Discussion

In this study, we found that calf survival during both summer and winter was influenced by several factors including the presence of predators, landscape and climatic features, human activities such as hunting, and behavioural traits such as the migratory strategy of the mother. Specifically, we observed that summer calf survival decreased with increasing bear density and varied significantly between years, giving support to one of the two predation hypotheses and to theclimate hypothesis . During the summer season, brown bears are one of the main predators of moose calves. Previous research has shown that moose calves can constitute a significant portion, ranging from 36% to 44% of the total energy content in the diet of brown bears (Opseth, 1998). In south-central Sweden, studies estimated a predation rate of 26% for moose calves by brown bears, with individual female bears killing an average of 6.8 calves per year (Swenson et al., 2007). The majority (92%) of brown bear predation on moose calves occurs when the calves are less than one month old (Swenson et al., 2007; Rauset et al., 2012). Our study revealed that the risk of calf mortality is more than twice as high in areas with high bear density compared to areas with low bear density. These findings align with a previous study that found a negative association between bear density and calf/cow ratio (i.e., the number of observed calves per female moose in autumn) in Sweden (Tallian et al., 2021). The lower summer survival rates and decreased autumn calf/cow ratios observed with increasing bear density likely reflect the impact of bear predation on neonate moose.
While we do not have information on cause-specific mortality for each moose calf included in the study, we were able to confirm at least two documented bear predation events during the first study year: one in the northernmost part of the study area, where a pair of twin calves were killed by bears, and one in the easternmost part where the female moose and her calf were observed being followed by a female bear and her two yearling cubs, and the calf was not seen again the following day. Overall, these findings highlight the significant influence of bear predation on the survival of moose calves during the summer season.
In ungulates, the survival of adult females is a crucial driver of population growth, but calf survival can account for up to 75% of the variation in population growth rates (Gaillard et al., 2000, Raithel et al., 2007). Previous research from Sweden asserted that bear predation was not an influential driver of moose population dynamics (Haglund 1974) but that study was made during a time with a much lower brown bear population size and did not have access to the modern tools of GPS-technology for field checks. Instead, our study supports the notion that bear predation on moose neonates is likely an influential factor driving population dynamics of moose (Swenson et al. 2007, Rauset et al. 2012, Tallian et al. 2021). This is likely to become even more relevant in the future as the brown bear population has been expanding in their southern range (Kindberg et al. 2011; Kindberg 2010) and now overlaps with the core areas of the wolf population.
In Scandinavia, wolves are important predators of new-born moose calves, which make up approximately 90% of wolf kills during the summer season (Sand et al., 2008, Tallian et al., 2017). Surprisingly, we did not detect a relationship between wolf presence and summer calf survival. The lack of such an association might be related to the spatial location of the collared female moose and wolf space use and behaviour during denning time. During the summer season, breeding wolves exhibit more restricted space use compared to winter, focusing their activities around the den area where they care for their pups (Fritts and Mech, 1981; Jędrzejewski et al., 2001; Walton et al., 2001; Zimmermann et al., 2019). During the denning period, Scandinavian female and male wolves had a mean attendance at the den site of 68% and 70%, respectively (Alfréeden, 2006). Consequently, the likelihood of a moose calf being killed by wolves may vary within wolf territories, with higher predation risk closer to the den site compared to farther away.
In our study, we characterized wolf presence based on whether a moose home range was inside or outside a wolf home range. By using this broad classification, we might have overlooked the spatial variation in predation risk within wolf home ranges. This could have resulted in a failure to detect a correlation between wolf presence and calf survival, which may have been influenced by a bias introduced by the sampling of collared females. In other words, our data might have included an overrepresentation of collared female moose that utilized areas with low wolf activity during the summer season.
We offer several potential ad hoc explanations for the lack of support for our forage opportunity hypotheses (habitat productivity and proportion of young forests) in our analysis of summer survival. Previous studies have indicated that during the summer season, the composition of moose home ranges tends to resemble the overall landscape, suggesting that moose utilize a greater variety of habitats compared to winter (Hjeljord et al., 1990; Nikula et al., 2004). This may indicate that although clearcuts and young forests are important habitats for moose, the more homogenous distribution of high-quality, nutritional plant forage during summer allows moose to find optimal feeding patches in different habitats, also explaining why productivity was unrelated to calf mortality. Moreover, Nicholson et al. (2014) found that during summer, females with calves had lower selection strength for young forests and tended to select older forests, which may also explain why we did not find a correlation between calf mortality and the proportion of young forests.
The observed annual variation in summer survival of neonates is likely an effect of differing climatic conditions. Neonate growth and survival can be negatively affected by hot and dry summers due to nutritional constraints (Cook et al. 2004), but also by the environmental conditions experienced by their mothers in the preceding summer and during pregnancy in the preceding winter and spring (Forchhammer et al. 2001; Eacker et al. 2016; Bastille‐Rousseau et al. 2016).
During autumn-winter, calf survival exhibited a negative association with increasing snow depth in the presence of wolves (predation*snow hypothesis ) and an increasing proportion of young forests in the mother’s home range (habitat hypothesis ). In addition, calf survival was higher for stationary moose calves compared to migratory calves, contrary to our initial prediction (migration hypothesis ). Deep snow can hinder the movement of both wolves and moose, but since the former have a lighter foot loading, they can often travel on top of the snow crust (Peterson 1974). Hence, deeper snow generally increases the vulnerability of moose to predation and therefore increases the hunting success of wolves (Kolenosky 1972; Peterson and Allen 1974; Haber 1977) and kill rate (Nelson and Mech 1986; Huggard 1993; Post et al. 1999). Contrary to expected, we found that deeper snow was positively correlated with calf survival the absence of wolves. Although surprising, this unexpected result may be confounded by spatial variations in snow depth within our study area: in the northernmost part, where wolves are absent, snow depth averaged 45 cm, whereas average snow depth in the central and southern parts (where wolves are present) was only 22cm. Thus, the absence of wolves and the generally higher snow depths in the northernmost region might be contributing to the observed positive relationship between calf survival and snow depth.
As food resources become scarcer during winter, moose commonly aggregate in young forest plantations to feed (Gundersen 2003), with Scots pine being the quantitatively most important food source (Månsson, 2007). In our study, we observed a negative relationship between calf survival and the proportion of young forests during autumn-winter. This finding supports the notion that clearcuts and young forests represent riskier habitats for moose, both in terms of predation by wolves (Gervasi et al., 2013; Ausilio et al., 2022) and hunting by humans (Ausilio et al., 2022).
Game harvesting has become the leading cause of mortality in many ungulate species (Allendorf et al. 2008; Darimont et al. 2015), with moose calves accounting for about 40% of total harvest of moose in Scandinavia (www.algdata.se; www.ssb.no). We did not find a relationship between harvest density and moose calf survival, but we found a significant negative correlation between hunting risk and autumn-winter survival: calves exposed to high hunting risk had twice the risk of mortality compared to calves exposed to low hunting risk. The lack of a relationship between harvest density and survival is most likely a result of the spatial scale at which both harvest statistics and harvest density were calculated, which is the moose management unit (MMU). Moose management units are larger (average ± SE = 1830 ± 140 km2,Wikenros et al. (2020)) than an average moose home range (68 ± 7 km2), which means that one MMU can include several hundreds of moose home ranges. Moose management units are in turn made up by many hunting teams. Unfortunately, harvest density at the hunting team level was not available, so we decided to average harvest density at the MMU-level, which may be a too coarse spatial scale to identify a significant correlation between harvest density and calf survival. Hunting risk was, on the other hand, estimated at a much finer spatial scale (hunting team level, Ausilio et al 2022) and was negatively correlated to calf survival, providing support for the hunting part of our predation-hunting hypothesis .
In our study area, the moose population is partially migratory, with some individuals moving between high-altitude summer ranges and low-altitude winter ranges. Migration is usually the product of balancing the costs and benefits of remaining in a range all year-round or moving to new areas. We found that the migratory strategy of females was unrelated to summer calf survival, while during autumn-winter, migratory females were associated with a lower probability of calf survival compared to stationary ones. This outcome was opposite to ourmigration hypothesis . A possible explanation is that migratory moose in our study area moved from summer ranges, that were often located in Norway where the hunting season stopped already in December, to winter ranges in Sweden where the hunting season continued to February. It is also possible that migratory moose were forced to select sub-optimal habitats that increased their exposure to hunting and wolf predation compared to stationary ones, which can select optimal habitats based on the perceived risk of hunting and foraging needs.
Both the wolf and brown bear populations have increased during the last decades and expanded their geographical distribution in Scandinavia (Liberg et al., 2012; Wabakken et al., 2011; Kindberg et al., 2011), and in many areas these two predators have overlapping ranges (Kindberg et al. 2011). A future scenario likely to arise is therefore increased predation pressure on moose, which will intensify the competition with human hunters for this shared prey (Jonzén et al. 2013). In areas where wolves and bears coexist with human hunters, the additive effects of two large carnivores and human harvest will most likely result in greatly reduced calf survival, which may influence population dynamics. In order to ensure a sustainable harvest of the moose population, calf survival is a crucial demographic parameter to take into consideration. The coexistence of wolves, brown bears, and human hunters presents challenges that can lead to reduced calf survival, ultimately influencing population dynamics. To ensure a balanced and sustainable harvest of the moose population, comprehensive management strategies need to account for the intricate interplay between predation, hunting, habitat conditions, and climatic factors.