Keywords (6-10)
Reproduction, phenotypic plasticity, microevolution, climate change, quantitative genetics, Marmota flaviventer

Abstract:

With global climates changing rapidly, animals must adapt to new environmental conditions with altered weather and phenology. Key to adapting to these new conditions is adjusting the timing of reproduction to maximize fitness. Using a long-term dataset on a wild population of yellow-bellied marmots (Marmota flaviventer ) at the Rocky Mountain Biological Laboratory (RMBL), we investigated how the timing of reproduction changed with changing spring conditions over the past 50 years. Marmots are hibernators with a four-month active season. It is thus crucial to reproduce early enough in the season to have time to prepare for hibernation, but not too early so as snow cover prevents access to food. Importantly, climate change in this area has increased spring temperatures by 5 oC and decreased spring snowpack by 50 cm over the past 50 years. We evaluated how female marmots adjust the timing of their reproduction in response to the changing conditions and estimated the importance of both genetic variance and plasticity in the variation in this timing. We showed that, within a year, the timing of reproduction is not as tightly linked to the date a female emerges from hibernation as previously thought. We reported a positive effect of spring snowpack but not of spring temperature on the timing of reproduction. We found inter-individual variation in the timing of reproduction, including low heritability, but not in its response to changing spring conditions. There was directional selection for earlier pup emergence date since it increased the number and proportion of pups surviving their first winter. Taken together, the timing of marmot reproduction might evolve via natural selection, however, plastic changes will also be extremely important as long as plasticity is not limited. Further, future studies on the marmots should not operate under the assumption that females reproduce immediately following their emergence.
Introduction :
Life history traits are those that impact the fitness of an individual through survival and/or reproduction (Braendle et al., 2011). The seasonal timing of these traits is heavily dependent on environmental conditions (Brommer, 2000; Bronson, 2009). These environmental conditions can vary inter-annually (Bright Ross et al., 2020) and seasonally, in both the mean value of the environment and in the timing of important events (e.g., when food becomes available; Nusseyet al., 2005). Animals must react to these yearly and seasonal variations by adjusting the timing of their life history traits to coincide with the environmental conditions that will maximize survival and/or reproduction. For example, timing egg laying date so that offspring emerge when food availability is at its highest (Nusseyet al., 2005), timing changes in coat colouration to match seasonal changes in the environment and thus avoid predation (Millset al., 2013), and timing emergence from hibernation to emerge late enough that food resources are not covered by snow, but early enough to maximize the length of the active season (Edic et al.,2020). However, climate change may directionally shift when these environmental conditions occur (Mills et al., 2013; Nusseyet al., 2005). Indeed, this has been seen with shifts in the timing of food availability (Nussey et al., 2005) and changes in average season lengths (Cordes et al., 2020). These changes can lead to mismatched timing between animal behaviours and optimal environmental conditions if animals are not able to adjust their timing adequately. Consequently, fitness can be negatively impacted, and indeed declines in both reproductive success (Nussey et al., 2005) and survival (Cordes et al., 2020) have been reported. Animals can alter the timing of their life history traits to coincide with the changed timing of environmental conditions through phenotypic plasticity and/or microevolution (Boutin & Lane, 2014).
Phenotypic plasticity occurs when a phenotype changes in response to a changing environmental condition (Nussey et al., 2007). This can be measured in a wild population by observing how a trait that is expressed multiple times during an individual’s life changes in response to changes in climate (Nussey et al., 2007). Phenotypic plasticity is an important mechanism by which individuals respond to their environment as it allows for a fast change in the phenotype that can accurately track sudden changes in environmental conditions (Charmantier et al., 2008). Plasticity therefore also provides a potentially important solution in terms of climate change response because while the trend in climatic changes is expected to be directional (Boutin & Lane, 2014), variability is expected to increase (Childs et al. , 2010). Indeed, Charmantier et al. (2008) outlined how the egg-laying date of a population of great tits (Parus major ) in the UK has become earlier by 14 days. They attributed this finding to phenotypic plasticity in response to changing spring temperatures with microevolution playing no role. While there is population-level plasticity, they reported no inter-individual variation in plasticity amongst females. This lack of variation indicated that there is no individual by environment interaction (IxE; Nussey et al., 2007) for the date of egg-laying in this population.
However, the capacity of phenotypic plasticity to respond to these global environmental changes may be limited over the long-term (Boutin & Lane, 2014). Certain studies have shown that to fully adapt to changes in climatic conditions, populations will need to undergo microevolutionary changes as phenotypic plasticity will not be enough on its own (Phillimore et al. , 2010; Mills et al., 2013). For instance, Phillimore et al. (2010) examined the response of first spawning date in British populations of the common frog (Rana temporaria ) to changes in temperature. They found that even though spawning date is plastic, plasticity alone will not be enough to adapt to the degree of change in temperature expected in these areas. They outlined that the upper limit of plasticity in this population is a 5–9-day advancement in spawning date. However, to maintain fitness, these populations will need to spawn 21-39 days earlier. Therefore, microevolutionary changes will be needed to allow the population to fully adapt. Further, the timing at which certain animals molt to match their environment has become mismatched with current environmental conditions (Mills et al., 2013). In these systems, spring snow melt now begins before coat colouration has changed, leaving individuals vulnerable to predation (Mills et al., 2013). Mills et al.(2013) studied wild populations of snowshoe hares (Lepus americanus ) and determined that while there was plasticity in the rate of molting in the spring, it is not enough to cope with the magnitude of expected climate change. Indeed, they reported that with current climate change projections, in the next hundred years hares will be exposed four to eight times longer than presently unless microevolutionary changes occur in the timing of spring molt.
These microevolutionary changes can occur in plasticity or in the mean of the trait if there is inter-individual variation, heritability, and selection. In terms of microevolution in plasticity, Nussey et al. (2005) found increased selection pressures for higher plasticity in the egg-laying date in a population of great tits (Parus major ) in response to warming temperatures. They concluded that selection for increased plasticity is being observed because those individuals that can adjust the timing of their egg-laying date to better match environmental conditions (i.e., those that have increased plasticity), will have more food available for their young when they hatch. This would increase offspring survival compared to those individuals with lower plasticity. They also found that plasticity is heritable and there is inter-individual variation in the degree of plasticity expressed. In terms of microevolution on the mean of the trait, Réale et al. (2003) reported that over the course of a decade, the date of reproduction in a wild population of red squirrels (Tamiasciurus hudsonicus ) became 18 days earlier. By finding that this trait is heritable and under strong negative selection, they concluded that this change in birth date is due, in part, to microevolutionary changes in the mean of the trait (but see Boutin & Lane, 2014). However, while microevolution may offer a long-term solution in responding to climate change, there may be a mismatch between the two (Gienapp et al., 2007; Boutin & Lane, 2014). Microevolution is a relatively slow, and not easily adjustable process in contrast to phenotypic plasticity; this may prove problematic as a response to climate change which can occur quickly and vary inter-annually (Charmantier et al., 2008).
Despite the extensive background research on these topics, studies examining the relative contributions of phenotypic plasticity and/or microevolution in the response of a wild population to climate change, remain rare (de Villemereuil et al. , 2020; Nussey et al. , 2005; Lane et al., 2018). This is because studies of this nature require long-term data, a known pedigree, a sizeable population (Boutin & Lane, 2014; Nussey et al., 2005), and a study site that is impacted by climate change (Lane et al., 2018). One example of a study system that meets these requirements are the yellow-bellied marmots (Marmota flaviventer ) of the Rocky Mountain Biological Laboratory (RMBL) in Colorado, USA. Since 1962, this study system has generated annual data on individual marmots, maternity has been assigned behaviourally since the study’s beginning and paternity assignment began in 2000. For the past 50 years, there has been a steady increase in mean temperatures and decrease in mean snowpack, with one of the fastest reported changes globally. Specifically, there has been an increase of 5oC in average spring temperatures and a decrease of 50 cm in average spring snowpack over the past 50 years (Figure S1).
Coupled with these climatic changes are changes in the marmots’ life history: adult emergence date from hibernation has advanced (Edicet al., 2020), pups are being weaned earlier (Ozgul et al., 2010), and overwinter survival is decreasing while summer survival is increasing (Cordes et al., 2020). Indeed, a marmot’s life history is heavily constrained by climate (Cordes et al., 2020). During the short four-month growing season, marmots must gain as much weight as possible to survive hibernation (Ozgul et al., 2010). Notably limiting the ability of marmots to gain mass is when they are born their first year and when they reproduce the years after. Individuals that are born later are less likely to survive overwinter than those born earlier in the season (Monclús et al. , 2014). Similarly, if a female is investing energy and resources into lactating late into the season, she may also have a harder time surviving overwinter than those that invest earlier (Andersen et al. , 1976). We might assume that marmots should emerge and reproduce earlier to increase the length of this crucial growing season. However, emerging and reproducing too early also poses problems. If there is still snow on the ground covering food resources when marmots emerge, they must draw on depleted energy stores for longer (Cordes et al., 2020). This could potentially lead to starvation, decrease the number of pups in a litter, or cause marmots to forgo reproduction altogether (Inoyueet al., 2000). Nevertheless, there has been a documented advance in the emergence date from hibernation (Edic et al., 2020).
Since marmots are thought to reproduce immediately following emergence, we expected the timing of reproduction, pup emergence date, and adult emergence date to be strongly linked and to follow a similar pattern. However, how the timing of reproduction varies from year to year and the impact of climate change on the timing of reproduction remains unknown. Therefore, we were interested in examining whether the timing of reproduction is changing in response to changes in average spring temperatures and average spring snowpack. Given that these changes can occur through microevolution and/or phenotypic plasticity, we further investigated the relative contributions of each by examining whether the trait is heritable, whether there is selection on the timing of reproduction in response to climate change, and whether there is phenotypic plasticity in the trait in response to changes in both average spring temperature and snowpack. As reproduction is expected to occur immediately following adult emergence, we expected the results for the timing of reproduction and adult emergence to be similar. Therefore, following Edic et al. (2020), we expected there to be low, but estimatable heritability for the trait, strong plasticity, and an impact of both spring temperature and snowpack. We further expected IxE in plasticity and stabilizing selection on the timing of reproduction.

Methods: