184 SHORT COMMUNICATIONS

ican birds. Vol. 5. Yale University Press, New WHITE, C. M., AND T. J. CADE. 1971. Cliff nesting Haven. raptorsand ravens along the Colville River in arc- POOLE,K. G., AND R. G. BROMLEY. 1988. Natural tic Alaska. Living Bird 10:107-150. history of the Gyrfalcon in the central Canadian WHITE, C. M., AND D. G. ROSENEAU.1970. Obser- Arctic. Arctic 41:3 l-38. vations on food, nesting, and winter populations USPENSKII,S. M. 1969. [Life in high latitudes:a study of large North American falcons. Condor 72: 113- of bird life.] Mysl Publishers, Moscow. 115. WALKER,D. A., ANDK. R. EVERETT.1987. Road dust and its environmental impact on Alaskan taiga and tundra. Arc. and Alp. Res. 19:479489.

The Condor 93:1X4-185 0 The CooperOrnithological Society I99 1

WHY HOUSE WRENS DESTROY CLUTCHES OF OTHER BIRDS: A SUPPORT FOR THE NEST SITE COMPETITION HYPOTHESIS’

STANISLAVPRIBIL AND JAROSLAVPICMAN Department of Biology, Universityof Ottawa, Ottawa, Ontario KIN 6N5, Canada

Key words: House Wren; Troglodytes aedon; egg receding from the House Wren nest; the first box was destruction;cannibalism: competition. 20 m from the wren nest. Each nesting box contained a dry-grass nest with one quail (Coturnix chinensis) House Wrens, Troglodytesaedon. are notorious for egg.The wren nestscontained either eggsor small nest- destroying clutches of other birds, including those of lings of the first brood. The nestingboxes were checked conspecifics.The destruction usually involves pecking after 6 hr, 1 day and 3 days. During the three-day holes in eggs and removing the soft lining from the period. males in six of 25 (24%) territories started nest cup; ifsmall nestlingsare present, they may also building a nest in one of the boxes (males build a rough be killed (Kendeigh 1941). It has been urouosed that twig nestwhich femalescomplete with soft lining; Ken- wrens attack clutches 1) to acquire suitable cavities for deigh 1941). The males first punctured and removed their own use, 2) to consume the contents of broken the quail egg, then removed the grassnest, and finally eggs,3) to force other birds to breed farther away, hence startedbringing in twigs. Five males brought in several reducingcompetition for food, and 4) to free potential centimeters of twigs, one male completed three-quar- mates (in caseof conspecificnests), thereby increasing ters of the nest. The males choseboxes which were 20 chancesof becomingpolygamous (Belles-Isles and Pic- m (one bird), 80 m (one bird) and 100 m (four birds) man 1986). Becauselittle evidence for the hypotheses from their nests. We assume that the majority of the is available (see Quinn and Holroyd 1989), we report males were resident males because a) in a separate severalcases of clutch destructionfollowed by a cavity experiment, we equippeda nestingbox with a trap and takeover, and provide evidence against the egg con- placed it 20 m from several active House Wren nests sumption hypothesis. (seven out of ten (70%) trapped wrens were resident Observations reported here are part of a long-term males); and b) the transectsof nesting boxes used in study of the House Wren breeding ecologybeing con- this test were always directed away from neighboring ducted in the Mer Bleue Bog conservation area near House Wren nests. Ottawa, Ontario, Canada. In May-July 1989, we con- In the second experiment, a nesting box was suc- ducted two tests in which we offered breeding House cessivelyintroduced into territories of 11 males. Each Wrens nesting boxes with experimental clutches (un- nesting box contained a House Wren nest and one published data). The nesting boxes were made of ply- House Wren egg. The nests and eggs were obtained wood and each was attached to a stake 1.5 m above from failed nesting attempts of other pairs. The males ground. Wren responsesto those boxes provide data were either unmated males defending a territory, or on the plausibility of two of the above hypothesesfor mated males whose females were incubating. We di- the function of egg-destructionby House Wrens. rectly observedall 11 males enter the box and remove In the first test, we introduced five nestingboxes near the egg, usually by carrying it in their beaks through eachof 25 active House Wren nests(125 boxesin total). the entrance and then dropping it below the box, or The boxes were placed 20 m apart along a transect flying a short distanceand dropping it into vegetation. Each male, except two, spent less than 8 set inside the box; one male spent 11 and one 23 sec. The fact that all eggswere removed from the nestsand dropped into I Received 7 May 1990. Final acceptance15 October vegetation,and that the males remained inside the nest 1990. for a short time suggeststhat conspecificeggs are not SHORT COMMUNICATIONS 185 pecked to accessegg contents. Instead, breakagemay imum amount of time handling the broken eggs)argues facilitate graspingand removal of the egg, or simply against the egg consumption hypothesis. Tests of the render the egg inviable. Consumption of egg contents remaining hypotheseswill require more information demandstime, and the benefitsof leaving immediately on the degree of foraging similarity between House may outweigh the nutritional gains.The resident male Wrens and sympatricpasserines, the movement ofbirds guardshis territory againstintruders (Johnson and Ker- whose nests are destroyed by House Wrens, and the mott 1989) and checks the nesting cavity during the chance that the intruding male has to mate with the female’s absence(i.e., when the female is foraging; S. female whose nest the male destroyed. Pribil, personal observation). If the intruder is con- This work was supportedby NSERC operatinggrant fronted inside the cavity, it may sustaininjuries or may to J. Picman. even be killed by the resident male, as was suspected in two casesreported by Belles-Islesand Picman (1987). LITERATURE CITED Nesting cavities suitable for House Wrens are lim- ited in our area. This is evident from the fact that the BELLES-ISLES,J. C., ANDJ. P~CMAN.1986. House Wren introduction of nesting boxes was followed by a dra- nest-destroyingbehavior. Condor 88: 190-l 93. matic increasein size of the studypopulation from less BELLES-ISLES,J. C., ANDJ. PICMAN. 1987. Suspected than 2-4 pairs to 3845 pairs in different years (J. adult intraspecifickilling by House Wrens. Wilson Picman and S. Pribil, unpublished data). The obser- Bull. 99(3):497498. vations of eggdestruction followed by a cavity takeover JOHNSON,L. S., AND L. H. KERMO~T. 1989. Terri- supportthe hypothesisthat the eggdestroying behavior torial intrusions in the House Wren Troglodytes in House Wrens may have been favored by intense a&on: evidence for the sperm competition hy- competition for nesting cavities. We cannot, however, pothesis. Omis Stand. 20:89-92. establishthe plausibility of the remaining hypotheses KENDEIGH,S. C. 1941. Territorial and mating be- becausea) the test specificallyexamined the nest site havior of the House Wren. Ill. Biol. Monogr. 18(3): competition hypothesis;b) results of the test are not l-120. inconsistentwith the other hypotheses;and c) the four QUINN, M. S., AND HOLROYD,G. L. 1989. Nestling hypothesesare not mutually exclusive. On the other and egg destruction by house wrens. Condor 9 1: hand, the fact that House Wrens apparently failed to 206-207. consumecontents of the broken eggs(i.e., spent a min-

The Condor93:185-189 0 The Cooper Ornithological Society 1991

MITOCHONDRIAL DNA VARIATION INDICATES GENE FLOW ACROSS A ZONE OF KNOWN SECONDARY CONTACT BETWEEN TWO SUBSPECIES OF THE BROWN-HEADED COWBIRD ’

ROBERTC. FLEISCHER~ Department of Biology, Universityof North Dakota, Grand Forks, ND 58202

STEPHEN I. ROTHSTEIN Department of Biological Sciences,University of California, Santa Barbara, CA 93106 LINDA S. MILLER Department of Biology, Universityof North Dakota, Grand Forks, ND 58202

Key words: mitochondrial DNA; Brown-headed introgressionof various traits or allelesis directly mea- Cowbird;gene flow; hybridization. sured within and around such zones. However, few studieshave histor$al data on the timing of secondary The dynamics of gene flow are often easily observed contact such that actual rates of introgressioncan be and analyzed where differentiated taxa meet and in- calculated (Endler 1977, Rand and Harrison 1989). terbreed in hybrid zones (Endler 1977; Barton and Studieswith historical data often involve specieswith Hewitt 1985, 1989). In most studies, the amount of well-documented range extensions (e.g., Gill 1980, Cookeet al. 1988,Fleischer and Rothstein 1988,Echelle and Connor 1989). Our studiesof geographicvariation I Received 21 May 1990. Final acceptance29 Oc- in morphometric and calorimetric characters in the tober 1990. brown-headed cowbird have indicated recent and ex- z Present address:National Zoological Park, Smith- tensive geneflow between the differentiated subspecies sonian Institution, Washington, DC 20008. Molothrus ater obscurusand M. a. artemisiae in the