Body mass declines during wing moult in numerous, but not all, populations of Anatidae. We assessed two leading hypotheses for body mass dynamics during the wing moult: (1) body mass dynamics are adapted to attain a target body mass at the end of the wing moult (restraint hypothesis), versus (2) body mass dynamics reflect environmental constraint on the nutrient-energy balance during wing moult (constraint hypothesis). We used regressions of mass of breeding female Black Brent, Branta bernicla nigricans, on ninth primary length (a measure of moult stage) for each of 16 years to assess mass dynamics during wing moult and used regression equations to predict mass at the beginning and end of wing moult each year.  We also included gosling mass at 30 days (an indicator of forage availability) in models of adult mass to assess how mass dynamics varied as a function of foraging conditions. Predicted body mass (± 95% CI) at the start of wing moult (ninth primary = 0 mm) varied significantly among years ranging from 1032 ± 52g to 1169 ± 27g.  Similarly, predicted mass in late wing moult (ninth primary = 142 mm) ranged from 1048 ± 25g to 1222 ± 28g. Rate of mass gain was significantly related to gosling mass at 30 days: interaction between adult ninth primary length and gosling mass = 0.0031 ± 0.0020 (P = 0.003). Females initiated wing moult at lower body masses, gained mass more rapidly, and ended wing moult heaviest when goslings were heaviest. Body mass dynamics of female Black Brent during wing moult were consistent with the constraint hypothesis.  Positive association between gosling mass and rate of body mass gain by adult females during wing moult was also consistent with the constraint hypothesis. 

Moulting adults and broods were herded into corral traps during the adult remigial moult (9th primary x  = 59.56 mm ± 30.62 [SD]) after which we applied metal U. S. Geological Survey bands and uniquely engraved plastic tarsal bands (Sedinger et al. 1997).  We determined sex by cloacal examination and assessed breeding status of female Brent based on the presence or absence of a brood patch (Owen 1980). We measured body mass of goslings and adults using an electronic balance (± 1.0 g, after 2000) or a spring scale (± 10 g, before 2000), and total tarsus and culmen using dial calipers (± 0.1 mm; Dzubin & Cooch 1992).  Goslings had been tagged in their webs with uniquely stamped fish fingerling tags while in the nest (Alliston 1975, Sedinger & Flint 1991) so we knew their age in days.  We measured ninth-primary length (± 1 mm) of female Brent with brood patches on the distal side of the rachis between the ninth and tenth primaries (Dzubin & Cooch 1992, Taylor 1995).  All handling of animals was approved by the Institutional Animal Care and Use Committee of the University of Nevada Reno, most recent protocol no. 00056.