INTRODUCTION
Evolution has selected to endorse resilience for buffering environmental
impacts, protecting biological systems from failure. Complex responses
to these impacts occur at all organizational levels, from cells and
organs to populations and ecosystems. At the individual level,
physiological responses are at the basis of resilience to enhance
survival and reproduction. One particular life history strategy that has
evolved to cope with environmental stress is dormancy or torpor.
Dormancy is a physiological adaptation in some plants and animals, that
can remain torpid for weeks, months and even years (Nowack et al.2017; Withers & Cooper 2019). In some mammals inhabiting seasonal
ecosystems, dormancy during winter, also called hibernation, reduces the
impacts of climatic stress. Hibernating species may reduce their
metabolic rate for long periods, which results in higher survival and
lower fecundity than close phylogenetic species that do not hibernate
(Withers & Cooper 2019). The advantages of hibernating may explain the
evolutionary success of ancestral mammals that survived the mass
extinction at the Cretaceous–Palaeogene boundary following an
unprecedented environmental perturbation (Lovegrove et al. 2014).
A much higher than expected rate of recent extinctions in mammals has
been recorded for homeothermic species, whereas hibernating species seem
to cope better with environmental impacts due to anthropogenic global
change (Geiser & Turbill 2009). Some theoretical models show that in
harsh environments, hibernation may be all that allows population
persistence (Tuljapurkar & Istock 1993).
The understanding of the physiological dynamics by which air temperature
influences the seasonality of life histories is crucial to assess the
resilience of hibernating species to global warming (Caro et al.2013). In recent times, wearable electronic loggers have allowed
researchers to analyze the dynamics of physiological systems, such as
body temperature fluctuating between activity and hibernation states
(Chmura et al. 2018; Scheffer et al. 2018). A
straightforward pattern emerges when examining these studies: body
temperature in hibernating mammals such as squirrels, marmots, tenrecs,
echidnas, dormice and hamsters, changes abruptly between these states
(Hut et al. 2002; e.g. Talaei et al. 2011; Williamset al. 2011; Hoelzl et al. 2015; Iwabuchi et al.2016). However, this highly stochastic physiological system has never
been explored to assess whether these sudden switches correspond to
tipping points between alternative basins of attraction through critical
transitions. Critical transitions are non-linear, abrupt responses of
some biological systems subjected to some type of environmental stress.
These transitions occur when a threshold value for resilience has been
crossed due to the cumulative stress, and beyond this tipping point,
there is a sudden shift of state (Scheffer et al. 2001; Scheffer
2009). Hibernation has either been studied from the perspective of
resilience, which is the property that mediates transition between
alternative stable states. Hibernation dynamics also appear as a
promising candidate for assessing our capacity for anticipating
transitions between states. The anticipation of responses to stress,
especially when responses are non-linear (e.g. critical transitions)
remains a challenge for most biological systems (Scheffer et al.2009; Ghadami et al. 2018).
For hibernating dynamics, critical transitions would be a particular
type of transition in which a gradual change in air temperature, once
past a threshold value, would trigger an overwhelming shift of body
temperature between the contrasting states of hibernation and activity
(Scheffer 2009). Other physiological systems, such as functional
heterogeneity of some progenitor blood cells, show critical transitions
through bifurcation thresholds (Mojtahedi et al. 2016). Here, we
deployed body temperature loggers to a small mammal, the edible dormouse
(Glis glis ), to assess the occurrence of criticality in
hibernation dynamics and to deepen our understanding of resilience of
hibernating animals to cope with environmental stress. We also assessed
the performance of statistical leading indicators for anticipating a
critical transition between stable states of activity in summer and
hibernation in winter. Interestingly, the studied dormice are from a
habitat at the edge of the distribution range, where patches of
cold‐temperate forests, their preferred habitat, are regressing due to
rising temperatures in recent decades (Peñuelas & Boada 2003).