674 Animal Behaviour, 33, 2 vegetation was lush, and forage apparently unli- mited. Infanticide in the laboratory occurred des- pite the abundant provision of commercial rodent food, fresh vegetation and water. Yellow-bellied marmots live in groups composed of one or more matrilines with an associating territorial male (Armitage 1984). Infanticide by males should be rare, because all juveniles in the group are likely to be the male's own offspring (Armitage et al. 1979). Females may benefit fi'om infanticide by remov- ing potential competitors from the colony. How- ever, in colonies composed of related individuals, a female committing infanticide may reduce her indirect fitness by killing relatives (Armitage et al. 1979). This was the case at Picnic colony: 573 killed a cousin, thereby decreasing her indirect fitness. However, if by committing infanticide an indi- vidual increases the probable success of her off- spring, the gain in direct fitness should greatly exceed the cost. Females normally share their burrows only with their young; this may be a means of reducing the opportunity of others to destroy their offspring. Infanticide has been observed only three times in marmots of the East River valley in over 2000 h of observation during which it could have occurred (Armitage, personal communication). Because of its apparent rarity, and in light of the costs and benefits cited above, it is difficult to determine the importance of infanticide as a reproductive stra- tegy of marmots. It may be an infrequently employed means to reduce competition for limiting resources and thereby provide an advantage to direct descendants. We thank K. B. Armitage, P. W. Sherman, and two anonymous reviewers for helpful comments on a previous draft. This work was supported in part by NSF grant DEB 81-21231 to K. B. Armitage. The sequence of authorship was determined alpha- betically. ALISON K. BRODY JAYE MELCHER* Department of Systematics and Ecology, University o f Kansas, Lawrence, KS 66045, U.S.A. * Present address: Zoology Research Building, University of Wisconsin, Madison, WI 53706, U.S.A. References Armitage, K. B. 1974. Male behaviour and territoriality in the yellow-bellied marmot. J. Zool. Lond., 172, 233-265. Armitage, K. 13. 1981. Sociality as a life-historytactic of ground squirrels. Oecologiea, 48, 36-49. Armitage, K. B. 1984. Recruitment in yellow-bellied marmot populations: kinship, philopatry, and indi- vidualvariability. In: Biology of Ground-dwelling Squir- rels: Annual Cycles, Behavioral Ecology, and Sociality (Ed. by J. O. Murie & G. R. Michener), pp. 377-401. Lincoln: University of Nebraska Press. Armitage, K. B., Johns, D. & Anderson, D. C. 1979. Cannibalism among yellow-bellied marmots. J. Mam- mal., 60, 205-207. Balfour, D. 1983. Infanticide in the Columbian ground squirrel, Spermophilus columbianus. Anita. Behav., 31, 949-950. McLean, I. G. 1983. Paternal behaviour and killing of young in Arctic ground squirrels. Anim. Behav., 31, 32-44. Michener, G. R. 1982. Infanticide in ground squirrels. Anim. Behav., 30, 936-938. Sherman, P. W. 1981. Reproductive competition and infanticide in Belding's ground Squirrels and other animals. In: Natural Selection and Social Behavior (Ed. by R. D. Alexander & D. W. Tinkle),pp. 311-331. New York: Chiron Press. Sherman, P. W. 1982. Infanticide in ground squirrels. Anim. Behav., 30, 938-939. Waterman, J. M. 1984. Infanticide in the Columbian ground squirrel, Spermophilus eohLmbianus. J. Mam- mal., 65, 137-138. (Received 14 March 1984; revised 30 May 1984; MS. number: As-268) Asynchrony of Song Series in the Bewick's Wren and Wrentit A strong, unambiguous song is important for long-distance avian communication. Many vari- ables within the sound environment potentially hinder the effective transmission of song (Wiley & Richards 1982), including the vocalizations of other organisms (Marler 1960). Adaptive re- sponses to minimize interference may include character and variance shifts in the vocal structure, as well as spatial or temporal shifts by the organism (Miller 1982). Two types of temporal displace- ments have been reported in birds: vocalizing between the individual vocalizations of other birds (Ficken et al. 1974; Gochfeld 1978) and asynch- ronous bouts of vocalizations (Cody & Brown 1969). Cody & Brown (1969) used the technique of cross-covariance to demonstrate that Bewick's wrens (Thryomanes bewickii) and wrentits (Cha- maea faseiata) sing their bouts of song asynch- ronously. This statistic is not standardized, and hence significance levels could not be ascertained. A comparable, standardized method is cross-corre- lation. In this study we re-examine the data of Cody & Brown (1969) using cross-correlation. In addi-