results in light of potential experimental bias

In many bird species early breeders have higher reproductive performance than late breeders from the same population. This could be caused by a reduction in environmental factors related to date per se (Date Hypothesis), or because poorer performers nest later (Parent Quality Hypothesis). We manipulated hatch date of Tree Swallows Tachycineta bicolor by switching clutches with different lay dates, generating broods with advanced or delayed timing, and assessed the impact of the experiment on nestling mass. The Date Hypothesis better explained the decline in nestling mass in the first half of the season, while the Parent Quality Hypothesis was supported in the second half. We also found that female mass loss was unintentionally reduced in advanced females and suggest that such impacts of the experiment on parent quality, or correlations between nestlings and their actual parents via heritability or maternal effects, could bias hatch-date manipulation experiments towards supporting the Date Hypothesis. Differential costs of incubation, either due to naturally low temperatures early in the season, or due to the unintentional manipulation of female incubation costs, appear to have driven support for the Date Hypothesis early in the season. Seasonal declines in reproductive performance have been observed in a wide variety of avian taxa. These declines result from a reduction, over the course of the season, in fitness-related parameters such as nestling mass and growth (Hochachka 1990, Sedinger & Flint 1991), clutch size (Winkler & Allen 1996), nestling survival (Norris 1993, Wiggins et a