We collected and analysed hourly GPS data from Antarctic fur seal pups tracked from birth until moulting in order to investigate the drivers of movement patterns and activity levels during this critical life stage. We found that pup movement was characterised by a star-like pattern, with individuals repeatedly returning to a central location of low activity after bouts of directed exploration. The HMM showed that the probability of such active movement was highest during the day and was mainly influenced by breeding season, colony of birth, and age. Our findings provide new insights into the movement patterns of pinnipeds prior to nutritional independence and highlight the importance of the life-history stage and the neonatal environment on behaviour.
Movement in fur seal pups
A recent study of movement in Antarctic fur seal pups found evidence for sex-specific differences in habitat use prior to nutritional independence and the onset of overt sexual size dimorphism. Specifically, male pups were found to exhibit more risk prone behaviour with increasing age, traveling further at sea and spending more time in exposed beach habitats26. While this study highlighted the importance of an intrinsic factor, sex, on neonatal movement, the contributions of other factors, both extrinsic and intrinsic, remained largely unexplored. Our HMM incorporated multiple environmental and individual-based variables including the time of day, air temperature, wind speed, age, body condition, and sex. Rather than focusing on animals from a single colony, our study used tracking data from two breeding colonies with contrasting population density and terrain, while replication across two consecutive breeding seasons also allowed us to account for inter-annual variation in food availability.
Our hourly GPS data revealed a distinct pattern of terrestrial pup movement that has not been reported in previous studies. Specifically, we documented a star-like pattern characterised by directed movement away from and subsequent return to a location of low activity. This appears to be a genuine behavioural pattern rather than an artefact of our temporal scale of sampling, as GPS locations collected every five minutes from a subset of individuals revealed a very similar picture (Supplementary Fig. S4). Star-like movement patterns are typified by central place foragers, which make round trips between a central location and a foraging patch37. However, as fur seal pups are entirely reliant on their mothers for food until they moult, repeated movements away from a central location are unlikely to be related to resource acquisition. Instead, our results appear to be indicative of bouts of exploration within a defined radius around a suckling location, which would imply that Antarctic fur seal pups are ‘central place explorers’.
We would expect central place exploration in fur seal pups to be adaptive given that it entails energetic costs37 and might even make it more difficult for pups to be located by their mothers when they return from foraging trips38. We can envisage a number of possible explanations for this behaviour. First, high levels of activity may facilitate the development of the muscle mass necessary for future foraging success. Second, movement both within and beyond the confines of the beach where a pup is born may increase the scope for social interactions, which might be important for future mating success18. Third, a transition from beach habitats to the tussock grass offers increased protection against harsh weather conditions and may reduce the risk of a pup either being crushed by a territorial male or predated.
While all tracked individuals displayed this pattern of central place exploration, the distance travelled during bouts of exploration differed depending on the colony of birth. On average, pups born at FWB travelled shorter distances and stayed closer to their colony of birth, whereas SSB pups moved farther and in a larger variety of directions. Given that SSB pups tend to transition into the tussock grass later in life (see discussion below), one explanation for this difference may be that mother–pup pairs need to travel further inland to find an appropriate, unoccupied suckling location.
Results of the HMM
We used an HMM framework to infer the discrete behavioural states underlying pup movement and the probability of switching between these states given certain environmental and individual covariates. This integrated approach to decompose behavioural patterns has been applied across a wide range of species and habitats and has emerged as a powerful tool for animal movement modelling5. Our model uncovered a strong influence of year, with the probability of occupying the active state being higher in the first breeding season (2019), particularly for FWB pups. One possible explanation for this pattern could be that inter-annual differences in movement are adaptive. The first breeding season in our study had among the lowest breeding female numbers, pup birth weights, and foraging trip durations on record28, reflecting poor environmental conditions and low prey abundance39. In 2019, local densities at both colonies were correspondingly below average28. Given a greater risk of predation at low density28, pups might increase their activity to avoid harassment, pecking injuries, and predation by generalist seabirds such as southern and northern giant petrels40. Alternatively, higher levels of activity could be a consequence of increased stress hormone levels. As mothers spent more time at sea foraging in 201928, pups were subjected to longer periods of food deprivation. Such extended bouts of starvation in pinniped pups have been associated with an increase in glucocorticoids41, which are known to increase activity levels in some species42. Understanding the interrelationships among prolonged fasting, glucocorticoids levels, and activity in fur seal pups may help to shed further light on individual responses to environmental challenges.
Another clear extrinsic determinant was the time of day, with activity levels peaking just before solar noon. This observation is in line with a previous study of Galápagos fur seal pups18, but otherwise very little is known in general about diurnal patterns of activity among fur seals while ashore. Daily meter-resolution location data from densely packed territory holding Antarctic fur seal males reveal negligible movements from day to day29, implying that these animals remain more or less stationary. By contrast, telemetry studies of adult females have reported associations between nocturnal foraging behaviour and diel variation in the time of arrival and departure from the breeding colony43. Consequently, it is unclear to what extent our results can be extrapolated across the life history, especially given the likelihood of an ontogenetic shift in times of peak activity as pups transition from being reliant on milk to nutritional independence.
The HMM also revealed an increase in pup activity with age, but only for pups born at FWB. One explanation for this colony-specific difference could be that FWB pups are able to express their full behavioural repertoire throughout the course of ontogeny, becoming more active as they grow, whereas pups born on SSB are constrained by the high density of conspecifics. In particular, pups that traverse tightly packed harems on SSB run a higher risk of being crushed by a territorial male or bitten by a breeding female21,22. Alternatively, we recently found that FWB pups are more likely to be predated by generalist seabirds28. This might translate into an increase of predator-avoidance activity as pups mature, which would help to explain why activity no longer increases after around 40 days of age, when the majority of individuals have transitioned to the more sheltered tussock grass.
We found that pups from FWB tended to move into the tussock grass earlier than pups from SSB and spent proportionally more time in the tussock grass than in their colony of birth. While this earlier shift toward residing mainly in the tussock grass could simply be a consequence of increased activity, it might also be adaptive for FWB pups to move inland as quickly as possible given the higher risk of predation at low density28. In other words, pups may move into the more sheltered tussock grass earlier in life and remain there for longer periods of time in order to avoid predatory seabirds. However, differences in the topography of the two colonies might also influence the timing of this transition, as a steep gully separates the beach at SSB from the tussock grass, while FWB offers a more gradual transition between the two habitats. To disentangle the effects of density and topography would require a larger study embracing a greater diversity of breeding colonies.
Contrary to our initial expectations, the HMM found little to no effect of air temperature or wind speed on pup activity. This is surprising given that the pups in our study lacked the water-repellent fur of adults and were thus poorly protected from the elements23. Reduced activity35 and huddling behaviour44 have been documented in pinniped neonates as effective thermoregulatory behaviours to withstand the respectively hottest and coldest daily temperatures, so we originally anticipated a reduction in pup activity under marginal weather conditions. However, large variation in temperature during the course of this study was not observed on Bird Island (2019 mean temperature = 3.7 °C ± s.d. 1.3; 2020 mean temperature = 3.9 °C ± s.d. 1.4). Consequently, behavioural adjustments in activity for effective thermoregulation may not have been necessary in the context of our study. Alternatively, other climatic variables that we could not account for, such as precipitation or humidity, may have a disproportionate influence on activity levels. Future studies involving the direct observation of focal individuals under specific climatic conditions would help to address this question.
The HMM also showed that pup activity levels were largely unaffected by body condition. This was unexpected because Antarctic fur seal pups rely on their mother’s milk for nutrition before moulting and must tolerate bouts of starvation lasting up to 11 days while their mothers forage at sea19. As a result, pup growth is known to decline with prolonged maternal absence45, implying that fewer resources should be available for movement. Taken at face value, the lack of a relationship between body condition and movement may suggest that the short-term benefits of high activity, such as muscle mass development and increased social interaction, may outweigh the costs associated with diverting resources from growth. This would be in line with a previous study that found that Galápagos fur seal pups maintained high activity levels throughout bouts of starvation18. However, it is also possible that an effect of body condition could not be detected in our study because our model assumed that condition was constant between successive measurements, whereas in practice it will vary to some extent from day to day.
Despite a number of studies having shown that sex-specific differences in activity, habitat use, foraging, and diving behaviour are established early in life in several pinniped species13,14,15,16,17,46, including Antarctic fur seals26, our HMM did not uncover any obvious sex differences in activity. This is probably a consequence of the timeframe of our study. Jones et al.26, for example, only detected sex-specific differences in habitat use in Antarctic fur seal pups greater than 41 days old. Given that we focused on the time window from birth until moulting at around 60 days of age, the results of these two studies are consistent and lend support to the notion that sex-specific movement patterns take several weeks to become established.
Results of the post-hoc analysis
The post-hoc analysis showed that pups were significantly less active when their mothers were ashore, suggesting that pup activity is correlated to some extent with maternal foraging behaviour. This association is most likely a reflection of the utmost importance of milk consumption for pup survival and development. A mother may spend as little as 24 h and on average only two days ashore during each nursing bout43, so pups must maximise nutrient update during this time. Moreover, adult female fur seals frequently display aggressive behaviour towards foreign pups21, potentially limiting a pup’s opportunity for social interactions when the mother is present18. It is important to note, however, that although significant, the effect size of the correlation between pup activity and maternal presence was fairly small. This suggests that pups may not remain with their mothers for the entire maternal attendance period, but rather alternate between nursing and bouts of activity.
We also found that pups that died tended to be considerably less active. In our dataset, cause of death was assigned to all but one pup as either starvation (n = 5) or predation (n = 6), two factors that are difficult to untangle given that smaller, weaker (i.e. starving) pups may be more likely to be predated. It is possible that the reduced activity of pups that died could be related to poor body condition. While this would appear to contradict the lack of an overall association between activity and condition in the HMM, it is possible that activity may decline only after body condition falls below a critical threshold18, an effect we may not have captured due to the small number (n = 13) and early occurrence (median = 15 days) of mortalities in our dataset. Alternatively, causality could flow in the other direction, with less active pups being more likely to be predated. This would be more in line with our HMM results as well as with the hypothesis that increased activity is related to predator-avoidance behaviour.
Strengths, limitations, and future directions
Recent advancements in bio-logging technology and analytical methods have made tracking studies of small, juvenile individuals ethically feasible, cost effective, and analytically approachable. In this study, we have taken advantage of these advancements to implement one of the first in-depth analyses of the movement and activity patterns of Antarctic fur seal pups prior to moulting. We were able to uniquely incorporate individual, ecological, and environmental variation into our analyses by collecting time-series biometric data for all tagged individuals, which were sampled at two breeding colonies across two subsequent breeding seasons. Finally, we were able to tease apart how these intrinsic and extrinsic factors may influence movement by inferring behavioural states using a HMM.
An important limitation of our study, however, is our inability to establish causal relationships between hypothesised explanatory factors and activity. While this is an ever-present challenge when working with wild populations, future studies may consider pairing GPS tracking data with detailed behavioural observations of the focal individuals. This would provide a behavioural context for movement patterns and possibly allow for a more nuanced interpretation of the underlying behavioural states. A higher GPS sampling rate might also facilitate the resolution of finer-scale states. While this was not possible in the current study due to the limited battery capabilities of our tags, as technologies continue to improve future studies may not need to sacrifice sampling resolution for study duration. Finally, it would be interesting in the future to consider GPS tracking both mothers and pups to better understand the relationship between maternal attendance behaviour and pup activity. This approach might also shed light on how mothers find their offspring after returning ashore from foraging trips.
Source: Ecology - nature.com