INTRODUCTION
The adaptive value of phenotypic traits can be identified only when functional correlations with environmental variables are considered. The variation of the phenotypes between populations and individuals is often used to address this. However, individual phenotypic flexibility, where the adult phenotype can still be modified in response to environmental change, can be hidden by this approach (Piersma & Drent, 2003). A special case of such phenotypic change is life-stage cycling, i.e. seasonal changes along the lifetime of individuals that are reversible. Studying life-stage cycling allows inferring mechanisms of adaptation to the environment as the changes are well marked and predictable.
An outstanding case of seasonal phenotypic flexibility are the drastic but reversible morphological changes called Dehnel’s Phenomenon, observed in some small, short-lived mammals with high metabolic rates. In this phenomenon young animals reach a first maximum size in their first summer, followed by a size decline reaching a minimum size in winter, They then regrow in the spring along with sexual maturation. Best studied in the common shrew (Sorex araneus ), Dehnel’s Phenomenon entails a decrease in overall size, the size of the skull and other parts of the skeleton, but also the brain and many other organs and tissues, followed by regrowth (Dehnel, 1949; Pucek, 1965). Brain mass, for example, decreases by up to 30% from summer to winter and increases again by 10-17% during the next spring and summer (Bielak & Pucek, 1960; Lázaro et al. , 2018a). Braincase height, often used as a proxy for braincase size, decreases by up to 18% and regrows by up to 15% (Crowcroft & Ingles, 1959; Homolka, 1980; Yaskin, 1994). Importantly, Dehnel’s Phenomenon causes not just a rescaling of the animal, but each organ and even each brain region shows a unique pattern of the direction and magnitude of changes, resulting in several completely different phenotypes along the year (Yaskin, 1994; Lázaroet al. , 2018b). Also the length of the spine decreases and regrows seasonally as a result of shrinkage of the inter-vertebral disks (Saure & Hyvärinen, 1965; Hyvarinen, 1969). Some other species of shrews, and, as has recently been found, some mustelids, also show seasonal reversible shrinkage and regrowth at least of their skulls and brains (Dechmann et al. , 2017; LaPoint et al. , 2017).
Species known to exhibit Dehnel’s Phenomenon are small short-lived predators with very high metabolic rates, which do not hibernate or migrate during winter (Taylor, 1998; Ochocińska & Taylor, 2005). They remain active and dependent on high quality food year-round and the reversible changes of body and brain are hypothesized to be a winter adaptation to save energy (Mezhzherin, 1964; Pucek, 1970; Yaskin, 2011). While direct evidence of a link between the changes in overall size or specific organs such as the brain and individual survival is still lacking, reducing metabolically expensive organs, including the brain during winter, is thought to decrease overall energetic needs and thus food intake (Churchfield, 1982; Schaeffer et al. , 2020). This would then compensate for the disadvantages of being small, such as an increasingly unfavorable volume to surface ratio in winter (Bergmann 1848; Yom-Tov & Yom-Tov, 2005). In support of this, mass corrected energy consumption remains constant across seasons despite large differences in ambient temperature, which means overall energy use and thus food requirements of the size-decreased subadult winter shrews is lower than in the juvenile summer animals and especially in the adult individuals, whose mass doubles in the spring (Gębczyński, 1965; Taylor, Rychlik, & Churchfield, 2013; Schaeffer et al. , 2020). In addition, although this seasonal cycle occurs in every free-ranging individual (Lázaro et al. , 2017), the intensity of the size changes is exceptionally flexible. Captive shrews differ in the magnitude of seasonal change of skull size when ambient temperature is manipulated (Lázaro et al. , 2019).
Ambient conditions thus play an important role for Dehnel’s Phenomenon, but whether as triggers or evolutionary drivers or both, remains unclear. Braincase changes associated with Dehnel’s Phenomenon in weasels (Mustela erminea and M. nivalis ) vary greatly in intensity and timing between populations at different geographic locations (LaPoint et al. , 2017). Previous studies on common shrews suggested a greater winter decrease in skull and body size in Northeastern Europe compared to Southwestern populations (Pucek, 1970; Spitzenberger, 2001). Similarly, the reorganization of brain structure differs greatly between two populations in Radolfzell (Southern Germany, Lázaro et al. 2018b) and Russia (Yaskin, 1994). This increase in the extent of seasonal size change from regions with milder winter conditions to regions with harsher winter conditions supports the hypothesis that Dehnel’s Phenomenon is a winter adaptation. In fact, this trend towards lower body size when facing harsh conditions also fits well to the contradictive morpho-geographical patterns observed inSorex shrews. Some Sorex species show lower body size at higher latitudes (Ochocińska & Taylor, 2003), again contradicting Bergmann’s rule that predicts larger bodies in colder climates to increase heat-preservation efficiency (Bergmann 1848, but see Zeveloff & Boyce 1988). Several small mammal species follow a ‘resource rule’ where body size is directly predicted by resource availability (McNab, 2010). This would explain why common shrews decrease body size in winter, and predicts a more pronounced summer to winter shrinkage in environments with harsher winters – at high latitudes. However, a review of latitudinal differences in seasonal body mass decline did not find any significant trend (Ochocińska & Taylor, 2003).
We compiled all published work on Dehnel’s Phenomenon, discuss progress made since the last literature review in 1970 (Pucek, 1970) and take advantage of the larger currently available dataset to tatistically test for the influence of geographic and climatic variables on the magnitude of Dehnel’s Phenomenon in S. araneus . We collected information from those studies which include changes in skull size and/or brain mass. From these studies, we also collected total body mass when reported and explored correlations of the intensity of Dehnel’s Phenomenon with climatic and geographic variables. We added our own data on braincase size, brain mass and body mass from new populations in Poland and body mass and braincase size from the Czech Republic to this dataset. We expected to find increasing strength of Dehnel’s Phenomenon along geographical gradient that fits increasingly harsh environmental conditions related to seasonality, as predicted by previous authors. In addition, we compiled information on Dehnel’s Phenomenon in other species and compared their results with S. araneus . Finally, we specifically investigated the variation in the structural changes within the brain associated with Dehnel’s Phenomenon between populations. We compared the divergent results from southern Germany (Lázaro et al. , 2018b) and Russia (Yaskin, 1994) with new data from a population in Northeastern Poland, situated geographically between these two. We expected to find intermediate values of structural change that would fit into a gradual, geographic pattern in this Polish population. The aim of this review is to establish an updated framework to study the evolutionary aspects of this fascinating phenomenon.