The snow and the willows : the ecological responses of the alpine dwarf shrub "Salix herbacea" to climate change

Current changes in shrub abundance in alpine and arctic tundra ecosystems are primarily driven by warming, changes in snow cover time and shifts in biotic interactions. However, while taller shrub communities are expanding, dwarf shrubs show reductions under climate warming. The aim of this thesis was to disentangle the phenological, growth and fitness response of a widespread prostrate shrub to warming and changes in snowmelt time, investigate shifts between competition and facilitation in communities along environmental stress gradients, and determine trait plasticity in response to snowmelt change. To this end, we investigated phenology, vegetative growth and reproductive traits in Salix herbacea, a widespread, long-lived alpine and arctic prostrate shrub, along its elevational and snowmelt microhabitat range over three years in Davos, Switzerland. To examine environmental drivers of trait variation, community interactions and potential for plastic responses to global changes, we used a space-for-time substitution study, a neighbor removal experiment, and reciprocal transplant experiment.
Earlier snowmelt was associated with longer phenological development periods, an increased likelihood of herbivory and fungal damage, fewer stems and no increase in end-of-season wood reserve carbohydrates. Furthermore, while early snowmelt was associated with an increased proportion of flowering stems, the number of fruiting stems was not, as fruit set decreased significantly with earlier snowmelt. Warmer temperatures at lower elevations were associated with decreases in stem number and wood low-molecular weight sugars, and increases in proportion of stems fruiting.
In the reciprocal transplant experiment, S. herbacea generally started to grow earlier, but had a longer development time and produced smaller leaves on ridges relative to snowbeds. The phenological changes did not influence clonal or sexual reproduction, but smaller leaves in early-snowmelt sites were associated with reduced sexual reproduction, a potentially maladaptive response of leaf plasticity. Clonal and sexual reproduction generally showed no response to changes in snowmelt timing. Moreover, we found no home site advantage in terms of sexual and clonal reproduction. Leaf damage probability depended on snowmelt timing and thus exposure period, but it had no short-term effect on fitness traits.
In the neighbor removal study, the majority of S. herbacea traits were controlled primarily by snowmelt timing. However, neighbour removal directly reduced the number of days required for fruit production relative to control shrubs; however, it also increased the likelihood of leaf tissue herbivory. Effects of neighbour removal also changed along the environmental gradients, with neighbour removal leading to reduced leaf size on earlier snowmelt sites, and increased fungal damage with increasing elevation.
We conclude that Salix herbacea exhibits strong reductions in performance under earlier snowmelt, with few benefits of warming temperatures. Further, plasticity in leaf traits may lead to fitness reductions under earlier snowmelt. With accelerated spring snowmelt, environmental stress will likely increase due to increasing exposure to cold temperatures and damage agents, which reduce the leaf crops and flowering probability in the following summers. However, detrimental effects of early snowmelt may be at least partly mediated by facilitative interactions by neighbours. Under climate change, we may begin to see reductions in clonal and sexual reproduction, leading to fitness declines in S. herbacea, which in turn may reduce persistence of populations in arctic and alpine ecosystems, particularly those growing on current early exposure microhabitats as spring snowmelt accelerates. The results of this thesis suggest that earlier spring snowmelt timing may represent one of the critical global change drivers reducing dwarf shrub performance in arctic and alpine tundra ecosystems.