Spatial variation in age‐specific probabilities of first reproduction for Weddell seals

Abstract

Spatial variation in vital rates can affect the dynamics and persistence of a population. We evaluated the prediction that age‐specific probabilities of survival and first reproduction for Weddell seals would vary as a function of birth location in Erebus Bay, Antarctica. We used multi‐state mark–resight models and 25 years of data to estimate demographic rates for female seals. We predicted that probabilities of survival and first reproduction would be higher for seals born at near‐shore colonies or more southerly‐located colonies with consistent ice conditions. Contrary to predictions, results revealed higher age‐specific probabilities of first reproduction at offshore colonies relative to near‐shore colonies and no spatial variation in survival rates. For 7‐year old females (average age at 1st reproduction=7.6 years old) born at offshore colonies to mothers aged 10.8 years (average maternal age), probability of first reproduction was 0.43 (SE=0.07), whereas probability of first reproduction for females born at near‐shore colonies was 0.30 (SE=0.05) based on estimates from our top‐ranked model. Breeding probabilities following first reproduction were also higher at offshore colonies. Thus, our results (1) provide evidence of spatial variation in breeding probabilities, (2) reveal the importance of birth location on a female's vital rates, and (3) suggest that the effect persisted for many years. Birth‐colony effects may be attributed to spatial variation in prey availability, or to heterogeneity in female quality in this population. If females who are superior competitors consistently chose offshore colonies for pupping, pups born at these locations may have inherited those superior qualities and displayed higher probabilities of first reproduction, relative to seals born at other colonies. Further research into physical or food‐related differences among colonies may offer insight into spatial variation in breeding probabilities documented in this paper.