Nesting Biology of the Rosy Finch on the Aleutian Islands, Alaska

Condor 851447452 Q The Cooper Omithologtcal Society 1983

NESTING BIOLOGY OF THE ROSY FINCH ON THE ALEUTIAN ISLANDS, ALASKA

RICHARD E. JOHNSON

ABSTRACT. -The nesting biology of the Aleutian Rosy Finch (Leucosticte griseonucha(arctoa) griseonucha)was studied on the maritime tundra of Amchitka Island in the Aleutian Islands,Alaska. Fifty active nestswere found, all in buildings or other man-made structures.Nests were re-used with new material added each year. Their densitieswere not limited by the availability of buildings or good nest sites.The maximum density was 13.9 nests/100 ha. Unlike their alpine relatives, these birds were double-brooded, with eggsin the nest from late April to early June and again in late June through at least mid-July. Mean hatch date for the first brood was 26 May. Mean clutch size was 4.5 eggs(range = 3-6). Incubation lasted 12-14 days, the nestling period 15-22 days (K = 18 days). Egg survival during incubation was 7 3.9%, hatching success9 3.9%, and nestlingsurvival 79.1%, for an overall nest successof 54.9%. These high values are attributed to the near absenceof predators on the island and to protected nest sites.

Rosy finches (Leucosticte spp.) breed in two disjunct habitats: above timberline in the al- STUDY AREA AND METHODS pine zone of the major mountain ranges of All data were collected from 14 May to 16 July Asia and western North America, and at sea 1968 on Amchitka Island (5 l”30’N) in the Rat level on maritime tundra on the islands of the Island group of the Aleutian Islands, about Bering Sea. The breeding biology of popula- 2,200 km WSW of Anchorage, Alaska. Am- tions in alpine habitat has been studied by sev- chitka is about 65 km long and 2-7 km wide, eral workers (Twining 1938a, b, 1940; French and is oriented SE-to-NW. The southeastern 1959; Johnson 1965) but less information is two-thirds of the island is mostly below 100 available concerning maritime populations m elevation and is covered by wet maritime (Hanna 1922, Shreeve 1980). Rosy finches in tundra containing many ephemeral ponds and both habitats sharemany common life history underlain by peat. By contrast, the northwest- traits including cliff nesting, unbalanced sex em third is relatively barren and mountainous ratios, and floating territories centered on the (maximum elevation 354 m) with alpine female. On the other hand, they differ mor- meadows, fell-fields, scree, and talus. The phologically, the maritime birds having nearly physiography and vegetation were described twice the body weight of any alpine member by Shacklette et al. (1969). of the genus and a bill that is distinctly more During the breeding season, Amchitka is slenderthan all but one of them (Johnson1972, cool, windy, and blanketed by fog. Between 1977). The slenderbill and lowland tundra and May and July the sky is cloudy or overcast beach habitat suggesta differing feeding ecol- over 90% of the time and rain falls about 35% ogy, while the longer season associatedwith of the time (Armstrong 1977). Temperatures the low elevation and maritime climate im- during May, June, and July average 3.9”, 5.6”, plies major adjustments in breeding schedule and 7.8”C, respectively,with a mean daily range and perhaps productivity. of about 3.5”C. Wind speedsduring the sum- In this paper, I describe the nesting biology mer average about 26 km/h. of the Aleutian Rosy Finch (L. g. griseonucha, Kenyon (196 1) and White et al. (1977) have nomenclature follows Johnson 1977 [The describedthe avifauna. Land mammals are ab- American Ornithologists’ Union Committee sent except for the Norway rat (R&us nor- on Classification and Nomenclature (1982) vegicus),which probably became established now placesall North American speciesof Leu- during World War II (Murie 1959). Rats ap- costicte in arctoa.-Ed.]), and especially its parently causedthe decline and near extinction breeding chronology and output, and compare of the Winter Wren (Troglodytes troglodytes) these data with those for alpine populations at and Song Sparrow (Melospiza melodia) on the the same latitude. Also, I will present data on island, but seem to have had little effect on the nest densitiesand nestingsuccess for the Aleu- rosy finch. tian birds, information previously unavailable The study was centered on the southeastern for any member of the genus. end of the island, where about 2,000 aban-

14471 448 RICHARD E. JOHNSON

TABLE 1. Timing and duration of major events in the breeding cycle of rosy finches and Lapland Longspurson Amchitka Island, Alaska, and rosy finches in the Rocky Mountains of Montana.

Amchitka Island Rocky Mountains Rosy finch Lapland Longspur Rosy hnchb Egg-laying 28 April-26 May 30 May-30 June 16 June-6 July Incubation 3 May-8 June 4 June-10 July 21 June-19 July Nestlings 16 May-20June 16 June-20 July 4 July-6 August Fledging 3-20 June 7-20 July 22 July-6 August Egg-layingto 54 days 68 days 53 days fledging

aFrom Wdbamson and Emison (1971) and White et al. (1977). b R. E. Johnson (unpubl. data). Based on seven nests from Glacier National Park and three from the Mlssion Range, Montana; all between 47”20’N and 49”N latitude.

doned buildings provided potential nest sites. RESULTS AND DISCUSSION From 14 May to 16 July 1968 I examined the beaches, sea cliffs, and tundra on this part of TIMING OF NESTING the island, as well as the buildings. In addition, When I arrived on 14 May, birds had already I made short trips to the central highlandsand begunto lay, and by 8 June all eggshad hatched. the small lowland section at the northwestern Based on 38 nests containing what appeared tip of the island. to be first broods, I calculated a mean hatch I examined more than 500 buildings for date of 26 May, with a range of hatching from nests, recording the type of building (quonset 16 May to 8 June. Assuming a 13-day incu- or frame), presenceof active nests,clutch size, bation period and four to five days for egg- stage of nesting, height of the nest above the laying, I estimated that laying for all nests oc- floor, nest location (shelf, rafter, other), and curred between 28 April and 26 May (Table presence of old nests. Two areas of similar 1). Females were incubating from 3 May to 8 habitat, but where the density of buildings dif- June and young fledged from these nests be- fered, were canvassedthoroughly for nests to tween 3 and 20 June. determine breeding density. Although I did not attempt to locate new Nesting successwas determined by both nestsafter 2 1 June, evidence indicated that this Mayfield’s exposure method and the standard specieshad a secondbrood in late June or early method (Mayfield 196 1). For comparison with July. First, I found one nest in which eggswere existing data on Lapland Longspurs(Calcarius being laid on 1 and 2 July in a building where lapponicus),nesting successwas also calculat- an earlier nest had successfullyfledged young ed asthe percent of nestsfound before hatching on 10 June; and another in which eggs were that fledged at least one young. I estimated being incubated on 5 July (incubation contin- clutch size before mortality by using the mor- ued through my last observation on 15 July). tality rates of eggsand of young and the percent Second, copulations again became frequent, hatching determined by Mayfield’s procedure, beginning about 28 June. Finally, adult fe- and then worked backward to calculate loss. malescollected throughout the summer showed Lost individuals were then added to the num- a resurgenceof gonadal activity. From 22 May ber actually found in nests, and a mean ad- to 25 June only one of eight females collected justed for loss was calculated. had an ovary > 10 mm long or ova > 1 mm I generally visited each active nest every oth- in diameter, and that bird was one of three er day. I define “incubation period” asthe time obtained on the last day of this period. How- from laying of the last egg to the hatching of ever, between 29 June and 13 July, four of six the last egg. Young were considered to have had ovaries > 14 mm and ova >6 mm, and fledged when the last individual left the nest, two had eggsin utero. The long delay after the regardlessof whether it left the nest building first wave of laying (about 30 days), the syn- immediately or remained inside for one or chrony within the first wave, and the apparent more days. Mean hatching date was deter- synchrony in the second,support the interpre- mined by direct observation, or by counting, tation that these were second nestings,not re- either backward from fledgingor forward from placements of failed nests. completion of laying using an incubation pe- Stejneger (1885) found that rosy finches on riod of 13 days and a nestling period of 18 the Commander Islands were double-brooded days, as establishedin this study. and Kenyon (1961) suggestedthat those on NESTING OF ALEUTIAN ROSY FINCHES 449

Amchitka occasionally have three broods, but TABLE 2. Distribution and height of nestsof rosy finch- neither provided specificrecords. Secondnests es in 279 abandonedbuildings on Amchitka Island, Alas- are probably possible because the climate of ka. the Bering Sea region is both predictable and Build- uniform. Furthermore, the entire Aleutian Nurfber mgs wth “,,,, Number good active chain, including Amchitka, is 5-10” south of Type of of build- active nest nestsb Anchorage, Alaska, and hence the summer is bulldmg ings nests sites’ (X i SE) not nearly as short as that on the Alaskan Quonset 220 13 99 1.94 * 0.13 mainland or arctic tundra. Frame 59 16 41 2.47 i 0.25 Lapland Longspurs, the most common ’ Buildings wth at least one potential nest site that was mdistmguishable from other locations used successfully. breeding birds on the island, occurred in the o There IS no significant difference (P > 0.05) between these nest heights; same habitat as the finches, but were single- but see text. brooded (Williamson and Emison 197 1). Rosy finches were able to breed twice by starting to breed a full month earlier than the longspurs were not paneled and their exposedrafters and and by completing the nestingcycle more rap- wall supportsoffered a wide array of nest sites. idly (Table 1). The latter was possible even In contrast,quonset huts had smooth, rounded though the nestling period of the rosy finch interiors that provided few nest sites; 55% of was longer than that for the longspur (18 vs. them lacked shelves or irregularities where a 11 days),because the population of finchesbred nest might be placed. Also, potential nest sites with greater synchrony (Table 1). were generally higher above the floor in frame Rosy finches on the mainland in the alpine buildings. Although the difference in mean zone have not been reported to be double- heights of nests in the two types of buildings brooded, but they do have the same tight syn- is not significant for this prefledging survey chrony as those on Amchitka (Table 1). Pre- (Table 2) it is significant if the sample size is sumably, this synchrony first evolved in alpine increased by including nests found later areas as an adaptation to the short summer (frame = 2.94 m above floor, quonset = 1.73 season,and has subsequentlyprovided a pre- m; P < 0.001, n = 46). adaptation to multiple-brooding in locations Nests were on rafters (15) and wooden such as Amchitka where the seasonis longer. shelves(26) behind broken panels (3) behind The snow-free period in the lowland on Am- the top rung of a ladder nailed to a wall (1) chitka (1 May-3 1 October; Armstrong 1977) and in a compartment in a tool cabinet (1). is about twice as long as it is in alpine areas Metal shelves were avoided. of the mainland at the same latitude (15 June- Nests appeared to be used for many years, 31 August; R. E. Johnson, unpubl. data). becominglarge as new material was added each year. Old nestswere usually multilayered, with NEST LOCATION old eggsand dead young sometimes covered- I found 50 active nests associatedwith man- over rather than removed. Other nestshad ex- made structures,but none in the natural hab- panded sideways,containing as many as three itats (e.g., tundra, cliffs) where I spent most of nest cups in a line within one mass of nest my time. One nest was in a rock wall, another material. Some covered entire shelves, the in an oil drum partially buried in the tundra, largestbeing about 35 x 100 cm in lateral di- two were on the exterior of buildings, and the mensions and 15 cm deep. remainder inside buildings. Only one pair of I frequently visited cliff areasabove beaches birds usually occupieda building, although two where the finchesoften fed, but found no nests pairs nested successfullyat opposite ends of there. Cliffs of this sort apparently harbored one extremely large sprawling structure. No most nests before construction of buildings other speciesof bird nested in buildings. during World War II (Dal1 1873) and were I surveyed 279 buildings on the eastern end still commonly used in the 1950s (Krog 1953, of the island before the earliest known date of Kenyon 196 1). Some birds probably continue fledging. Twenty-nine had active nests and to nest on cliffs on other parts of the island, another 42 contained old used nests. Thus, at where I found birds but no buildings. least 25% had served asnest sitesat some time. Cahn (1947) suggestedthat rosy finchesnest- Although 79% of these buildings were quonset ed in the tundra on Amaknak Island, but I huts, less than half of the active nestswere in found no evidence of this on Amchitka. them (Table 2). In other words, finchesshowed a significantpreference (x2 = 22.47, P < 0.00 1, NEST DENSITIES 1 df) for frame structures.Several factors prob- I thoroughly censusedtwo areasbefore the ear- ably account for this: most frame buildings liest known date of fledging in order to deter- 450 RICHARD E. JOHNSON

TABLE 3. Comparison of nest and nest-site densitiesof particularly suspectbecause I observed it only rosy finchesin two study areason Amchitka Island, Alas- once, midway through the nestlingperiod. Us- ka. ing Mayfield’s (196 1) procedure,I calculatethat six individuals were probably lost from this Study area A B group of nests.When these are included in the calculation of clutch size, the mean becomes Number of active nests 11 19 4.54 eggs. Size of area (ha1 79 382 Active nests;1 Ob ha 13.9 5.0 Earlier records of clutch size on the Aleutian Number of buildings/ 100 ha 162 24 Islands are all from well east of Amchitka (11 Buildings with good nest nestson Adak [Shreeve 19801, one on Amak- sitesWOOha 110 13 nak [Cahn 19471,one on Unalaska [Dal1 1873]), Percent of buildings with good nest sitesa 67.9 54.2 and they yield a mean of 4.08 eggs.This value Old nest+/ 100 ha 19.0 7.1 is significantly lower than the Amchitka mean (P < 0.05) and may be due to geographicvari- a Sites indistinguishable from other locations used successfully. b Buildings with nests from prewous years, but without active nests during ation along the Aleutian chain or between-year the period of this study. variation, but more likely is due to inclusion of incomplete clutches.For example, the series of nests from Adak all contained young and mine nest densities (Table 3). Area A, which had an unusually high proportion of clutches had the higher density (13.9 nests/l 00 ha), also of three chicks (3 of 1 l), especially for this had more buildings, suggestingthat the avail- latitude (Johnson 1965). These facts suggest ability of acceptable nest sites might be lim- that some mortality preceded observations. iting. This idea was reinforced by my obser- The Adak sample may also have contained vations that few birds were present in areas secondbroods sincethree late nests(3-2 1 July) that lacked buildings. Two lines of evidence, were included, which on average had fewer however, indicate that something other than (3.33 vs. 4.25) young than the eight nests ob- the number of nest sitescontrolled population served there in May and early June. size. First, ifnest siteswere limiting, one would Clutch size on Amchitka was slightly, but expect nest densities to be proportional to not significantly (P > 0.05) higher than that building densities,which was not the case.Area for a sample of 10 nests of L. t. tephrocotis A had more than 6.5 times the density ofbuild- (X = 4.20 eggs) from the alpine zone in the ings of area B, but only 2.8 times the density Rocky Mountains at a similar latitude (47-52” of nests. Becausebuildings vary in suitability, N). Thus, altitudinal differences seem not to one might argue that fewer buildings in area influence clutch size. A were suitable.However, area A had 8.5 times as many buildings with good sites as area B INCUBATION PERIOD (Table 3). The incubation period was 12- 14 days for one Second, many old nests were in buildings eggin a nest that was abandoned by the female and sites not then in use in either area (Table during hatching and 12-13 days in another 3). If nest siteswere limiting, nearly all suitable where only three of four eggshatched. These sites should have been occupied. In these two figuresare similar to those published for other areas combined, more than 58% of all build- taxa of rosy finch on the mainland well to the ings that had been used earlier were not in use south (L. tephrocotis dtiwsoni: 14-16 days during 1968. Moreover, 78% of those that I [Wheelock 1920, Twining 1938a]; L. &rata: judged to be suitablewere not usedin that year. 12-14 days [French 19591;L. australis: 13-14 Taken together, these facts indicate that the days [Bailey and Niedrach 19651).Hence, they availability of nest sites did not determine do not support my earlier suggestion(Johnson population levels, at least within areas A and 1965) that the rate of development diminishes B. with increasingclutch size to the north in this genus.It is also evident that maritime and al- CLUTCH SIZE pine populations do not differ in their devel- Clutch size was estimated from 46 active nests opment. believed to contain first broods, either eggs that were being incubated or young chicks. NESTLING PERIOD These included one nest with three, 26 with For 10 nests followed from completion of four, 18 with five, and one with six eggs or hatching to fledging,the nestlingperiod ranged young, yielding a mean clutch size of 4.4 1 eggs. from 15 to 22 days and averaged 18 days. Nest- This value is probably low, because loss was lings that were 15 days of age or older became possible at most nests before I first examined restlesswhen approached and sometimes flew them. The nest that contained three young is or jumped from the nest to the floor. They NESTING OF ALEUTIAN ROSY FINCHES 451

often remained on the floor or werejust outside Two nests with nestlings were abandoned. the building for one or two days thereafter. Also, one chick was lost from each of 11 oth- This suggeststhat the mean nestlingperiod for erwise successfulnests, and two from another, undisturbednests is at least two days longer. in 1,662 nestling-days of observations. Four The few data available for other members of thesenestlings were found dead beneaththeir of the genus appear in general agreement: 14 nests, one was on the nest shelf, two were in days for L. g. littoralis (Shaw 1936); 18-20 their nests, and the remainder simply disap- days for L. atrata (French 1959); 14-15 days peared. Those that were dead in nestsprobably for L. australis(Bailey and Niedrach 1965); starved, and many of the others may have also. and 14-21 days for L. t. dawsoni(Wheelock Becauserosy finchesre-use old nests,nest par- 1920, Dixon 1936, Twining 1938a). (All fledg- asitesmay also be a mortality factor (Ricklefs ings at the sixteenth day or earlier, except that 1969). for littoralis,occurred in response to the ap- Both neststhat were abandoned during the proach of an observer.) Thus, rosy finches at nestling period were located on the outer side all elevations have nestling periods of similar of buildings, and subject to wetting by wind- duration. blown rain. These nests represent 50% of all nestsfound outside buildings, and their failure NESTING SUCCESS provides evidence of strong selection for their I determined nesting successfrom 35 nests placement within buildings offering total pro- containing first broods that were visited two tection from rain and wind. or more times and did not appear disturbed I found no evidence of predation on nest- by human activity. I found few nests during lings or eggs.None of the nests showed signs laying or early incubation; consequently nest- of disturbance, and dead chicks were not vis- days of observation prior to hatching (86) were ibly injured. The only ground predator on the fewer than those during the nestling period island is the Norway rat, which is largely re- (410) and my calculations of successare ac- stricted to beach areas and riparian meadows cordingly more reliable for the latter period. (Williamson and Emison 197 1). The presence From these data I calculated the following of several old eggsand dead birds on the floors probabilities: eggsurvival to hatching, 73.9%; of the buildings suggeststhat rats did not fre- hatching success,97.9%; and nestling survival quent them during my study. Avian predators from hatching to fledging, 79.1%. The product (e.g., Parasitic Jaegers, Stercorariusparasiti- of these values yielded a nesting success(i.e., cus)were uncommon, and probably had little the probability that eggswill survive from lay- direct effect on nests because these were in- ing to fledging)of 57.2%. This value, calculated doors. Williamson and Emison (197 1) also by the Mayfield (196 1) procedure,is lower than found that predation was low for nestsof Lap- that determined by the standard method land Longspurson the island. (64.1%) and provides a better estimate when The nestingsuccess of these rosy fincheswas few nests are found before the start of incu- high compared to values for most avian species bation. and higher than that for the Lapland Longspur, The only mortality that I saw during incu- the only other passerineon Amchitka for which bation occurred in two nests that were aban- data are available. Successfor the longspur, doned. No individual eggsdisappeared from calculated as percent of nestsfledging at least any nestsand only one of 47 eggsthat I watched oneyoung, was 57% (Williamson and Emison closely failed to hatch. However, this sample 197 l), compared to 73% for the finch. The high may have underestimated hatching failure be- successof finchesis probably becausethey have cause eight nests containing chicks, as well as few predatorsand useprotected nest sites.Lack several nests from previous years, contained (1954), Ricklefs (1969), and others have shown singleunhatched eggs. When the eight eggsfrom that hole-nestersand speciesthat nest in build- active nests were incorporated into the cal- ings have higher successrates, and the values culations, which required inclusion of all 23 they cited are comparableto thosein this study. nests not observed before hatching, hatching I know of no published figures for nesting successdropped to 93.9%. This figure exag- successof rosy finches in alpine habitat, but I geratesfailure if any of the unhatchedeggs were can make a tentative estimate from the frag- eggsremaining from a previous nesting or if mentary information presently available. Four individual nestlingswere lost prior to my first of 24 nestsof L. t. dawsonifound by Twining nest observation, and exaggeratessuccess if (1938a, 1940) were lost to predators, and the unhatched eggswere removed by the parents. rest apparently fledgedyoung, for a successrate These factors probably cancel one another. If (nestsfledging at least one young) of 83%. All this lower (93.9%) estimate of hatching success five nestsof L. t. tephrocotisthat I monitored is used, nesting successbecomes 54.9%. (Johnson1965) were successful(= 100%). These 452 RICHARD E. JOHNSON figuresare probably biased upward becauseall KENYON,K. W. 1961. Birds of Amchitka Island, Alaska. of my nestsand 11 of those reported by Twin- Auk 78:305-326. KROG,J. 1953. Notes on the birds of Amchitka Island, ing were found after hatching; thus, the ob- Alaska. Condor 55:299-304. servedlosses were probably fewer than for nests LACK,D. 1954. The natural regulation of animal num- watched for the entire nest period (see May- bers. Clarendon Press, Oxford. field 1961). These data nevertheless suggest MAYFIELD,H. 1961. Nesting successcalculated from ex- that the nesting successof alpine populations vosure. Wilson Bull. 73:255-26 1. M&E, 0. J. 1959. Fauna of the Aleutian Islands and of rosy finches is also high. Alaska Peninsula. N. Am. Fauna 6 1:l-406. RICKLEFS,R. E. 1969. An analysis of nesting mortality ACKNOWLEDGMENTS in birds. Smithson. Contrib. Zool. No. 9. This study was incidental to investigations of the avian SHACKLETTE,H. T., L. W. DURRELL,J. A. ERDMAN,J. R. ecologyofAmchitka Island directedby F. S. L. Williamson KEITH,W. M. KLEIN, H. KROG,H. PERSSON,H. SKUJA, for Battelle Memorial Institute under contract from the AND W. A. WEBER. 1969. Vegetation of Amchitka Atomic EnergyCommission. I am grateful to D. W. John- Island, Aleutian Islands,Alaska. U.S. Geol. Surv. Prof. ston, R. E. LeResche, R. C. Wood, and C. G. Yarbrough Pap. No. 648. for assistancein the field, and especiallyto Dr. Williamson SHAW,W. T. 1936. Winter life and nesting studies of whose enthusiasticsupport and help made the study pos- Hepburn’s Rosy Finch in Washington State. Part 2. sible. Summer: nesting. Auk 53: 133-149. SHREEVE,D. F. 1980. Differential mortality in the sexes LITERATURE CITED ofthe Aleutian Gray-crowned Rosy Finch. Am. Midl. Nat. 104:193-197. AMERICANORNITHOLOGISTS UNION.’ 1982. Thirty-fourth STEJNEGER,L. 1885. Results of ornithological explora- supplement to the American Ornithologists’ Union tions in the Commander Islandsand in Kamtschatka. check-list of North American birds. Suppl. Auk 99, U.S. Natl. Mus. Bull. 29. no. 3. TWINING,H. 1938a. Natural history of the SierraNevada ARMSTRONG,R. H. 1977. Weather and climate, p. 53- Rosy Finch (Leucostictetephrocotis duwsoni). M.A. 58. In M. L. Merritt and R. G. Fuller [eds.], The thesis, Univ. of California, Berkeley. environment of Amchitka Island, Alaska. ERDA Rep. TWINING,H. 1938b. The significanceof combat in male TID-267 12, National Technical Information Service, Rosy Finches. Condor 401246-247. Springfield, VA. TWINING,H. 1940. Foraging behavior and survival in BAILEY,A. M., AND R. J. NIEDRACH. 1965. Birds of the Sierra Nevada Rosy Finch. Condor 42:64-72. Colorado. Denver Museum of Natural History, Den- WHEELOCK.I. G. 1920. Birds of California. A. C. McClure. ver, CO. and &I, Chicago. CAHN, A. R. 1947. Notes on the birds of the Dutch WHITE, C. M., F. S. L. WILLIAMSON,AND W. B. EMISON. Harbor area of the Aleutian Islands. Condor 49:78- 1977. Avifaunal investigations, p. 227-260. In M. 82. L. Merritt and R. G. Fuller [eds.], The environment DALL, W. H. 1873. Notes on the avi-fauna of the Aleu- of Amchitka Island, Alaska. ERDA Rep. TID-267 12, tian Islands. from Unalashka eastward. Proc. Calif. National Technical Information Service, Springfield, Acad. Sci., First Series 5:25-35. VA. DIXON, J. B. 1936. Nesting of the Sierra Nevada Rosy WILLIAMSON,F. S. L., AND W. B. EMISON. 1971. Vari- Finch. Condor 38:3-8. ation in the timing of breeding and molt of the Lap- FRENCH,N. R. 1959. Life history of the Black Rosy land Longspur(Calcarius lapponicus)in Alaska, with Finch. Auk 76:159-180. relation to differencesin latitude. Bioscience 2 1:70l- HANNA.G. D. 1922. The Aleutian ROSYFinch. Condor 707. 24:88-9 1. JOHNSON,R. E. 1965. Reproductive activities of Rosy Department of Zoology and CharlesR. ConnerMuseum, Finches. with suecial reference to Montana. Auk 82: Washington State University, Pullman, Washington 190-205. - 99164-4220. Received 16 April 1981. Final acceptance JOHNSON,R. E. 1972. Biosystematicsof the avian genus 28 June 1983. Leucosticte.Ph.D. diss., Univ. ofcalifornia, Berkeley. JOHNSON,R. E. 1977. Seasonalvariation in the genis Leucostictein North America. Condor 79176-86.