- numberof timesthat perchedbirds caughtsight duction and incubation are done totally by the of predators before I did suggeststhat birds female, the former requiring energy,the latter, that devote more time to perching are better time; the protectionof offspringis skewed to- able to protect themselves,and their offspring, ward the male and demandsthe use of energy from predation. and involvesrisk; and the feeding of offspring Between-pair comparisonssuggest that vari- demands the use of time and energy for prey ation in territorial quality can have pro- capture and appears to be female-biased dur- nounced fitness effects. Pair number 1 were the ing the nestling phase but male-biased during first to form a pair bond and appeared to have the fledglingphase. This summarysuggests that the greatestresource density on their territory. there is a qualitative balance between parents They utilized smallerhome ranges,they spent in parental investment. Males have higher significantlyless time in flight than other pairs DEE,(.•than females over the totalbreeding cycle in the pre-nest-building (P < 0.05) and nest- (Table 6). Therefore,males may have a slightly ling phases(P < 0.05) (the only phaseswhere greater energeticinvestment in offspring than sample sizes were large enough to warrant females. comparisons), they appeared to make more Taken together, these two measuresindicate feeding visits to the nest per unit time, and that there is little sexual bias in parental in- they were the only birds to raise an offspring vestment (male and female investmentsare both that survived after leaving the nest. needed throughout the cycleif offspring are to SEASONAL SUMMARY OF PARENTAL be raised), and this is a major determinant of INVESTMENT the overwhelmingly monogamousmating sys- tem of Mockingbirds. A breeding-cyclesummary indicates that males have both a larger total power consump- INTERPOPULATION COMPARISONS tion and a larger power consumption due to productionand activity than do females(Table The behavioralcategories used in this study 6). This is causedprimarily by the extended match well only the flight categoryin Utter's period of fledgling care, which elevatedmale (1971) study of Mockingbirds. Mockingbirds power consumption during times when fe- breeding in California and New Jerseyspend males incubated. similar amountsof time in flight (Table 7). The Parentalinvestment may be defined as "any mean percentageof time that male birds in investmentby the parent in an individual off- Californiaspent in flight during the whole cycle spring that increasesthe offspring'schance of was 5.5%, whereas Utter's values for two male surviving at the cost of the parent's ability to populations were 7.3% and 6.0%. A compari- invest in other offspring" (Trivers 1972). Rel- son of home-rangesizes shows that birds in ative parental investmentplays a large role in New Jerseyhad slightlylarger average maximal the evolution of mating systems and mating home ranges(7.9 and 7.6 ha in New Jerseyvs. behaviors. In Mockingbirds it can be divided 6.5 ha in California), which may accountfor the into five components:territorial defense, pro- lower flight timesof the Californiapopulation. duction of eggs,incubation, protection of off- A major difference between birds in New spring, and ratesof feedingoffspring. Because Jerseyand Californiais that the changesin time these components have different units (i.e. spent in flight between phasesdo not parallel time, energy,or risk), it is difficult to distill a each other. New Jersey males reached a pri- single quantitative value for parental invest- mary peak with fledglingsand a secondarypeak ment made by males and females.Neverthe- while females incubated. In California, flight less, there are two approaches that can act as means peaked with nestlings, and a much January1983] MockingbirdTime-Energy Budgets 159

TABLE7. Interpopulationcomparison of percentage LITERATURE CITED of active day in flight. ASCHOFF,J., & H. POHL. 1970. Rhythmicvariations in energymetabolism. Fed. Proc., Fed. Amer. Utter's study (1971) Soc. Exp. Biol. 29: 1541-1552. This Popu- Popu- BENt, A. C. 1948. Life histories of North American Phaseof cycle studya lation I lation 2 nuthatches, wrens, thrashers, and their allies. U.S. Nat. Mus. Bull. 195. Unmated 3.6 6.1 3.2 Pre-nest-building 4.3 6.5 5.3 BRETT,J. R. 1971. Energeticresponses of salmon In cub ation 3.9 7.4 7.8 to temperature.A study of somethermal rela- Nestlings 8.0 6.4 5.3 tions in the physiologyand fleshwaterecology Fledglings 7.8 10.1 8.3 of sockeyesalmon (Oncorhyncus nerka). Amer. Mean % 5.5 7.3 6.0 Zool. 11: 99-113. CALDER,W. A., & J. R. KING. 1974. Thermal and Utter usedactual means, so I presentactual means for comparison. caloric relations of birds. Pp. 259-413 in Avian Biology,vol. 4 (D. S. Farnerand J. R. King, Eds.). New York, Academic Press. CAMPBELL, G. C. 1977. An introduction to envi- smallerpercentage of time was devotedto flight ronmentalbiophysics. New York, Springer-Ver- during incubation(œ = 3.9%) than was found lag. in New Jersey(7.4 and 7.8%). Ecologicalcauses 1981. Fundamentals of radiation and tem- for these differences remain obscure, as Utter peraturerelations. Pp. 11-40 in Encyclopediaof plant physiology-newseries, vol. 12A (O. L. gave no reasonsfor fluctuationsin flight times Lange,Ed.). New York, Springer-Verlag. between phases. ETTINGER,A. O., & J. R. KING. 1980. Time and Simpson(1978) also collecteddata on resting, energy budgetsof the Willow Flycatcher(Em- foraging,and flight for Mockingbirdsin Mary- pidonaxtrailli) during the breeding season.Auk land. No numerical results are presented, but 97: 533-546. time spent resting significantlyincreased dur- HANNA,W. C. 1924. Weightsof aboutthree thou- ing the fall and winter, while flight time sig- sandeggs. Condor 26: 146-153. nificantly increasedduring the nestling phase HART, J. $., & M. BERGER.1972. Energetics,water (P < 0.05). economy and temperature regulation during flight. Proc. 15thIntern. Ornithol. Congr.: 189- Hence, over a broad geographicrange Mock- 199. ingbirdsexpend greater amounts of energyand HINDS, D. S., & W. A. CALDER. 1973. Temperature time being activein the nestlingand fledgling regulationof the Pyrrhuloxiaand the Arizona phasesthan in other portions of the breeding Cardinal. Physiol.Zool. 46: 55-71. cycle.If increasedflight time due to nutritional HUE¾,R., & R. D. STEVENSON.1979. Integrating demandsis mandatory, then a greater degree thermalphysiology and ecologyof ectotherms: of constraint is imposed on the time budget a discussionof approaches.Amer. Zool. 19: 357- during these phases. Partitioning the neces- 366. saryflight time nearlyequivalently between the HUTCHINSON,G. E. 1965. The ecologicaltheater and the evolutionaryplay. New Haven, Connecticut, male and the female (Table 3) may be a tactic Yale Univ. Press. that minimizes the selectiveimpact of this con- KENDEIGH,S. C. 1969. Toleranceof cold and Berg- straint on the pair. mann's rule. Auk 86: 13-25. 1972. Monthly variations in the energy budget of the House Sparrow throughoutthe ACKNOWLEDGMENTS year. Pp. 17-44 in Productivity,population dy- namicsand systematicsof granivorousbirds (S. I thank WesleyWeathers for helpful advice and C. Kendeigh and J. Pinowski, Eds.). Warsaw, assistancein getting the project started. T. Mc- Polish Sci. Publ. Kewan, K. Whitelaw, and J. Vance helped in the mist KING,J.R. 1973. Energeticsof reproductionin birds. netting of birds. M. Murphy did the bomb calorim- Pp. 78-107 in Breedingbiology of birds (D. S. etry, and J. R. King, G. S. Campbell,J. N. Thomp- Farner,Ed.). Washington,D.C., Nat. Acad. Sci. son, J. N. Mugaas, and an anonymousreviewer pro- 1974. Seasonal allocation of time and en- vided valuable critiques of the manuscript. Partial ergy resourcesin birds. Pp. 4-70 in Avian en- fundingfor this studywas providedby a President's ergetics(R. A. Paynter, Jr., Ed.). Cambridge, UndergraduateFellowship at the University of Cal- Massachusetts, Publ. Nuttall Omithol. Club No. ifornia at Davis. 15. 160 DOUGLASW. BmDENWEG [Auk, Vol. 100

LACK,D. 1968. Ecologicaladaptations for breeding the Mockingbird, Mimus polyglottos(Abstract). in birds. London, Metheun. J. Elisha Mitchell Sci. Soc. 94: 106. LASKEY,A. R. 1962. Breedingbiology of Mocking- STEELS,R., & J. TORRIS. 1960. Principles and pro- birds. Auk 52: 370-381. cedures of statistics. New York, McGraw-Hill. MA•ONSY, S. A. 1976. Thermal and ecological TRtVERS, R. L. 1972. Parental investment and sex- energeticsof the White-crownedSparrow (Zo- ual selection.Pp. 136-179 in Sexualselection and notrichialeucophrys) using the equivalentblack- the descent of man (Bernard Campbell, Ed.). body temperature. Unpublished Ph.D. disser- Chicago,Aldine Publ. Co. tation. Pullman, Washington, Washington State UNSWORTH,M., & J. L. MONTEITH. 1975. Long- Univ. wave radiation at the ground. I. Angular distri- --, & J. R. KING. 1977. The use of the equiv- bution of incoming radiation. Quart. J. Royal alent black-bodytemperature in the thermal en- Meteorol. Soc. 101: 13-24. ergeticsof small birds. J. Thermal Biol. 2: 115-- UTTER,J. M. 1971. Daily energy expenditures of 120. free-living Purple Martins (Prognesubis) and MAYR, E. 1961. Cause and effect in biology. Sci- Mockingbirds (Mirnuspolyglottos) with a com- ence 134: 1501-1506. parison of two northern populations of Mock- MUCAAS,J. N. 1976. Thermalenergy exchange, mi- ingbirds. Unpublished Ph.D. dissertation.New croclimateanalysis, and behavioralenergetics of Brunswick, New Jersey,Rutgers Univ. Black-billed Magpies, Pica pica hudsonia.Un- WALSBERG, G. E. 1978. Brood size and the use of published Ph.D. dissertation.Pullman, Wash- time and energyby the Phainopepla.Ecology 59: ington, Washington State Univ. 147-153. PORTER,W. P., & D. M. GATES. 1969. Thermody- ., & J. R. KING. 1978a. The relationshipof the namic equilibria of animals with environment. external surface area of birds to skin surface area Ecol. Monogr. 39: 245-270. and body mass. J. Exp. Biol. 76: 185-189. RICKLEtS,R. E. 1969. An analysisof nesting mor- , & -- 1978b. The heat budget of in- tality in birds. SmithsonianContrib. Zool. 9: 1- cubating mountain White-crowned Sparrows 48. (Zonotrichia leucophrysoriantha) in Oregon. 1974. Energeticsof reproduction in birds. Physiol. Zool. 51: 92-103. Pp. 152-292 in Avian energetics(R. A. Paynter, WARNER, R.R. 1980. The coevolution of behavioral Jr., Ed.). Cambridge,Massachusetts, Publ. Nut- and life-history characteristics.Pp. 151-181 in tall Ornithol. Club No. 15. Sociobiology: beyond nature/nuture? (G. W. ROnINSON, D. E., G. S. CAMPBELL, • J. R. KING. Barlow and J. Silverberg, Eds.). Boulder, Colo- 1976. An evaluation of heat exchangein small rado, Westview Press. birds. J. Comp. Physiol. 105: 153-166. WILSON, E. O. 1975. Sociobiology.Cambridge, SIMPSON,B. S. 1978. Annual time expenditure by Massachusetts,Belknap. HARLAN'S HAWK (BUTEO JAMAICENSIS HARLANI): A VALID SUBSPECIES

DAVID P. MINDELL • Bureauof LandManagement, 4700 E. 72ndAvenue, Anchorage, Alaska 99507 USA

ABSTRACT.--Duringsurveys of 19 different rivers in central and southwesternAlaska, north of the Alaska Peninsula,all Red-tailed Hawks (Buteojamaicensis) seen were either B. j. harlani or harlani intergrades,demonstrating that harlani does have a breeding range exclusiveof other Red-tailedHawk subspeciesand is, therefore, a legitimate subspecies. The rangeof harlaniintergrades was widespreadand overlappedthe harlanibreeding range entirely. The findings presentedsuggest that the breeding rangefor harlaniand its intergrade forms, exclusiveof other Red-tailed Hawk subspecies,is: west to Norton Sound, eastto the AlaskaRange west slope,north to the tree line, and southto the Alaska Peninsula.Harlani may have diverged morphologicallyfrom other Red-tailed Hawks by isolation in a Pleis- tocene epoch glacial refugium, as has been suggestedfor other northern bird groups. The possibility of paripatric speciation, however, must also be considered.The characteristics of red in the tail (even small amounts) and white body plumage are discussedas likely intergradetraits. Received22 March 1982, accepted23 September1982.

OF the seven North American Red-tailed Hawk was simply a local color phase of the Hawk subspeciesrecognized by the American Red-tailed Hawk, which did not require a sci- Ornithologists' Union (1957, 1973), harlani is entific name of its own. At Ridgway's (1890) the most stronglydifferentiated and, together suggestion,the Harlan's Hawk was regarded with alascensisand perhaps umbrinus,the least as a subspeciesof the Red-tailed Hawk (A.O.U. well known. Extensivevariation of plumage 1895). Some, however, still considered Har- within subspecies,combined with consider- lan's Hawk a distinct species (Peters 1931). able subspecificbreeding-range overlap and Harlan's Hawk was again considereda species interbreeding, complicatesthe application of in 1957(A.O.U. 1957)but then reassignedsub- the rules of taxonomy. specific status in 1973 (A.O.U. 1973). Confusion concerning the Harlan's Hawk Because harlani interbreeds freely with form can be traced back as far as 1829, when neighboring subspecies,it is clearly unaccept- no less a person than John JamesAudubon de- able as a speciesunder the biologicalspecies scribedas Falcoharlani (after R. Harlan a phy- concept (Mayr 1969). Subspeciesby definition sician and naturalist) a hawk he had shot in must possessa geographicbreeding range ex- Louisiana and called the Black Warrior (Au- clusiveof conspecificsubspecies, however, and dubon 1830: Plate LXXXVI; 1840). Despite no suchrange was previouslyfound for harlani doubtsabout the eventualadult plumageof the (Taverner 1936, Gabrielson and Lincoln 1959, apparently immature type specimenspainted Godfrey, 1966). Taverner (1927: 8) analyzed by Audubon (Taverner1927), one type speci- recordsin the Yukon Territory and northern men in the British Museum showed the char- British Columbia and concluded that "If har- acteristic Harlan's Hawk mottled tail pattern lani is not racially pure here it seemshopeless (Hellmayr and Conover1949). It graduallybe- to look for racial purity in it anywhere." came evident that the Harlan's Hawk was a This paper presents new information from memberof the genusButeo and was relatedto the field showingthat harlanidoes have a dis- the Red-tailed Haw.k, but whether it should be crete breeding range exclusiveof other Red- considered a separate speciesor a subspecies tailed Hawk subspecies,along with discussion of the Red-tailed Hawk has remained in some of causesfor divergenceand intergrade traits. doubt. Another opinion was that Harlan's STUDY AREA AND METHODS

• Presentaddress: Department of Zoology, Brig- Subspecificidentifications were made in the field ham Young University, Provo, Utah 84602USA. during June, July, and August 1979, 1980, and 1981

161 The Auk 100: 161-169. January1983 162 DAVIDP. M•taDELL [Auk, Vol. 100

E] HARLANI OR HARLANI INTERGRADES (1979,1980) O HARLANI INTERGRADES (1981)

ß HARLANI (1981) YUKON RIVER

50 KM

tSKOKWIM RIVER

Fig. 1. Study area and distributionof Red-tailedHawk sightingsin southwesternAlaska during 1979- 1981.All Red-tailedHawks seen were either Buteo jama[•s• harlanior B.j. harlaniintergrades. Distinction between thesetwo groupswas made during 1981 only. Surveyswere conductedon all rivers shown.

in central and southwestern Alaska north of the Alas- Red-tailed Hawks having tails with various com- ka Peninsula (Fig. 1). Nineteen rivers were travelled binations of white and dark-brown mottling, freck- in the Kuskokwim, Yukon, and Unalakleet river ling, and streakingand with lessthan approximately drainages.Rivers in the Kuskokwim drainage in- 10% red color were counted as harlani. Harlani is cludedthe Kusk•kwim,Takotna, Tuluksak, Stony, characteristicallydark and lacks the predominating Selatna,Tatlawiksuk, Cheeneetnuk,Gagaryah, East red color in the tail that other Red-tailed Hawk sub- Fork George, and Oskawalik. Thosewest of or with- specieshave. Harlani frequently has a light or mot- in the Yukon River drainage included the Yukon, tled bib; it is not consistentlyfound, however. Light North, South, Chiroskey, Old Woman, Golsovia, Ot- spotting on the breast, back, and wing linings is also ter Creek, Canyon Creek, and Stuyahok. All rivers a good field mark, when present.Hawks with tails were in the range of boreal forest and between 60ø havingmore than 10% red colorbut still havingwhite and 65ø latitude and 154ø and 162ø longitude. Prin- and/or brown mottling and streaking were consid- cipal vegetationtypes were spruce-hardwoodforest, ered harlaniintergrades. Hawks with tailshaving less consistingof white and black spruce(Picea glauca, than 10% red color but with extensivebarring (n = P. mariana), paper birch (Betulapapyrifera), and bal- 2) were also classed as intergrades. Intergrades sam poplar (Populusbalsarnifera), shrub thickets, and showed wide variation in the amounts of red, mot- moist tundra (Viereck and Little 1972). Fuller de- tling, and barring in the tail and in the amount of scriptionsof the survey rivers are found in Mindell light breastcolor. Hawks with tails of more than 85% (1983). red color and lacking any of the distinctive harlani During 1979 and 1980, adult Red-tailed Hawks, mottling were classedas hawks with red-coloredtails. which I could seewell through 7 x 50 binocularsor Three Red-tailed Hawk subspecies,with tails pre- a 20x spotting scope,were classifiedas either being dominantly red in color, have apparentzones of con- or not being of the group harlani and harlani inter- tact with harlani. Calurusfrequently has a light col- grades. No attempt was made to distinguish inter- ored breast, although often washed with rufous. gradesfrom "pure" harlani.In 1981,Red-tailed Hawks Melanistic calurusindividuals can be distinguished were classifiedas either harlani,harlani intergrades, from harlani by their red-coloredtails or lack of mot- or hawks with red-coloredtails (Fig. 2). I used dorsal tling and lack of light color in the tails. Kriderii could views of the tail in classifying Red-tailed Hawks be mistaken for a light-phase harlani intergrade, as whenever possible. both may have tails with white and varying amounts January1983] Harlan'sHawk 163

harlani intergrade calurus

Fig. 2. Three forms of Red-tailed Hawk (Buteojamaicensis) found in Alaska. The drawings are typical representations;characteristics varied widely within eachof the forms,however. Widely spacedstipplings in the tails of the caIurus(entire) and intergrade (distal half) forms denote red color. of rufous color;however, krideriiare apt to be even in 1981, when a total of 82 Red-tailed Hawks paler than light harIani intergrades. AIascensis,al- was seen at close range on the Yukon, Kus- though slightly smaller and darker, is virtually in- kokwim, and Stony rivers. Of these, 73 (89%) distinguishable from calurusin the field (see Fried- were pure harlani and 9 (11%) were harlani in- mann 1950 for subspeciesdescriptions). Red-tailed tergrades.Ratios of harlani to intergradeswere hawks in juvenile plumage were not classified to subspecies. 89% to 11%, consideringthe KuskokwimRiver alone (n = 52), and 90% to 10% on the Yukon

RESULTS AND DISCUSSION River alone (n = 19). Field observations are an imperfect substitute for examination of pre- Subspeciesand intergrade identification.--A to- served specimens.Specimens collected during tal of 187 sightingsof adult Red-tailed Hawks migration or winter, however, are inappro- was made at a rangeclose enough to allow sub- priate for a determination of breeding range, specific identification in 1979-1981. All were and collectedAlaskan breeding birds are few. eitherharlani or harlaniintergrades (Fig. 1). One The field identifications probably provide a hawk had a tail about 70% red and 30% white conservativeestimate of the frequency of oc- with mottling,and two had tailsabout 60% red currence of intergrades, because they were and 40% white with mottling.The resthad tails based only on tail characteristics,and some that were 25% or lessred. The majority showed hawks with subtle intergrade charactersmay no discernable amounts of red. The Alaskan have been misclassified as harlani. breeding range found to be exclusiveto Red- Taxonomicstatus and breedingrange.--Be- tailed Hawks of the harlani subspeciesran from causeharlani freely interbreedswith other Red- Farewell at the east slope of the Alaska range, tailed Hawk subspecies in zones of contact north to Tanana and the Yukon River, west to (Taverner1927, Lowe 1978),it cannotbe a good Norton Sound,and southto St. Marys and An- species.Treatment of harlani as a semispecies iak. (sensuAmadon 1966) would assume partial re- I differentiated between individuals of ap- productive isolation in contactzones with oth- parently pure harlani and partial harlani (inter- er subspecies.The evidenceavailable does not grades) ancestryon the basis of tail plumage suggestthis; more data on subspecificratios 164 DAVIDP. MI•DELL [Auk,Vol. 100

ß HARLANI OR HARLANI INTERGRADE

o CALURUS *ALASCENSIS ALASK•

]Fig.3. Distributionof Red-tailedHawk subspeciesin Adaskaand adiacentCanada. Locations for c½l,rus and ½l½sce•sisare basedon specimensonly. Records not fromthe presentstudy are from: Grinnell (]909), ,Swarth(1911), Cureming (1931), Gabrielson and Lincoln (1959), Wd]iamson eta]. (1965),Eesse] and ,Springer (1966),White and Haugh (1969),Lowe (1978),Ritchie and Curato]o(1980), and from D. Weir, C. M. White, ,S.Ambrose, and M. Amaral (pets. comm.).

within pairs in contact zones, however, are morphs of the same subspecies,one would ex- needed. This leaves two possibilities:harlani pect to see both varieties in mixed occurrence is either a distinct subspeciesof the Red-tailed through most of their range. Overlap in breed- Hawk or a colorphase of anothersubspecies, ing range has been recorded (Taverner 1936, either calurusor kriderii(Mayr 1942:150). Gabrielsonand Lincoln 1959),although to only Two factorsmade the possibilityof harlani a minor degree. The Alaskan harlani breeding being a colormorph of krideriiinitially attrac- population and the kriderii population of the tive. Harlaniand krideriishare the morpholog- north-central U.S. and southern prairie-prov- ical characterof oftenhaving large amounts of inces of Canada remain distinct, with little white in the tail, a trait unique among Red- mixing (seeFriedmann 1950 for generalkriderii tailed Hawks. Harlani and krideriioverlap in range). The same can be said of calurus and the locationof corewintering areas on the Great harlani,although their zone of breedingrange Plains of the south-central United States. Dur- contact is broader than that for harlani and kri- ing migration,harlani passes over the breeding derii. As distinct populations with different range of kriderii in southern west-central Can- geographiccore areas,harlani, kffderii, and cal- ada and adjacentportions of the United States, urus are good subspecies.The similarities be- and overlapsthe generalmigratory route of kri- tween harlani and kriderii suggesta close his- derii as well. Calurushas a stronglymarked toricalaffinity. melanistic phase and can resembleharlani; cal- The fact that all individuals seen in central urus, however, retains the red tail color. and southwestern Alaska were either harlani or Despite similarities, however, harlani does harlani intergradesclearly indicates the exis- not qualify as a polymorph variant of either tence of a breeding range for harlani in which kriderii or calurus.Polymorphism is an intra- other Red-tailed Hawk subspeciesdo not occur populationphenomenon, and the genes in- (Fig. 3). Subspeciesby definition must possess volved in polymorphismhave, in general,con- an exclusivegeographic breeding range, and, spicuousdiscontinuous effects (Mayr 1970:90). thus, the field observationsfrom Alaska pro- Thus, if kriderii and harlani were different vide the information that has previously been January1983] Harlan'sHawk 165 lacking to support subspecificstatus for harlani by him had someless visible harlani tail char- (cf. Mayr and Short 1970:38, 89). acteristics. The presence of intergrades within the Five specimensidentified by Gabrielsonand breeding range of a subspeciesdoes not in- Lincoln (1959) indicate that harlani's zone of validate the exclusivenessof the range. Inter- contact in central and eastern Alaska is with breeding can be expected to occur where two calurus, which is generally the predominant subspeciescome into contact,and intergrade subspeciesthroughout most of westernNorth forms will often penetrate farther into the op- America (Friedmann 1950,Brown and Amadon posing parental ranges than will pure forms 1968). Alascensis,a smaller and richly pig- (Mayr 1942, 1963). The zone of intergradation mented subspeciesof southeastAlaska and for harlani entirely overlapsits breeding range coastalBritish Columbia (Grinnell 1909, Cum- within the study area (Fig. 1) and, I suspect, ming 1931),probably contacts both harlani and in the remainder of its range as well. Inter- calurus inland. Borealis, with the core of its grades in the study area occured at a south- range in easternNorth America, has poorly western extreme of Pilot Village, east to the known western range limits and no demon- McGrath vicinity, and north to Galena. Inter- strated contact zone with breeding harlani. gradesoutside the studyarea are recordedfrom Todd's (1950) B. j. abieticolamay replacebo- Fairbanks (Lowe 1978), Atlin, British Colum- realisin the spruce-firbelt of Canada,although bia, and two southwestern Yukon Territory lo- it is infrequentlyrecognized. Kriderii, generally cations (Taverner 1927). Harlani remains a of the centralNorth American plains, may con- meaningfulsubspecies, regardless of the wide tact harlani, although the evidenceis small, intergradation zone, due to its recognizability based on possible intergrades discussedby in migration and on wintering groundsand the Taverner (1927) and a juvenile specimenfrom usefulnessof this trait in studying geographic Eagle, Alaska identified by variation in the species. Even individuals of (Bailey 1916). partialharlani ancestry can often be recognized The summerand breeding recordsof calurus as such. closestto the easternlimit of the present study Harlani makes the longest migration of any area are from the Chatenika River and Minto Red-tailed Hawk subspeciesand passesover Lakes west of Fairbanks, the Kasilof River and (leapfrogs)some populations of other subspe- Tustumena Lake on Kenai Peninsula (Gabriel- cies. This further illustrates harlani's unique- son and Lincoln 1959), and in the vicinity of ness and weakens any considerationof it as a Ferryalong the FairbanksHighway (Kessel and color morph of another subspecies.Differences Springer 1966). Some of these sightings may in length, duration, and chronologyof migra- have been of harlani intergrades. Specimens tion for different subspeciesinfluence differ- identifiedas calurus by Gabrielsonand Lincoln ences in physiological timing of breeding (1959)came from SalchaSlough (50 km south- cycles.Vaurie (1961a) and White (1968) point east of Fairbanks), Chitina River (a Copper out the importance of migratory habits in River tributaryin southcentral Alaska), Denali showing a taxon's nomenclatural distinctive- National Park, and Circle. Additional calurus ness. reports are from the ChandalarRiver, Robert- The exclusivebreeding range of harla,i is son River, Circle, Gulkana, and Chitina River not strictlylimited to the presentstudy area, (Gabrielson and Lincoln 1959). althoughit appearsto includethe core.Reports In the only detailedprevious study of Red- of Red-tailed Hawks identified to subspecies tailed Hawks in Alaska, Lowe (1978) reports in Alaska and adjacent Canada are few, and that, of 285 Red-tailed Hawk sightingswithin sometimes unreliable. Twitchel]'s reports of a 300-mi2 study area in the Fairbanksvicinity, calurus from the Innoko, Nushagak, lower 83.9% were pure harlani. Although Lowe clas- Kuskokwim, and Iditarod rivers (Gabrielson sified the remainder as either calurus, borealis, and Lincoln 1959) around 1917 are not included or unknown, he reveals the mixed characterof in Fig. 3. These locations are in or near the mostof them in stating(1978: 32), "Most of the presentstudy area and conflictwith recentob- hawks classified as borealis or calurus were not servations. The reports make no mention of typicalof theseforms . . . manyhad tails only harla, intergrades, and it seemspossible that washed with red . . . and portions being mot- Red-tailed Hawks with red-colored tails seen tled, barred or white." Lowe gives a photo- 166 DAVIDP. MI•DELL [Auk, Vol. 100

graph of an adult labeled as calurusthat clearly variants within a widespread species are allo- has intergrade characters. patric and paripatric (also alloparipatric) (En- In the Yukon Territory, harlani are appar- dler 1977) or geographicand semigeographic ently predominant in the southwesternportion (Mayr 1942, 1963). Allopatric speciation in- of the province(Yukon River drainage),harlani volves the isolation and subsequent diver- intergrades and hawks with red tails are mixed genceof a population causedby extrinsic, geo- on the west slope of the Ogilvie Mountains, graphic barriers. Allopatric speciation due to and hawks with red tails are common northeast glaciation and concomitant biotic refugia has of the Ogilvie Mountains in the Peel and Por- been hypothesizedby Hult6n (1937) to explain cupine river drainages (David Mossop pers. phytogeographic patterns and the present comm.). More fieldwork and specimens are rangesof Alaskan plants. Pleistoceneglaciation needed throughout northwestern Canada and as an isolating mechanismallowing speciation Alaskato fill in detailsof Red-tailedHawk geo- among northern birds has been discussedby graphic variation. Austin (1932), Rand (1948), Pitelka (1950), Dru- Reportsof harlani occurthroughout central ry (1953), Cade (1955), Fay and Cade (1959), and easternAlaska south of the BrooksRange. and Selander (1965). Much supportive infor- The northernmostrecord is from the Porcupine mation on geology,oceanography, palynology, River near the Alaska-Yukon border (Ritchie and paleontology around the Bering Land and Curatolo 1980). Additional records are from Bridge is presentedby Hopkins (1967). the Yukon River between Circle and the Forty- The morphologicallydistinctive harlani pop- mile River confluence(White and Haugh 1969), ulation may also have developed through iso- the Fairbanks vicinity (Kessel and Springer lation in a Pleistoceneepoch glacial refugium. 1966, Lowe 1978), Ferry and Denali National The Bering Sea-YukonRefugium and its pos- Park (Kessel and Springer 1966), Tok and sible relationshipwith geographicvariation in Northway (D. Weir pers. comm.), the Charley the Rough-leggedHawk (Buteo )has River (S. Ambrose, M. Amaral, and C. M. White been discussedby Cade (1955). A similiar ar- pers. comm.), and Kechumstukand McCarthy gument for the evolution of harlani in Alaska (Gabrielson and Lincoln 1959). Williamson et can be made. A northern population of B. ja- al. (1965)report seeingonly harlani during the maicensismight have becomeisolated by sur- breeding seasonat unspecificlocations in the rounding ice sheetsin a land area encompass- area "west of the Copper River Valley, includ- ing southwesternAlaska, part of easternSiberia ing the Cook Inlet region." They also record and the interlying Bering Sea. Morphological nestsof harlani near Anchorageand at Hidden divergence occurring during isolation devel- Lake on the Kenai Peninsula and two speci- oped more rapidly than did any intrinsic iso- mens collected at mile 20 on the Tok Cutoff lating mechanismsthat might have prevented Road. breeding with the parental stock.Thus, harlani The findingssuggest that the breedingrange was still able to breed with Red-tailed Hawks for harlani and its intermediate forms, exclu- from unglaciated regions in southern North sive of other Red-tailed Hawk subspecies,can America when corridorsopened during the re- be outlined as follows: west to Norton Sound treat of the last glaciation. Evolution of har- and the western limit of the taiga approaching lani's migration is another matter, as present the Yukon-Kuskokwim Delta; east to the Alas- migration patterns for all specieshave appar- ka Rangewest slope;north at leastto the Koyu- ently evolved since the end of the last glacia- kuk River mouth and probably up to tree line tion (Moreau 1972, see Dingle 1980). south of the Brooks Range; and south to the Buteojamaicensis and B. buteoare considered Iliamna Lake region. Even if a few other Red- by some to be members of the same superspe- tailed Hawks are located in this area, domi- cies group (Voous 1960, Mayr and Short 1970), nation of harlani in subspecificratios would although relationships in this group remain not be changed.Breeding sites for harlanifound obscure. Harlani isolation in the Bering Sea- during the present study representa westward Yukon Refugium suggeststhe possibility of extension in the Red-tailed Hawk's reported contact(gene flow) during that period with the range (A.O.U. 1957,Brown and Amadon 1968). easternmostform of B. buteo,japonicus. Buteo Causesfor divergence.--The two basic pat- buteo has not been recorded in the Soviet Union terns of speciation resulting in geographic east of the Kolyma River system, including January1983] Harlan'sHawk 167

Kamchatka Peninsula (Vaurie 1961b); the east- age for hawks with harlanicharacteristics may ern limit ofjaponicus,however, may be incom- also be the result of subspecificinterbreeding, pletely known. similar to red colorin the tail. The greaterper- Glaciation and paleoclimatologicalrefugia centageof light-breastedforms among inter- represent only one of the two broad patterns grades than among harlani seen during the of speciation.Although Mayr (1963) has pre- present study suggeststhis possibility. Inde- viously argued against the likelihood of pari- pendent segregationof such traits can also be patricspeciation, Endler (1977) presents a strong expectedand would accountfor the three hawks casefor its feasibility and further demonstrates seen in 1981 with light breasts but no red in that it is not possibleto distinguishbetween the tail. Consideringthe degreeof color vari- the result of paripatric or allopatricspeciation ation seen in other Red-tailed Hawk subspe- simply by interpreting a particulargeographic cies, however, the existence of light phase pattern. Under the paripatric mode of specia- "pure" harlani seemspossible. Observations tion, harlaniwould have developedthrough the on the parentage of light phase harlani are formationof clines,without any barriersto gene needed. flow occurringin the contiguousRed-tailed Hawk range. Isolation would be more by dis- ACKNOWLEDGMENTS tance from similar conspecificsthan by geo- I conductedthe field portion of this study while graphical, ecological,or temporal factors(Hux- working for the Bureau of Land Management, An- ley 1939, Timofeef-Ressovsky1940, Mayr 1942, chorageDistrict, primarily, and for the U.S. Fish and Murray 1972, Endler 1973). Wildlife Service, Office of EndangeredSpecies, An- The wide harlani intergraderange may im- chorage,secondarily. I particularlythank the Bureau ply incompletedivergence of morphological of Land Managementfor their interestand support. characters,assuming the variance arose through I gratefullyacknowledge assistance during the river primary intergradation and paripatric differ- surveysfrom RichardDotson, Lance Craighead, Bruce Campbell,and RobertSmall. This paperwas greatly entiation, or, if one assumes the events of sec- improved by thoughtful review comments from ondary intergradation and allopatric specia- Amadeo Rea, Dean Amadon, DouglasWeir, Clayton tion, the wide intergraderange may suggest White, Ralph Palmer, and Paul Matray. that the intergrades are not "unbalanced" enoughto producea well-definedintergrade LITERATURE CITED zone (seediscussion of hybrid zonesin Mayr AMADON,D. 1966. The superspeciesconcept. Syst. 1963, Endler 1977). In other words, the inter- Zool. 15: 246-249. grade charactersdo not cause a selective dis- AMERICAN ORNITHOLOGISTS' UNION. 1895. Check- advantage,thus allowing breedingsuccess and list of North American birds, second edition. spread of intergrade forms. Just as one would Amer. Ornithol. Union. expect disadvantageousgenetic combinations 1957. Check-list of North American birds, to be selectedagainst, one could expect that fifth edition. Baltimore, Maryland, Amer. Or- charactersnot disadvantageouswould pene- nithol. Union. trate farther into the parent populations.The 1973. Thirty-second supplement to the American Ornithologists' Union check-list of characterof red in the tail may fit the latter North American birds. Auk 90: 411•419. description.This could explain the frequent AUDUBON,J. j. 1830. Birds of America, Elephant occurrence of small amounts of red in the tails folio, vol. I. of adults that are otherwise"pure" harlani in 1840. The Birds of America, vol. I. New character.Friedmann (1950) describesharlani's York, GeorgeR. Lockwood& Son. tail as having "... from scarcelyany admix- AUSTIN, O. L., JR. 1932. The birds of Newfound- ture of ochraceous to a definite cinnamoneous land Labrador. Nuttall Ornithol. Club. MeT. 7. wash .... "It is questionablethat pure harlani BAILEY, B. H. 1916. Krider's Hawk in Alaska. Auk has any red color in the tail, and the not un- 33: 321. common"cinnamoneous wash" may be an in- BROWN,L., & D. AMADON. 1968. Eagles, hawks and falcons of the world, vol. 2. New York, tergrade trait. McGraw-Hill Book Company. On the Yukonand Kuskokwimrivers during CADE,T.J. 1955. Variation of the commonRough- 1981, 3 (6.5%) of 46 harlani seen were light leggedHawk in North America. Condor 57: 313- breasted,while 5 (41.6%)of 12 intergradesob- 346. servedwere light breasted.Light body plum- CUMMING, R. A. 1931. Some birds observed in the 168 DAVID P. MINDELL [Auk, Vol. 100

Queen Charlotte Islands, British Columbia. American birds. Publ. Nuttall Ornithol. Club. Murrelet 12: 15-17. No. 9. DINGLE,H. 1980. Ecologyand evolution of migra- MINDELL, D. P. 1983. Nesting raptors in south- tion. Pp. 1-101 in Animal migration, orientation western Alaska, status, distribution and aspects and navigation (S. A. Gauthreaux, Ed.). New of biology. Tech. Publ. Anchorage,Alaska, Bu- York, Academic Press. reau Land Mgmt., AlaskaState Office. DRURY, W. H., JR. 1953. Birds of the Saint Elias MOREAU, R. E. 1972. The Palearctic-African bird Quadrangle in the southwesternYukon Terri- migration systems.New York, Academic Press. tory. Can. Field. Natur. 67: 103-128. MURRAY,J. J. 1972. Genetic diversity and natural ENDLER,J. A. 1973. A study of morph-ratioclines. selection.Edinburgh, Scotland, Oliver and Boyd. Unpubl. Ph.D. dissertation, Edinburgh, Scot- PETERS,J. L. 1931. Check-listof birds of the world, land, Univ. of Edinburgh. vol. 1. Cambridge,Massachusetts, Harvard Univ. 1977. Geographicvariation, speciation,and Press. dines. Princeton, New Jersey,Princeton Univ. PITELI

--, & J. R. HAUGH. 1969. Recent data on sum- WILLIAMSON, F. S. L., L. J. PEYTON, & M. E. ISLIEB. mer birds of the upper Yukon River, Alaska, and 1965. New distributional and overwintering adjacentparts of the Yukon Territory, Canada. records of birds from south-central Alaska. Con- Can. Field Natur. 83: 257-271. dor 67: 73-79.

The American Ornithologists'Union will hold its Annual Meeting in New York from 26 Septemberto 1 October, 1983. A formal call for papers will be mailed to membersby mid-April with the deadline for submission of abstractsfor papers and poster sessionsof 15 June 1983. For further information on the scientificprogram, contactDr. George Barrowdough, Department of Ornithology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024.

Fellowsand ElectiveMembers are reminded that nominationsfor ElectiveMembers or Fellows may be submitted to the Secretaryon the prescribedform up until five months prior to the opening of the next dated meeting. The deadline for 1983 is 26 April. Nominations for Vice President and three Elective Councillorsmay be made in writing to the Secretaryat any time prior to the annual meeting.

The American Ornithologists'Union invites applicationsby I March 1983 for the following competitive research awards: JosselynVan Tyne Memorial Fund supportsprojects in any area of ornithology,with awards of a few hundred to a thousand dollars each. Successfulapplications usually focus on one or a few carefully de- fined, incisive questions,rather than on generalbiology or management. AlexanderWetmore Memorial Fund supportsprojects involving systematicsand/or the taxonomyof birds and, secondarily,studies of neotropicalbirds. Awards are of a few hundred to a thousanddollars. Suc- cessfulapplications are usuallybuilt arounda well-focused,interesting question. Herbert and Betty Cames Awards are intended to promoteexcellence in the scientificstudy of birds by permitting individuals an opportunity to expand the design of field or laboratory work and to explore new methods of data analysis. Applicationsare judged on the basis of scientificmerit, importance,and originality. One or two awards of $1,000-$2,000are made. Students,amateurs, and otherswith limited or no accessto major grantingagencies are encouragedto apply for the Van Tyne or Wetmore Awards, while the Carnes Award is open to all individuals. Appli- cation forms and further information on the Van Tyne and Wetmore Awards may be obtained from Dr. Erica Dunn, 30 Davidson Road, Aurora, Ontario L4G 2B1 Canada; applicationsfor the Cames Award shouldbe addressedto Dr. Ellen D. Ketterson,Department of Biology,University of Indiana, Bloomington, Indiana 47405 USA. 24 PAGEET AL. [Auk, Vol. 100

Bird Observatory. Ralph Widrig, Steven Cellman, tain Plover social systems.Living Bird 12: 69- Barbara Swarth, Susan Peaslee, Erica Buhrman, Don- 94. na Vaiano, Paul Neal, Sharon Goldwasser, Dennis HAYS, H., & M. LECI•o¾. 1971. Field criteria for Parker, John Harris, David Shuford, Stuart Johnson, determining incubation stage of the Common John and Ricky Warriner, Stephanie Zeiler, Elliot Tern. Wilson Bull. 83: 425-429. Burch, Brett Engstrom,Leslie Smith, BernadetteAl- HOLMES,R.T. 1966. Breeding ecologyand annual len, Greg Calllet, Tom Pogson,Tom Cassidy, Geoff cycle adaptationsof the Red-backedSandpiper Geupel, Doug Burgess,Judy Wagner, Kim Sullivan, (Calidrisalpina) in northern Alaska. Condor 68: JanetKjelmyr, and Tad Theimerhelped with the field 3-46. work and Philip Henderson, Carolyn Frederiksen, 1972. Ecologicalfactors influencing the and Frances Bidstrup provided valuable observa- breeding season schedule of Western Sandpi- tions. John and Ricky Warriner generouslyshared pers (Calidrisrnauri) in subarcticAlaska. Amer. with us their knowledge of Snowy Plovers at Mon- Midi. Natur. 87: 472-491. terey Bay, and CharlesSimis provided weather data. MAYFIELr•,H. F. 1961. Nesting successcalculated Jan Simis allowed us to camp on her land and ex- from exposure.Wilson Bull. 73: 255-261. tended to us very warm hospitality while we were --. 1975. Suggestionsfor calculatingnest suc- there. David Ainley, Peter Connors,David DeSante, cess. Wilson Bull. 87: 456-466. Marshall Howe, Harriet Huber, Frank Pitelka, and OI•INO, L. W., & M. L. KNOrrSON. 1972. Monogamy Chris Ribic provided helpful comments on the and polyandry in the SpottedSandpiper. Living manuscript.We thank theseorganizations and peo- Bird 11: 59-72. ple for their generosity.This is ContributionNo. 189 PITELKA, F. A., R. T. HOLMES, & S. F. MACLEAN. of Point ReyesBird Observatory. 1974. Ecologyand evolution of social organi- zation in arctic sandpipers.Amer. Zool. 14: 185- LITERATURE CITED 204. RITIINGHAUS,H. 1975. Seeregenpfeifer,Charadrius ANDEI•SSOtW,M., & C. G. WI•

The American Ornithologists'Union offers several Marcia Brady Tucker Travel Awards to help defray expensesof outstanding students wishing to present researchpapers at the society'sCentennial Meeting in New York (26 September to 1 October, 1983). To apply, send the following materials to Douglas Mock, Departmentof Zoology,University of Oklahoma,Norman, Oklahoma73019 by 1 May 1983:(i) ex- panded abstractof the paper (max. 3 pp, typed, double-spaced;to include methods,major results,and scientificsignificance), (ii) curriculumvita, (iii) your anticipatedbudget of expenses,and (iv) a letter of support from the academicadvisor supervisingyour research(which should be mailed by itself). Please note that you must provide seven(7) copiesof everything,including the letter of support,because each application is reviewed by the AOU Student Awards Committee's seven members. Also, be advised that applying for a MBT Travel Award does not guaranteea place on the Scientific Program--each student must make arrangementsfor that as per the instructions given with the Call for Papers announcement. 32 w. BRUCEMcGILLIVRAy [Auk, Vol. 100

1978. Seasonal variation in reproductive SCHIFFERLI,L. 1976. Weight conditions in the output of House Sparrows:the determinationof HouseSparrow particularly when breeding.Un- clutchsize. Ecology59: 1189-1199. published Ph.D. dissertation, Oxford, Oxford NEWTON,I. 1966. Fluctuationsin the weight of Univ. bullfinches. Brit. Birds 19: 89-100. SEEr, D.C. 1969. Food, feeding rates and body NORBERG,R. •. 1981. Temporaryweight decrease temperature in the nestling House Sparrow in breeding birds may result in more fledged (Passerdomesticus) at Oxford. Ibis 111: 36-47. young. Amer. Natur. 118: 838450. SMITH,J. N.M. 1981. Does high fecundityreduce PIANKA,E. R., & W. S. PARKER.1975. Age specific survival in Song Sparrows?Evolution 35: 1142- reproductivetactics. Amer. Natur. 109: 453-464. 1148. PINKOWSK•,B.C. 1977. Breeding adaptationsin , & D. A. RUFF. 1980. Temporal spacingof the Eastern Bluebird. Condor 79: 289-302. broods, brood size and parental care in song PINOWSKA,J. 1979. The effect of energy and build- sparrows (Melospiza melodia). Can J. Zool. 58: ing resourcesof females on the production of 1007-1015. House Sparrows Passerdomesticus L. popula- SNELL,T. W., & C. E. KINe. 1977. Lifespan and tions. Ecol. Polska 27: 363-396. fecunditypatterns in rotifers:the costof repro- R•CKLEES,R. E. 1974. Energeticsof reproduction duction. Ecology31: 882490. in birds. Pp. 152-292 in Avian energetics(R. A. STEARNS,S.C. 1976. Life-history tactics:a review Paynter, Jr., Ed.). Publ. Nuttall Ornithol. Club of the ideas. Quart. Rev. Biol. 51: 3-47. No. 15. SUMMERS-SMITH,D. 1963. The House Sparrow. SAPPINGTON,J. N. 1975. Cooperative breeding in London, Collins. the House Sparrow (Passerdomesticus). Unpub- WILL•AMS, G.C. 1966. Natural selection, the costs lished Ph.D. dissertation, Mississippi State, of reproductionand a refinementof Lack'sprin- Mississippi, Mississippi State Univ. ciple. Amer. Natur. 100: 687-690.

Now AVAILABLE

WorldInventory of Avian Skeletal Specimens. 1982. D. ScottWood, RichardL. Zusi, and Marion Anne Jenkinson. WorldInventory of Avian SpiritSpecimens. 1982. D. ScottWood, Richard L. Zusi, and Marion Anne Jenkinson. Published by The American Ornithologists'Union and The Oklahoma BiologicalSurvey, the cost of each inventory is $25.00, including surface mail postage.For air mail, add $5.00 (U.S.), $5.50 (Canada, Alaska, Hawaii), $9.00 (Mexico, Central America), $15.50 (Europe, South America), or $22.00 (Asia, Australia, New Zealand, Africa). Make checkor money order, in American dollars, payable to University of Oklahoma. Order from Dr. Gary D. Schnell, Oklahoma BiologicalSurvey, Sutton Hall, University of Oklahoma, Norman, Oklahoma 73019, U.S.A.

The First Conferenceon Birds Wintering in the MediterraneanRegion will be held on 23-25 February1984 at Aulla, Italy. Emphasiswill be on the ecology,ethology, distribution, and migration of birds wintering in this region. For further information and preregistrationmaterials contact Dr. Almo Farina, Museum of Natural History of Lunigiana, Fortezzadella Brunella, 54011Aulla, Italy.

The Raptor ManagementInformation System(RMIS) is a collectionof published and unpublishedpapers, reports, and other works on raptor management and human impacts on raptors and their habitats. It currentlyconsists of nearly 2,500 original papers, 160 keyworded notecarddecks comprised of 15,000key paragraphsfrom the original papers,and a computerprogram to retrieve partially annotatedbibliographies by species,by keyword, or by any combinationof keywordsand/or species. A geographicalindex is under development, and new papers are added as they are received. Originally designed to facilitate land-use planning and decisionmakingby governmentagencies and industry, the RMIS has since grown into a powerful researchand environmentalassessment tool for scholars,students, consultants, as well as land managersand their staff biologists. For more information write Dr. Richard R. Olendorff, U.S. Bureau of Land Management,2800 Cottage Way, Sacramento,California 95825 U.S.A., or phonecommercial (916) 484-4701or throughthe FederalTelephone System 468-4701. January1983] NestAttendance in PaintedSnipes 55

MAYFIELD,H. F, 1979. Red Phalaropesbreeding PARMELEE, D. F., & R. B. PAYNE. 1973. On mul- on BathurstIsland. Living Bird 17: 7-39. tiple broodsand the breedingstrategy of arctic ORING,L. W., & M. L. KNUDSON.1972. Monogamy Sanderlings.Ibis 115: 218-226. and polyandryin the SpottedSandpiper. Living RIDLEY,M.W. 1980. The breedingbehaviour and Bird 11: 59-73. feeding ecologyof Grey PhalaropesPhalaropus --, & S. J. MAXSON. 1978. Instances of simul- fulicariusin Svalbard.Ibis 122:210-226. taneouspolyandry by a SpottedSandpiper Ac- SCHAMEL,D., & D. TRACY. 1977. Polyandry, re- tiris rnacularia. Ibis 120: 349-353. placementclutches, and site tenacity in the Red OSBORNE,D. R., & G. R. BOURNE. 1977. Breeding Phalarope (Phalaropusfulicarius) at Barrow, behavior and food habits of the Wattled Jacana. Alaska. Bird-Banding 48: 314-324. Condor 79: 98-105.

IMPORTANT NOTICE TO AUTHORS

Effectivewith this issue (January1983), The Auk will adhere to use of the commonand scientificnames of North American birds as listed in the Thirty-fourth'Supplement to the American Ornithologists' Union Check-List of North American Birds (1982 Auk 99(3): 1CC-16CC). Please use this nomenclaturewhen pre- paring manuscriptsfor submissionto TheAuk. Exceptionsshould be justified in writing to the Editor. January1983] SharpbillBehavior 125

1960. Life histories of Central American ST•LES,F. G., & L. L. WOLF. 1979. The ecologyand birds, vol. II. Pacific Coast Avifauna No. 34. evolution of lek mating behavior in the Long- 1969. Life histories of Central American tailed Hermit hummingbird. Ornithol. Monogr. birds, vol. III. Pacific Coast Avifauna No. 35. No. 27. SLUD, P. 1964. The birds of Costa Rica: distribu- TRAMER,E. J., & T. R. KEMP. 1980. Foraging ecol- tion and ecology. Bull. Amer. Mus. Nat. Hist. ogy of migrantand residentwarblers and vireos 128. in the highlands of Costa Rica. Pp. 285-296 in SNow, B. K. 1970. A field study of the Bearded Migrant birds in the Neotropics (A. Keast and Bellbird in Trinidad. Ibis 112: 299-329. E. S. Morton, Eds.). Washington,D.C., Smith- 1972. A study of the Calfbird Perissosc- sonJan Inst. Press. phalus tricolor. Ibis 114: 139-162. WETMORE,A. 1972. The birds of the Republic of 1977. Territorial behavior and courtship of Panama, vol. 3. Smithsonian Misc. Coil. No. 150. the male Three-wattled Bellbird. Auk 94: 623- 645.

The FrankM. ChapmanMemorial Fund givesgrants in aid of ornithologicalresearch and alsopostdoctoral fellowships.While thereis no restrictionon who may apply, the Committeeparticularly welcomes and favorsapplications from graduatestudents; projects in game managementand the medicalsciences are seldomfunded. Applications for projectsin 1983should be submittednot later than 15 February;pros- pectiveapplicants and advisorsshould note there will shortlybe a changein the Chapmanmeeting and deadlineschedule. Application forms may be obtainedfrom the FrankM. ChapmanMemorial Fund Com- mittee, The American Museum of Natural History, Central Park West at 79th St., New York, New York 10024. Chapmangrants during 1982, totalling $40,453 with alcyon);Carlos A. Delannoy, Breedingbiology and a mean of $499, were awarded to: Jonathan L. At- ecologyof the PuertoRican Sharp-shinned Hawk wood, Speciation in the Black-tailed Gnatcatcher (Accipiter striatus venator); Kim Craig Derrickson, (Polioptilamelanura) complex; James C. Bednarz, Analysisof the Mockingbird (Mimuspolyglottos) vo- Cooperativepolyandry in theHarris' Hawk; CraigW. cal repertoire for behavioral and situational corre- Benkman,Food availability, foraging efficiency, and latesand seasonaltrends; David F. DeSante,Stability the regulation of crossbills(Loxia) in eastern North and dynamicsof a subalpinebreeding bird commu- America;Thomas K. Bicak,Food resources and forag- nity; Robert JackDowsett, Conservationplan for the ing behaviorof Long-billedCurlews (Numenius ameri- relict evergreenforests of Malawi and investigation canus) in western Idaho; David Edward Blockstein, of bird populationstherein; Paul J. DuBowy, Optimal Reproductivebehavior and parental investment in foraging, resource partitioning, and community the Mourning Dove, Zenaidamacroura; Sharon Ann structure of western North American Anatini; Patrick Brady, Effect of habitat size on the breeding and J. Dugan, Investigation of the information-center wintering ecology of the Ovenbird (Seiurusauro- hypothesisin tree-nestingherons; Bonita C. Eliason, capillus);Gregory S. Butcher,Sexual differences in the Mating system,parental care, and individual repro- color and behavior of the Northern Oriole; William ductive strategiesin the Blackpoll Warbler, Den- J. Carmen, Juvenile dispersal, flocking behavior, droica striata; Keith William Emmerson, Bird com- and habitat use in the CaliforniaScrub Jay (Aphelo- munity of the laurel forest of Tenerife (Canary Is- comacoerulescens californica); John H. Carothers,For- lands);Robert C. Fleischer,Host choiceby individual agingefficiencies in a nectar-feedingguild of Hawai- female Brown-headed Cowbirds; Frank B. Gill and ian honeycreepersat a commonfood source;Ralph Douglas Wechsler, Evolution of avian mating sys- V. Cartar,Incubation behavior of the White-rumped tems: breeding biology of a promiscuoustropical Sandpiper (Calidrisfuscicollis); Michael D. Carter, flycatcher,Mionectes oleagina; Steven M. Goodman, Socialorganization and parasitichabits of breeding Avifaunal survey of the central Egyptian eastern Bronzed Cowbirds (Molothrus aeneus); Christine desert; Kathleen Diane Groschupf, Song repertoires Copenhaver,Experimental analysis of decisionmak- and singing behavior of Rufous-wingedand Cassin's ing in hummingbirds:the effectof resourcedistribu- sparrows:functional significance of diverse singing tion on territory defense;Robert L. Curry, Evolution strategiesin Aimophila sparrows; Lise A. Hanners, and ecologyof communalbreeding in Galapagos Development of social behavior in Laughing Gull Mockingbirds;William JamesDavis, Significanceof (Larusatricilla) chicks;Linda Heald, Behavioralplas- vocalizationsin the Belted Kingfisher (Megaceryle ticity in a Tyrannid flycatcher:effects of environmen-

(continuedon p. 138) 138 RAIKOWAND CR•CR•VT [Auk, Vol. 100 methods, and his argument is little more than OLSON,S. L. 1983. Evidence for a polyphyletic or- an opinion. On the basis of present under- igin of the Piciformes.Auk 100:126-133. standing,we concludethat piciform monophy- ß & D. W. STEADMAN. 1981. The relation- ly remains the preferredhypothesis. ships of the Pedionomidae(Aves: Charadri- iformes). Smithsonian Contri. Zool. 337. ACKNOWLEDGMENTS RAIKOW,R. J. 1982. Monophyly of the Passeri- formes:test of a phylogenetichypothesis. Auk We thank Kenneth C. Parkes and D. Scott Wood, 99: 431-445. CarnegieMuseum of Natural History, for providing SIMPSON,S. F., & J. CI•ACI•AFT. 1981. The phylo- specimensused in dissection.Parkes, Wood, and genetic relationships of the Piciformes (Class Mary C. McKitrick providedhelpful criticismsof the Aves). Auk 98: 481-494. manuscript. Our research is supported by N.S.F. STEINBACFIERßG. 1935. Functionell-anatomische grants DEB-8010898(R.J.R.) and DEB-7921492(J.C.). Untersuchungenan Vogelftissenmit Wendzeh-

LITERATURE CITED en und Rticksehen.J. Omithol. 83: 214-282. SWIERCZEWSKI,E. V. 1977. The hindlimb myology BALLMAN, P. 1969a. Les oiseaux mioc•nes de La and phylogeneticrelationships of the avian or- Grive-Saint-Alban (Isbre). Geobios 2: 157-204. der Piciformes.Unpublished Ph.D. dissertation, 1969b. Die V6gel aus der altburdigalen Univ. Pittsburgh,Pittsburghß Pennsylvania. Spaltenffillungvon Wintershof (West) bei Eich- ß & R. J. RAIKOW. 1981. Hind limb mor- st•ittin Bayera. Zitteliana 1: 5-60. phology, phylogeny, and classificationof the BRODKORB,P. 1971. Catalogueof fossil birds: part Piciformes. Auk 98: 466-480. 4 (Columbiformes through Piciformes). Bull. WILEY, E. O. 1979. Ventral gill arch musclesand Florida State Mus. 15: 163-266. the interrelationshipsof Gnathostomes,with a ELDREDGEßN., & J. CRACRAFT.1980. Phylogenetic new classificationof the vertebrata. Zool. J. Lin- patternsand the evolutionaryprocess. New York, naean Soc. 67: 149-179. Columbia Univ. Press. 1981. Phylogenetics:the theoryand prac- FEDUCCIA,A., & L. D. MARTIN. 1976. The Eocene tice of phylogeneticsystematics. New YorkßJohn zygodactylbirds of North America (Aves: Pici- Wiley & Sons. formes). Smithsonian Contri. Zool. 27: 101-110.

(continuedfrom p. 125) tal variability; Gary Richard Hepp, Cost-benefit Ani (Crotophagaani); Michael P. Lombardo,Auxiliary analysisof pair-bond formationin wintering Green- birds at Tree Swallow (Iridoprocnebicolor) nests; winged Teal (Anas creccacarolinensis); Marc Herre- David B. McDonaldß Role of resource distribution mans, Intra- and inter-island variation in the Bul- and behavioralinteractions in a lek systemwith male- buls ( Microscelismadagascariensis/ M. crassirostris) on male cooperation;G. M. O. Maloiy, Environmental Grand Comoroand Moheli; Wendy L. Hillß Behav- physiology of two East African land birds: Kori ioralflexibility of reproductivebehavior in theAmer- Bustardand SecretaryBird; JeffreyS. Marks, Natal ican Coot; Mark Alan Holmgren, Geographicand philoparryand breeding-areafidelity in Long-eared songvariation in westernField Sparrows;William G. Owls; John L. Maron, Intraspecific competition Hoppes,Avian frugivoryand seeddispersal in north amongwintering Sanderlings;John Marzluff, Corre- temperateforests; Anne E. Houde, Nest site selec- latesof reproductivesuccess in the Pition Jay, Gym- tion by Common and Roseateterns; Laurie Ann norhinuscyanocephalus; Donald B. Miles, Patternsof Hunter, Cooperativebreeding behavior of Purple morphologyand ecologyin a SonoranDesert avian Gallinules; Kristine Johnsonß Mate choice in the community;Patrick J. MockßComparative breeding Pition Jay, Gymnorhinuscyanocephalus; Nancy Joste, ecologyand energeticsof storm-petrelspecies in Baja Electrophoreticanalysis of paternity in the Acorn CaliforniaßMexico; Michael L. Morrison, Geographic Woodpecker;Robert S. KennedyßSystematic review variation in morphologyand singingbehavior of the of Philippine birds for "A Checklist of Philippine Black-throatedGray Warbler in Oregon; CharlesA. Birds"; RichardLee and SusanKay KnightßValue of Munnß Socialorganization and ecologyof canopy flockingby Bald Eaglesin relation to foragingand flocks in Amazonian Peru; Gerald J. Niemi, Ecomor- predation: an experimentalstudy; RachelN. Levinß phologyof breeding birds in peatlandhabitats of Adaptive significanceof antiphonalsong in Thryo- North Americaand northernEurope; Manuel Nores, thorus nigricapillus;Stewart T. Levinson, Bird/fruit Studyof specimensof a new speciesof antbirdfrom interactions in northern Florida; Robert K. Loftin, the east of Brazil; StephenNowicki, Physiologyand Communalnesting behavior of the Smooth-billed physicsof harmonically-complexbird calls;Allan R. (continuedon p. 148) 148 SCOTTAND GRUMSTRUP-$COTT [Auk, Vol. 100

$ELANDER,R. K. 1964. Behavior of captive South preening invitation display of parasitic cow- American cowbirds. Auk 81: 394-402. birds. Auk 78: 473-504. --, & D. R. GILLER. 1960. First-year plumages STEVENSON,M. 1969. Agonistic behavior in the of the Brown-headedCowbird and Red-winged cowbirdMolothrus ater. UnpublishedPh.D. dis- Blackbird. Condor 62: 202-214. sertation. Manhattan, Kansas, Kansas State Univ. --, & C. J. LA RUE, JR. 1961. Interspecific ZAR, J. H. 1974. Biostatisticalanalysis. Englewood Cliffs, New Jersey,Prentice Hall Inc.

(continued from p. 138) Phillips, Examinationof specimensat the American Potential effects of food abundance and social envi- Museum of Natural History; Nina Pierpont, Foraging ronment on migratory distancein Dark-eyed Juncos behavior, competition,and relative abundanceof (Juncohyemalis hyemalis); Kimberly Titus, Nest-site woodcreepers(Dendrocolaptidae) at an Amazonian habitat selectionpatterns of woodlandhawks in three forest site; Melinda Pruett-Jones,Vocal copying in temperate forest regions; Jill M. Trainer, Song dia- Macgregor's Bowerbird (Amblyornismacgregoriae); lects, communication, and social organization in StephenPruett-Jones, Ecology and evolutionof social Chestnut-headedOropendolas (Zarhynchuswagleri) organizationin Parotialawesii, Lawes' Six-wired Bird and Yellow-rumped Caciques(Cacicus cela); Peter B. of Paradise;Mark D. Reynolds,Mobbing behaviorof Turchin, Determination of paternity in a polyandrous nestingBlack-billed Magpies; Mark F. Riegner, For- bird, Jacanaspinosa; Katherine A. Voss-Roberts, Pa- aging ecologyof the Yellow-crownedNight Heron; rental attentiveness and embryonic tolerance of Emily FrancescaRissman, Hormones and socialbe- thermal stressin Yellow-headed Blackbirds (Xantho- havior in the Spotted Sandpiper, Actitis macularia; cephalusxanthocephalus); Thomas C. Will, Behavioral Gary Ritchison, Singing behavior of the Cardinal, ecology of species replacement: Blue-winged and Cardinaliscardinalis, with emphasis on the function Golden-wingedwarblers in Michigan; PamelaL. Wil- of singing by females;Nicholas L. Rodenhouse,In- liams, Comparisonof colonialand non-colonialnest- dividual and regional variability in the foraging be- ing by northern orioles in central coastalCalifornia; havior of forestpasserines; Stephen R. Sabo,Mixed- Kathy A. Winnett-Murray, Behavioralecology of four speciesflocks of Hawaiian honeycreepers;Bette J. Central American wrens; Barbara L. Woods, Mate- Schardien,Behavioral ecology of a southernpopula- guarding and female reproductivesynchrony in Barn tion of Killdeer; Debra H. Schlenoff, The "startle" and Cliff swallows; David A. Yokel, Mating system responseof wild-caughtbirds to artificial prey; Hel- of the Brown-headed Cowbird in two different en- mut Sick, Study of specimensin the AmericanMu- vironments; Robert M. Zink, Allozymic correlates seum of Natural History and identification of birds of dominancebehavior in Dark-eyedJuncos: a reap- from Brazil; Thomas Allen Stevens, Food hoarding praisal. by Great Grey and LesserGrey Shrikes, Laniusexcu- Correction to 1981 list: MarthLeah Chaiken re- bior and L. minor, a quantitative and experimental ceived a grant for research on the functions and study; Stuart D. Strahl, Ecologicalcorrelates of the development of variability in some calls of the socialsystem and foraging behavior of the hoatzin European Starling ( Sturnus vulgaris ). (Opiathocomushoazin) in Venezuela;Scott B. Terrill, January1983] Harlan'sHawk 169

--, & J. R. HAUGH. 1969. Recent data on sum- WILLIAMSON, F. S. L., L. J. PEYTON, & M. E. ISLIEB. mer birds of the upper Yukon River, Alaska, and 1965. New distributional and overwintering adjacentparts of the Yukon Territory, Canada. records of birds from south-central Alaska. Con- Can. Field Natur. 83: 257-271. dor 67: 73-79.

The American Ornithologists'Union will hold its Annual Meeting in New York from 26 Septemberto 1 October, 1983. A formal call for papers will be mailed to membersby mid-April with the deadline for submission of abstractsfor papers and poster sessionsof 15 June 1983. For further information on the scientificprogram, contactDr. George Barrowdough, Department of Ornithology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024.

Fellowsand ElectiveMembers are reminded that nominationsfor ElectiveMembers or Fellows may be submitted to the Secretaryon the prescribedform up until five months prior to the opening of the next dated meeting. The deadline for 1983 is 26 April. Nominations for Vice President and three Elective Councillorsmay be made in writing to the Secretaryat any time prior to the annual meeting.

The American Ornithologists'Union invites applicationsby I March 1983 for the following competitive research awards: JosselynVan Tyne Memorial Fund supportsprojects in any area of ornithology,with awards of a few hundred to a thousand dollars each. Successfulapplications usually focus on one or a few carefully de- fined, incisive questions,rather than on generalbiology or management. AlexanderWetmore Memorial Fund supportsprojects involving systematicsand/or the taxonomyof birds and, secondarily,studies of neotropicalbirds. Awards are of a few hundred to a thousanddollars. Suc- cessfulapplications are usuallybuilt arounda well-focused,interesting question. Herbert and Betty Cames Awards are intended to promoteexcellence in the scientificstudy of birds by permitting individuals an opportunity to expand the design of field or laboratory work and to explore new methods of data analysis. Applicationsare judged on the basis of scientificmerit, importance,and originality. One or two awards of $1,000-$2,000are made. Students,amateurs, and otherswith limited or no accessto major grantingagencies are encouragedto apply for the Van Tyne or Wetmore Awards, while the Carnes Award is open to all individuals. Appli- cation forms and further information on the Van Tyne and Wetmore Awards may be obtained from Dr. Erica Dunn, 30 Davidson Road, Aurora, Ontario L4G 2B1 Canada; applicationsfor the Cames Award shouldbe addressedto Dr. Ellen D. Ketterson,Department of Biology,University of Indiana, Bloomington, Indiana 47405 USA.