Life history theory predicts allocation of energy to reproduction varies with maternal age but additional maternal features may be important to the allocation of energy to reproduction. We aimed to characterize age-specific variation in maternal allocation and assess the relationship between maternal allocation and other static and dynamic maternal features. Mass measurements of 531 mothers and pups were used with Bayesian hierarchical models to explain the relationship between diverse maternal attributes and both the proportion of mass allocated by Weddell seal mothers, and the efficiency of mass transfer from mother to pup during lactation as well as the weaning mass of pups. Our results demonstrated that maternal mass was strongly and positively associated with the relative reserves allocated by a mother and a pup’s weaning mass but that the efficiency of mass transfer declines with maternal parturition mass. Birthdate was positively associated with proportion mass allocation and pup weaning mass, but mass transfer efficiency was predicted to be highest at the mean birthdate. The relative allocation of maternal reserves declined with maternal age but the efficiency of mass transfer to pups increases, suggestive of selective disappearance of poor-quality mothers. These findings highlight the importance of considering multiple maternal features when assessing variation in maternal allocation.

Each year since 2002, attempts were made to weigh a sample of mother-pup pairs at parturition, and later at 20 days after parturition for a mid-lactation mass and 35 days after parturition for a late-lactation mass. Not all mother pup pairs were available for weighing on target dates therefore weights were obtained 1-4 days after parturition, 15-25 days after parturition and 30-40 days after parturition. Pups were weighed using a spring scale or digital weighing platform and mothers were weighed using either a digital weighing platform, photogrammetric methods, or both (Ireland et al. 2006; Paterson et al. 2016). Due to the difficulty in obtaining maternal mass measurements with a weighing platform throughout lactation, photogrammetry has been used to estimate the mass of Weddell seal mothers since 2002. Photographs of mothers from 2002-2010 were taken using a two-dimensional photogrammetry technique described by Ireland et al. (2006). For photogrammetry data collected during 2012-2016, photographs of mothers were taken and analyzed using a three-dimensional photogrammetry technique described by de Bruyn et al. (2009). For each of the two photogrammetric techniques morphometric measurements were related to mass measurements using linear regression to obtain raw mass estimates and prediction errors.

Proportion Mass Allocation Model Data

The Proportion_Mass.csv file contains data for the proportion mass allocation model in the following columns: speno (individual identification number), season (year of observation), partmass_k (raw estimated maternal parturition mass in kg), latemass_k (raw estimated maternal late-lactation mass in kg), skip (indicator variable for experienced breeder skipping reproduction the previous year), firstbreed (indicator variable for first time breeder the previous year), prebreed (indicator variable for prebreeder the previous year), male (indicator variable for whether the pup is male), part.pred.err (prediction error associated with raw estimated maternal parturition mass), late.pred.err (prediction error associated with raw estimated maternal late-lactation mass), momage.z (standardized maternal age), momage_quad.z (standardized and squared maternal age), days.z (standardized number of days between maternal parturtion and late-lactation mass measurements), age.first.z (standardized age at first reproduction), bday.z (standardized birthdate), bday_quad.z (standardized and squared birthdate)

Mass Transfer Efficiency Model Data

The Transfer_Efficiency.csv file contains data for the mass transfer efficiency model in the following columns: speno (individual identification number), season (year of observation), mom.partmass_k (raw estimated maternal parturition mass in kg), mom.midmass_k (raw estimated maternal mid-lactation mass in kg), mom.days (number of days between maternal parturition and mid-lactation mass measurements), skip (indicator variable for experienced breeder skipping reproduction the previous year), firstbreed (indicator variable for first time breeder the previous year), prebreed (indicator variable for prebreeder the previous year), male (indicator variable for whether the pup is male), part.pred.err (prediction error associated with raw estimated maternal parturition mass), mid.pred.err (prediction error associated with raw estimated maternal mid-lactation mass),  momage.z (standardized maternal age), momage_quad.z (standardized and squared maternal age), daily.pup.gain (standardized daily pup mass gain from parturition to mid-lactation), age.first.z (standardized age at first reproduction), bday.z (standardized birthdate), bday_quad.z (standardized and squared birthdate)

Pup Weaning Mass Model Data

The Pup_Weaning_Mass.csv file contains data for the pup weaning mass model in the following columns: speno (individual identification number), season (year of observation), mom.partmass_k (raw estimated maternal parturition mass in kg), pup.wean.mass (weaning mass of pups in kg), skip (indicator variable for experienced breeder skipping reproduction the previous year), firstbreed (indicator variable for first time breeder the previous year), prebreed (indicator variable for prebreeder the previous year), male (indicator variable for whether the pup is male), part.pred.err (prediction error associated with raw estimated maternal parturition mass), log.momage.z (standardized log of maternal age), pup.birth.mass.z (standardized pup parturition mass),  age.first.z (standardized age at first reproduction), bday.z (standardized birthdate), bday_quad.z (standardized and squared birthdate)

* All standardized variable were centered using the mean and scaled by two standard deviations.