SHORT COMMUNICATIONS 765

and T. J. Cade [eds.], The 1975 North American easternMexico, p. 377-388. In W. Swanson[ed.], PeregrineFalcon survey. Can. Field-Nat. 90:228- National Geographic Society Research Report, 213. 1977. Vol. 18. National Geographic Society, Hm, W. G., J. H. ENDERSON,D. V. LANNING, M. A. DC. Hrrcucoc~, AND B. S. JOHNSON.1988. Nesting OGDEN, V. T., ANDM. G. HORNOCKER.1977. Nesting Peregrinesin Texas and northern Mexico, p. 115- density and successof Prairie Falcons in south- 121. In T. J. Cade, J. H. Enderson, C. G. Thelan- western Idaho. J. Wildl. Manage. 4 1:l-l 1. der, and C. M. White [eds.], Pop- OLENDORFF, R. R., AND J. W. STODDART,JR. 1974. ulations: their management and recovery. Univ. The potential of management of raptor popula- Wisconsin Press, Madison. tions in western grasslands,p. 47-88. In F. N. INSTITUTODE GEOG~AF~A,UNIV. NAC. AuTCIN.Mnxrco. Hamerstrom. Jr.. B. E. Harrell. and R. R. Olen- 1970. Cartas de Climas: Chihuahua. Culiacan. dotIT [eds.], Management of raptors. Raptor Res. Hidalgo de1Parral, Monterrey, Nogales, Ojinaga; Rept. no. 2. and San Luis Potosi. Comision de Estudios de1 RUNDE,D. E.. AND S. H. ANDERSON.1986. Charac- Tenitorio National. 7 maps. teristics of cliffs and nest sites used by breeding LANNING,D. V., P. W. LAWSON,AND W. G. HUNT. Prairie Falcons. Raptor Res. 20:21-28. 1985. Ecology of the Peregrine Falcon in north-

The Condor93:765-768 0 The CooperOrnithological Society 1991

DEMOGRAPHY OF THE PIGEON ON SOUTHEAST FARALLON ISLAND, ’

DOUGLAS A. NELSON Department of Zoology, Universityof California, Davis, CA 95616

Key words: Demography; ; Cep- from their wintering areas,and continuedobservations phus columba; ;survivorship. until late July or August when chicks began to fledge. I also visited the island for one month in April 1981 (Alcidae) form an important component of ma- and May 1982. rine avifaunas in the Northern Hemisphere, in terms In 1977, I studied about 10 pairs of that of numbers of both individuals and species.In order nestedatop a 40 m high ridge on Shubrick Point. After to effectively manage and protect alcid populations, 1977, I also made observations on the southeastside basic demographic data are vital (Nisbet 1979). De- of the island where approximately 20 pairs nested in mographic parametershave been estimated for five of or near a rubble pile at the head of a surge channel the six speciesin the North Atlantic (Hudson 1985). (East Landing). In contrast, survivorship of banded is known for Guillemots nested in natural and artificial crevices only one Pacific Ocean alcid species(Ancient Murrelet, at East Landing (Nelson 1987). Eight-nine guillemots Synthliboramphusantiques, Gaston 1990). In this note, were capturedwith leg noosesand banded with unique I report survivorshipand recruitmentof a color-marked combinations of two color bands and a numbered mo- sample of Pigeon Guillemots (Cepphuscolumba) over nel U.S. Fish and Wildlife Serviceband (Nelson 1987). a five-year period. To minimize nest desertion, I allowed birds to occupy a site for at least one week before capture. Additional METHODS birds were identified by unique tears or holes in their I observedPigeon Guillemots in 1977 and from 1979- webbed feet or by plumage marks in the white wing 1982 on SoutheastFarallon Island, California (37”42’N, patches.Birds were sexedby copulation position (Nel- 123QO’W) a 44 ha island about 43 km west of San son 1987). The Point Reyes Observatory (PRBO) Francisco.Biological and physicalfeatures of the island also banded between 19 and 284 chicks annually since were describedbv Ainlev and Lewis ( 1974) and Ainlev 1970 with unique year-classcolor rings (Ainley et al. and Boekelheide-(1990): Among the ‘12 speciesof ma- 1990). Most of these birds were banded on Lighthouse rine birds breeding on SE Farallon is a population of Hill, about 200 m from my study sites.Several of these approximately 1,000 Pigeon Guillemot pairs (Ainley birds were in my study plot each year as territory own- et al. 1990). I beganobservations in March 1977, 1979 ers and non-territorial “loafers.” and 1980 before guillemots had returned to the island Survivorship of territory owners was estimated by resightingbanded birds and five birds with easily vis- iblei unique natural marks at Shubrick Point and East * Received 13 November 1990. Final acceptance5 Landing. This includesseveral pairs eachyear that held April 1991. territories but did not produceeggs (Nelson 1987). Sur- 766 SHORT COMMUNICATIONS

TABLE 1. Annual survivorship estimates for teni- TABLE 2. Association between site fidelity and sur- torial Pigeon Guillemots. vival of mate.

YelI Survivorship nb Site fidelitv

Mate SUY MOW 1977-1979” 0.89 10 1979-1980 0.84 37 Present 71 4 1980-1981 0.62 52 Not present 12 8 1981-1982 0.76 33

aAnnual survivorship over the two-year period was estimated as the square root of the percentage of birds that survived. b n birds alive at beginning of sample period. vorce” (both members of the pair were alive the next year, but one or both occupied different crevices). Most changesof nest site were associatedwith loss viva1 was defined as the proportion of birds alive in of a mate. However, loss of mate did not usually lead year x that were present in year 52+ 1. I observed at to a change of nest site because 12/20 birds (60%) East Landing almost daily, and at Shubrick Point one whosemates disappearedremained at the same crevice or two mornings a week. To estimate dispersal, I at- (Table 2). Both members of the pair had to be recog- tempted to locate all banded guillemots on the island. nizable to be included in this analysis. Males (n = 4) However, much of the island was inaccessible.Band and females (n = 4) were equally likely to changecrev- losswas not a seriousproblem over the relatively short ices after mate loss. term of this study. Three individuals lost single color Of 10 birds that lost their mates during the breeding bands but I was able to keep track of the birds. The season(April-August), nine remained at their crevices rate of plastic band loss was about 1% per band-year. while four of six whose mates did not return in April I never found a bird with plastic band(s) and no monel changedcrevices (Fisher’s Exact Probability = 0.036). band. The one female who did change crevices after losing her mate in June 1980 paired with a widowed male 3 RESULTS m away and also repeatedly defended her old crevice Survivorship.The mean annual survival of territory against other females in the same year. holders was 0.80 (n = 5 years, Table 1). This is a Predatorsand competitors.Pigeon Guillemots had minimum estimateand may includedisappearance due few predatorson SE Farallon. PeregrineFalcons (Falco to emigration. The value for 1980-198 1 is low because peregrinus)were rare on the island during the I missed two birds that were present in 1982. I also breeding seasonbut have been observed to chase and captured and banded 15 loafers. Eight were not seen captureguillemots (PRBO unpubl. data). Western again after their capture. One loafer was sighted about (Larus occidentalis)caught but never killed guillemots 450 m from East Landing 22 months after I banded it in the 14 capturesI witnessed. Most capturesby gulls (PRBO, unpubl. data). The remaining six along with seemed to be a matter of chance, although a few in- many other recognizable birds visited the study sites dividual gulls stalked and attempted to capture guil- almost every day. lemots. On one occasiona stole an octopus from Prebreedingdispersal.Two ofthree known-agemales a guillemot feeding chicks. (one seven-year-old, one two-year-old) held territories Six hole-nesting seabird speciesbreed on SE Faral- less than five meters from the crevice where they had lon, two storm-petrels (Ashy Storm-petrel, Uceanod- been banded as chicks.A seven-year-oldmale and two roma homochroaand Leach’s Storm-petrel, 0. leuco- eight-year-old females bred in crevices about 200 m rhoa: Ainley and Lewis, 1974), and four alcids. I never from where they had been raised. saw guillemots interact with either of the nocturnal Age atfrrst breeding.The youngestguillemot known storm-petrels, but conflicts with the three other alcids to breed on SE Farallon was a three-year-old male in were common. Cassin’s Auklet (Ptychoramphusaleu- 1979. This male acquired a territory and mate as a ticus) is a small species(mean mass = 165 gm; Ma- one-year-old in late June 1977, but did not breed until nuwal 1979) that visits the island only at night, except 1979; however, very few guillemots bred successfully when incubating eggsor brooding chicks. Egg-laying in 1978 on SE Farallon (Ainley et al. 1990). In 1979, usually began during March or April, two to three two seven-year-oldsand two eight-year-oldsalso bred. months before guillemotslaid. At EastLanding in 1979 These birds all bred in 1980, and there were no new and 1980, auklets laid eggs in 13 of 28 nests (461) known-agerecruits in 1980. Loafers were youngerthan used the same seasonby guillemots. None of the 13 most breeders.In 1979 there were three two-year-olds, auklet pairs successfullyreared a chick. I observedmale one three-year-old, one five-year-old, and one six-year guillemotschase auklets out oftheir nestsseveral times. old loafers. In 1980 there were three three-year-olds, Of the other alcids, the Tufted (Lunda cirrha- and one four-year-old. ta: 797 gm, Sealy 1973) is considerablylarger, and the Nest-site retention and pair-bond maintenance. Of RhinocerosAuklet (5 18 gm, Sealy 1973) slightly larger 95 year-year transitionsinvolving 43 different territory than the Pigeon Guillemot (490 gm, Nelson 1987). owners, birds returned subsequentlyto the same nest- Fewer than 150 pairs of either species occur on SE site in 83 cases(87%) (Table 2). Three males bred for Farallon, although the populations of both speciesap- at least six consecutiveyears in the same nests. pear to be increasing (Ainley and Boekelheide 1990). In 12 cases(13%), birds switched crevices between One pair took over a guillemot crevice years. Four of these switcheswere associatedwith “di- on Shubrick Point in 1977 and in late June and July SHORT COMMUNICATIONS 767

of that year (presumably young non-breeders) and adult survival is 80%, then 40% of fledged chicks inspectedmost of the guillemot crevicesI was studying. would need to survive to breeding age. Average ex- Guillemots with nests containing chicks gave alarm pectation of life after entering the breeding population calls, mobbed, and chased Tufted Puffins that ap- is about 4.5 years (Cormack 1964). Both of these es- proached their nests (Nelson 1985). Aerial chasesof timates are very sensitive to the estimated adult sur- puffins often included body contact. No Tufted Puffins vivorship (Hudson 1985). were ever observed at East Landing, but two or three The presenceof nest-site competitors has implica- pairs of Rhinoceros Auklets nested there and several tions for the SE Farallon guillemot population. If the pairs also nested near my Shubrick Point study plot. populationsof Tufted Puffins and Rhinoceros Auklets One guillemot crevice on Shubrick Point was taken continue to grow, then the guillemot population might over by Rhinoceros Auklets in 1979. Circumstantial decline (Ainley and Boekelheide 1990). The degree of evidence suggestedthat Rhinoceros Auklets were as- decreasewould be limited by the number and distri- sociatedwith crevice changesby guillemotsin two oth- bution of creviceslarge enough for guillemots, but too er casesin 1979. small for the larger species.In Bering Seacolonies, size- related interference competition for nest crevices also DISCUSSION occursamong hole-nestingauks (B&lard 1969, Divoky 1982, Knudtson and Byrd 1982, Piatt et al. 1990). Pigeon Guillemots on SE Farallon exhibited high an- nual adult survivorship, retention of the nest-site and It is a pleasure fo thank the staff and volunteers of mate, and deferral of first breeding until about three the PRBO who have helped me over the years. The or four years of age. These estimated demographicpa- San Francisco Bay National Wildlife Refuge granted rameters resemble those of other auks, especially the permissionto work on the Farallones.Grants from the congeneric (Cepphur grylle) (Asbirk Societyof Sigma Xi, the Frank M. Chapman Memorial 1979, Hudson 1985). The annual survivorship of 80% Fund of the American Museum of Natural History, the is similar to that of the Black Guillemot, but lower Division of Biological Sciences,Horace H. Rackham than estimates for most other auks,which average90- School of Graduate Studies and the Museum of Zo- 95%. An exception is the Ancient Murrelet, with an- ology of the University of Michigan, and a NSF grad- nual survival of 77% (Gaston 1990). Speich and Ma- uate fellowship supported the work. Logistic support nuwal (1974) estimated annual survival of Cassin’s was provided by PRBO and the Farallon Patrol of the Auklet on SE Farallon to be 83%. This estimate was Oceanic Societv (San Francisco Chanter). John Piatt based on the frequency distribution of gular pouch and an anonymous reviewer made useful comments lengthswithin the population, and shouldbe confirmed on the manuscript. with banding data. My estimate may be somewhatlow, LITERATURE CITED as sampling effort declined in later years (1 month in 1981 and 1982). This may be partly responsiblefor the AINLEY, D. G., AND T. J. LEWIS. 1974. The history apparent decline in survival rates. The year with the of Farallon Island marine bird populations, 1854- lowest survival (1980-l 98 1) was also the year I visited 1972. Condor 76~432-446. the island only in April, when colony attendance is AINLEY, D. G., AND R. J. BOEKELHEIDE [EDS.]. 1990. highly variable (Nelson 1987). Dispersalhorn my study Seabirdsof the : Ecoloay.___ dvnam- _ sites is an unlikely cause of low estimates, as I never its, and structure of an upwelling system com- sighted a banded breeder away from my sites. munity. Stanford Univ. Press, Stanford, CA. In common with most other seabirds,Pigeon Guil- AINLEY, D. G., R. J. BOEKELHEIDE,S. H. MORRELL, lemots retained the same mate and crevice from year AND C. S. STRONG. 1990. Pigeon Guillemot, p. to year in this and previous studies. On Mandarte Is- 276-305. In D. G. Ainley and R. J. Boekelheide land, Drent (1965) recorded six instancesof nest-site [eds.], Seabirds of the Farallon Islands. Stanford tenacity of at least’four years. On SE Farallon, Tenaza Univ. Press, Stanford, CA. ( 1966) recorded two pairs at the same crevices in two ASBIRK,S. 1979. The adaptive significance of the successiveyears. reproductive pattern in the Black Guillemot, Cep- The timing of mate loss and the availability of re- phus grylle. Vidensk. Medd. Dan. Naturhist. For- placement mates may have influenced whether or not en. 141:29-80. a bird switched crevices. Guillemots were more likely BBDARD,J. The nesting of the Crested, Least, and to move between nests if their mate did not return at Parakeet Auklets on St. Lawrence Island, . the beginning of the breeding seasonthan if the mate Condor 7 1:386-398. died during the breeding season.Since clutch size and CO~MACK,R. M. 1964. Estimates of survival from fledgingsuccess decline with time (Nelson 1987, Ainley the sighting of marked . Biometrika 5 1: et al. 1990), it may be advantageousfor a bird widowed 429438. early in the season to move and form a pair bond DIVOKY,G. J. 1982. The occurrenceand behavior of immediately with an available bird. Later in the season, non-breeding Homed Puffins at Black Guillemot there are few unoccupiedcrevices and more birds avail- coloniesin northern Alaska. Wilson Bull. 94:356- able to serve as replacement mates and competitors 358. for crevices. The best tactic then may be to hold onto DRENT,R. H. 1965. Breeding biology of the Pigeon one’s crevice and wait until next year. Guillemot Cepphuscolumba. Ardea 53:99-160. Average survival to age of first breeding can be es- GASTON,A. R. 1990. Population parameters of the timated if we make several assumptions.If pairs fledge Ancient Murrelet. Condor 92:998-1011. an average of 1.O chicks per year (Ainley et al. 1990) HUDSON,P. J. 1985. Population parameters for the 768 SHORT COMMUNICATIONS

Atlantic Alcidae, p. 233-26 1. In D. N. Nettleship 3 15. In J. C. Bartonek and D. N. Nettleship [eds.], and T. R. Birkhead [eds.], The Atlantic Alcidae. Conservation of marine birds of northern North Academic Press, New York. America. U.S. Fish and Wildlife Service. Wildlife KNUDTSON, E. P., AND G. V. BYRD. 1982. Breeding Research Report 11, Washington, D.C. ’ biology of Crested, Least, and Whiskered AukleG PIATT,J. F., B. D. ROBERTS,AND S. A. HATCH. 1990. on . Alaska. Condor 84: 197-202. Colony attendance and population monitoring of MANUWAL,D. A. 1979. Reproductive commitment Least and CrestedAuklets on St. Lawrence Island, and successof Cassin’s Auklet. Condor 8 1:11 l- Alaska. Condor 92:97-106. 121. SEALY,S. G. 1973. Breeding biology of the Homed NELSON, D. A. 1985. The syntacticand semantic or- Puffin on St. Lawrence Island, , with ganization of Pigeon Guillemot (Cepphuscolum- zoogeographicalnotes on the north Pa&c puffins. ba) vocal behavior. Z. Tierpsychol. 67:97-130. Pac. Sci. 27:99-l 19. NELSON,D. A. 1987. Factors influencing colony at- SPEICH,S., ANDD. A. MANUWAL.1974. Gular pouch tendanceby Pigeon Guillemots on SoGtheastFar- development and population structureof Cassin’s allon Island. California. Condor 89:340-348. Auklet. 9 1:29I-306. NISBET, I.C.T. i979. Conservation of marine birds TENAZA,R. R. 1966. Pigeon guillemot banding re- of northern -a summary, p. 305- turns. Bird Banding 37:288.

The Condor93176%773 0 The CooperOrnithological Society 1991

SHORT-TERM VARIATIONS IN WATER-VAPOR PRESSURE IN NESTS OF COMMON CANARIES ’

MICHAELD. KERN AND MICHAELKNAPIC Biology Department, The Collegeof Wooster, Wooster,OH 44691

Key words: Common Canary; nesthumidity; water- MATERIALS AND METHODS vaporpressure,. nesting behavior; Serinus canarius. We measured PN in 11 canary nests throughout days 1, 4, 7, and/or 10 of incubation. In all, we collected The discovery that eggs,regardless of their size, lose data for 28 separatedays of incubation (Table 1). Dur- similar proportions of water during incubation (Ar and ing 23 of these days, we measured PN every 15 min; Rahn 1980) generatedconsiderable interest among avi- during the other 5 days, at hourly intervals. Our canary an biologists in the water-vapor pressure of the nest hens were experienced birds which had raised broods (PN) since it determines how much water is lost from successfullyin previous seasons.However, the data the egg.As a result,P, has been measuredin numerous (see below) suggestthat some of the males with which avian species (Walsberg 1980, Rahn and Paganelli they were paired were inexperienced breeders. 1990). Pairs were kept in standard double-brooder cages However, most of these measurementswere made (23 x 36 x 28 cm) in an indoor aviary. They were with egghygrometers which provide only one value of exposedto a long d&ly photoperiod (16L:8D, lights on PN for periods of incubation typically spanningone or from 06:OOto 22:00 EDT). air temoeratures of 20.3- more days. Until recently, little information was avail- 25.0% (22.8% on average), and ambient vapor pres- able concerning variations in the moisture content of sures (P,) of 7.8-16.7 torr. Within any given day, P, nests over short intervals of time. We now have de- varied 1.4-6.5 torr (4.1 ton on average; Fig. 1). The scriptionsof suchchanges, but only from nestsof a few birds had accessto food and fresh water ad libitum, large birds, specifically those of (1) swans and geese, and received greensand vitamins once or twice a week. where PN fluctuates as much as 9 torr during 24-hr Each cagecontained a lined, plasticnest pan in which periods (Howey et al. 1984); (2) ostriches,where it only the hens laid and incubated clutchesof 4-5 eggs.The varies about 4 torr each day (Swart et al. 1987); and bottom of the nest pan was fitted with a 2-cm diameter (3) ptarmigans, where it scarcely changesat all (An- scinteredbronze (dust) cap. This enabled us to plug a dersen and Steen 1986). No comparable information relative humidity probe (Solomat Model HCl) into a appearsto exist for other birds, particularly small ones. nest for 24-hr periods without disturbing the incubat- We attempt here to remedy this situation partially by ing canary.Parts ofthe cap other than the surfacefacing describing short-term variations in the moisture con- the eggs were taped, and the cap itself taped to the tent of Common Canary (Serinuscanarius) nests. probe, so that the only air reaching the sensor came from the nest. When PN was not being measured, we kept a plug covered with Saran Wrap0 in the cap to ’ ’ Received 14 November 1990. Final acceptance2 1 prevent it from draining heat and moisture from the February 1991. nest.