Problems with Removal Experiments Designed to Test the Relationship between Paternity and Parental Effort in a Socially Polyandrous Bird

Aspects of sperm-competition research in birds have progressed from purely descriptive studies that establish multiple paternity (e.g. Gowaty and Karlin 1984, Gavin and Bollinger 1985, Westneat 1987) to a more experimental approach that examines the relationship between paternity and male parental effort (e.g. M0ller 1988, Koenig 1990, Davies et al. 1992, Whittingham et al. 1993). These experiments take the form of removing resident males during the female's fertile period as a means of disrupting one of the cues that males presumably use to assess their paternity (Schwagmeyer and Mock 1993), and then retuming them back to their territory and observing their behavior. Parental care by a removed male is predicted to decrease continuously with decreasing paternity in species with more than one male breeder and caregiver at the nest (Whittingham et al. 1992). In addition, removed males with low or no paternity have been known to destroy eggs upon being released back onto their territory, forcing females to renest and thus allowing the males a further reproductive opportunity (Koenig 1990, Robertson 1990). We attempted to test the general prediction that paternal effort should decrease with decreasing paternity using the polyandrous Pukeko (Porphyrio porphyrio). Pukeko exhibit a variable mating system, with about half of our study population composed of groups with two to three unrelated, co-breeding males. Alpha males do not guard the female during her fertile period and will tolerate copulations by other males within the group. This results in shared paternity and low reproductive skew among males (as shown by DNA fingerprinting analysis), and equal contribution in parental effort (Jamieson et al. 1994, Jamieson 1997). According to the predictions, Pukeko males removed early in the egg-laying period should show a reduced effort in parental care in response to decreased paternity, whereas control males removed after the laying period should exhibit normal levels of parental care. As we will show, the results of testing the relationship between paternity and parental effort were inconclusive because residents showed unusually high levels of aggression toward removed males that were released back onto their territories for both experimental and control groups. However, we present the results of the experiment to provide a cautionary note