GRAY-HEADED JUNCO 1121 



making for the valley." Howard Rollin (pers. comm.) saw a northern 

 shrike in January "chase and kill a pink-sided juuco" at his ranch on 

 the Colorado plains dm-ing severe weather wlien mice, the shrike's 

 usual diet, remained hi shelter. An occasiontd graj'-head loses its life 

 at a banding or feeding station; at the Grand Canyon a banded 

 dorsalis was reported killed in November "by an Abert Squirrel in a 

 trap" (unpublished banding notes), and at nearby Flagstaff, Hargrave 

 (1936) reports an immature bird killed in a trap by a shai-p-shinned 

 hawk m Februarj^ Also at Flagstaff, Hargrave (MS.) says the first 

 brood raised in 1940 in the "vine" nest previously mentioned was 

 "killed by dogs the day after they left the nest." 



Although H. F. Friedmaun (1949) lists the white-winged and the 

 Cassiar (J. hycmalis cismontanus) j uncos as victims of cowbird 

 parusilism, there seem to be no records of victimization of the gray- 

 headed junco. The breeding habitats of this junco and the brown- 

 headed cowbird (Molothrus ater) seldom, if ever, overlap. 



Fall. — After completing their breeding activities the northern 

 caniceps form small groups, presumably of adults and the young of the 

 season, and start a southward movement. Such groups are frequently 

 encountered weeks before late breeding is finished in the same area. 

 Twomey (1942) observed small flocks as early as July 25 at 6,000 feet 

 in northwestern Colorado. The late-summer migration appears to be 

 southward and frequently upward as higher mountains are encoun- 

 tered. Occasionally small moving bands occur 2,000 feet or more 

 above timberline, where I have seen them as late as August 29. The 

 first bh'ds usually reach the "lowlands" about October 1, but their 

 advent may be hastened by an e&rlj mountain snowstorm. Graj^- 

 heads soon join the migrating flocks of other junco species from the 

 north, which eastward are mostly J. o. mearnsi, and westward J. o. 

 montanus. In Arizona and New Mexico, the J. c. caniceps and the 

 Oregon j uncos associate commonly with the local, less migrator}' 

 dorsalis. 



The latitudinal limits of the winter range of caniceps are well south 

 of those of the breeding range — generaUy 100 to 150 miles in the north 

 and 500 miles or more in the south. Thus presumably the bhds 

 migrate not only vertically but a considerable distance southward. 

 In his study of the winter birds in Utah, Hayward (1935) concluded 

 that heavy snowfall covermg "a large part of the available ground 

 food * * * for a considerable length of time" determines the northern 

 limit of the wmter range of tlie gra3'-headed and other j uncos. 



To learn whether the wintering or migrating gray-headed j uncos at 

 Boulder, Colo., breed, as is generally presumed, in the mountams 

 immediately to the west or northwest, Dr. and Mrs. John N. Hough 

 color-banded 234 birds from October through April of 1957-58 and 



