Martin Wilmking

The stability of global forest ecosystems is challenged by the accelerated pace of climate change, raising questions about their role as future carbon reservoirs. Intensifying drought events affect forest productivity and dieback at regional scales, yet how drought sensitivity and resilience components vary across the full distribution of widespread tree species remains insufficiently quantified. Here, we compile a pan-Eurasian network of 534 Pinus sylvestris tree-ring chronologies spanning the species’ full ecological range and quantify growth departures and resilience components during the most extreme summer droughts of the past century. Drought impacts diverge systematically along major climatic gradients. Growth reductions exceed 70% in cold-dry and warm semi-humid regions, while cool–humid forests show muted and even positive departures. Resistance is lowest under prolonged droughts in continental cold–dry interiors and under short summer droughts in semi-humid lowlands, while recovery capacity is highest where growth losses are most severe, reflecting local adaptations in drier regions. Despite coherent biome-scale patterns, substantial site-level variability indicates that growth legacies often persist beyond the event year. Nonlinear models show that resilience-climate relationships are shaped not only by mean climate conditions, but also by atmospheric water demand, site-level water balance, and thermal stability across climatic domains. Together, these results provide a range-wide assessment of drought impacts, demonstrating how prevailing climate conditions modulates forest stability and carbon sequestration capacity under intensifying climate extremes.

Increasingly frequent summer droughts are intensifying stem dehydration across European forests, constraining tree growth and resilience. Even mild water-related stress can inhibit tree growth, although its impact is species-specific. Tree water deficit (TWD), a measure deduced from radial stem contraction during drought, reflects stem water status and can be used as a proxy for the physiological drought stress experienced by trees. Using high-resolution dendrometer measurements representing 29 tree species from 130 forest sites across Europe from 2003 to 2018, we examined how the length of TWD periods during summer is related to their recurrence within the same year, to TWD intensity, to climate factors, and to wood type (i.e., tracheid-based coniferous, ring-porous or diffuse-porous anatomy). A greater degree of incomplete rehydration, indicated by highest minimum TWD during TWD periods (70% increase of expected response for conifers and ring-porous species, 42% increase of expected response for diffuse-porous), together with warmer temperature (44% increase of expected response for all wood types) and stronger drought, prolonged the length of TWD periods. However, the recurrence of TWD periods only increased the length of the subsequent TWD period within the same year for diffuse-porous tree species (12% increase of expected response). Our results show that wood type and its associated functional traits influence tree responses to recurring stem dehydration periods. The greater sensitivity of diffuse-porous species to recurring TWD periods suggests that wood type anatomy may play an important role in determining how tree species cope with increasingly frequent summer droughts.

Reconstructing masting, the synchronized large interannual variations in seed production by plant populations, is key to assessing tree species response to past climate variability and provides a baseline for predicting forest ecosystem fecundity under ongoing global warming. Since masting is tightly linked to growth through shifts in resource allocation, previous reconstruction attempts relied on tree‐ring width as a proxy of reproductive investment. However, this approach shows limitations, as narrow rings do not exclusively originate after masting events but may also result from periods of drought or high temperature (temperature anomalies) and other biotic or abiotic factors reducing radial growth. To disentangle the effect of masting events and temperature anomalies on wood anatomy, we investigated a wide range of wood anatomical traits along with tree‐ring width, using time series (1970–2021) from beech increment cores sampled in northeastern Germany. We used multivariate regression to capture traits affected by masting events and developed a random forest model to evaluate the predictive power of these traits compared with tree ring width alone. Results suggested that mesic site conditions prevented temperature anomalies from emerging as a major confounding factor for tree‐ring width, which was significantly reduced only after masting events. However, quantitative wood anatomy showed concurrent anatomical signals associated with both extreme temperatures and masting. Following masting events, the anatomical response revealed a complex mechanism of carbon reallocation towards reproductive effort, while reflecting a compensatory strategy to maintain hydraulic function under resource‐constrained conditions. Estimates of hydraulic conductivity, anatomical wood density, and lignin content emerged as key predictors of masting events, matching tree‐ring width alone in capturing the reproductive signal. Synthesis . Understanding past tree reproduction is critical for predicting future forest dynamics in response to climate change, but existing methods struggle to distinguish reproductive signals from other growth‐limiting factors. Our findings establish a novel link between wood anatomy and masting events beyond ring width alone, demonstrating that quantitative wood anatomical traits offer a more accurate and ecologically relevant approach for reconstructing past masting dynamics.

Climate is a primary, but non‐stationary, driver of tree growth. Climate change is altering the sensitivity of forest growth to water availability and temperature over time. It is considered that pedunculate oak ( Quercus robur L.) will cope with the changing climatic conditions in Europe in the near future. However, while species distribution models project expansion zones, they also identify reductions in occurrence at the dry and warm distribution margins. Whereas species distribution models primarily rely on occurrence data, tree rings—given their long‐term perspective and their use in empirical models—can provide a mechanistic view of forest growth dynamics, including temporally changing climate responses. Increased climate sensitivity and growth synchrony are key dendroecological indicators of tree stress. Here, we used an unprecedented network of 150 Q. robur sites (over 3300 trees), covering the full projected range of contracting to persistent areas across Europe, to assess the dendroecological indicators over recent decades in relation to species distribution model predictions. We reveal that oaks in areas projected to experience range contraction exhibited greater sensitivity to current growing season climatic conditions, whereas those in persistence areas responded more strongly to previous season conditions. Growth synchrony among trees was higher in the contraction areas, but showed no significant increasing trend over the last 70 years, as expected from ecotone theory. Temporal shifts in climate sensitivity were stronger for temperature and vapor pressure deficit in the persistence areas, whereas the climatic water balance gained importance in the contraction zones. These findings suggest that Q. robur growth is not yet being severely affected by climate change, and that the species is currently coping well with the climate changes, even in regions with projected range contractions, thereby challenging statistically derived scenarios of range shift based on species distribution models.

Aim: Intraindividual trait variability (iITV), which is the variability among repeated architectural units within an individual, may represent a crucial dimension of functional diversity in plant eco‐evolutionary dynamics. Although inter‐ and intraspecific trait variability have been widely studied, the extent of iITV remains largely overlooked. Since iITV might be especially relevant in long‐lived clonal plants, we investigated sources of trait variability (species, site, clone, ramet, leaf) in tundra clonal dwarf shrubs, particularly focusing on iITV and its potential drivers. Location: Europe. Time Period: July 2019 to September 2020. Major Taxa Studied: Clonal dwarf shrubs. Methods: We sampled four widespread boreo‐alpine clonal dwarf shrub species ( Dryas octopetala , Empetrum hermaphroditum , Vaccinium myrtillus and Vaccinium uliginosum ) along elevation gradients in the Pyrenees and along a latitude gradient in Europe. At each site, we selected four clones per species and sampled five ramets per clone. We measured size‐architectural traits in each ramet and leaf traits in five leaves per ramet. We quantified interspecific, intraspecific and iITV, investigated the relationship between iITV and both climate and clone structural variables, and compared sampling protocols either accounting for or ignoring iITV. Results: Although interspecific trait variability was substantial, we found large proportions of iITV within species. Size‐architectural traits showed larger iITV (up to 100% of ITV), but leaf traits also showed remarkable values (up to 77%). Our results showed that iITV increased with mean annual temperature for specific leaf area. However, climate and clone structural variables were not predictive for iITV for any other trait. Conclusions: Our quantification of iITV in clonal dwarf shrubs evidences the importance of this source of variability and its potential ecological implications and emphasises the need to integrate it in sampling protocols, both to avoid bias in comparative studies and improve the predictive capacity of trait‐based approaches in population and community ecology.

Key message: Provenances show a high phenotypic plasticity and the ability to grow beyond the cold treeline. Local is best can still be applied. Boreal forests situated in high latitudes face heightened susceptibility to climate extremes and global warming. Understanding the relative influence of adaptation mechanisms like phenotypic plasticity or local adaptation on key traits is crucial to better understand and project species distribution, forest growth and vitality. To address this, we conducted a reciprocal transplant experiment featuring two white spruce (Picea glauca [Moench] Voss) provenances in Alaska, representing cold and dry treelines. Trees from each provenance were reciprocally transplanted across a gradient spanning from dry bluff sites, dry treelines via old-growth forests to cold-limited treelines and beyond. From 2015 to 2022, we monitored survival, vitality, growth, and various needle morphology traits. Results showed that the dry provenance had a superior performance in its home environment. Whereas both provenances performed similarly at the cold site. Survival and vitality rates indicated that elevated temperatures favoured tree growth. Seedling survival and growth are possible beyond the current cold treeline. Further, needle morphology traits were more influenced by the current environment than by origin, thus showing a high phenotypic plasticity. Nevertheless, significant differences in needle morphology among provenances hinted at a genetic base of these traits. Results suggested that local is best can still be applied.

Key message: Climate-growth correlations are non-stationary among all size classes, and large trees are becoming sensitive to August and September drought conditions in the year preceding growth during the last decades. Understanding tree growth and forest dynamics under climate change is paramount to predict changes in carbon cycling, forest development, and ecosystem services. At temperature limited sites, such as alpine treelines, tree growth is often assumed to benefit from rising temperatures, while increased drought may offset potential benefits. Tree size is known to be related to climate sensitivity and drought induced mortality, with large trees generally suffering the most from drought. To assess the relationship of tree size and climate sensitivity for Norway spruce trees at treeline, we collected 158 tree cores at treeline and the adjacent closed canopy forest in the High Tatra Mountains in Slovakia. Size classes were established based on size class isolation of the total sample set, yielding artificial tree ring chronologies with a constant size over time. We ran moving-window correlation analyses to assess the temporal development of climate sensitivity. We found climate-growth correlations to be non-stationary and with similar trends among size classes. In general, trees are temperature limited during the growing season, but correlations have shifted from June to July in recent decades. Additionally, the largest trees show an increased and significant sensitivity to August and September drought conditions in the year preceding growth. These findings emphasize the increasing influence of drought constraints on tree growth, even at supposedly temperature limited treeline sites.

With ongoing global warming, increasing water deficits promote physiological stress on forest ecosystems with negative impacts on tree growth, vitality, and survival. How individual tree species will react to increased drought stress is therefore a key research question to address for carbon accounting and the development of climate change mitigation strategies. Recent tree‐ring studies have shown that trees at higher latitudes will benefit from warmer temperatures, yet this is likely highly species‐dependent and less well‐known for more temperate tree species. Using a unique pan‐European tree‐ring network of 26,430 European beech (Fagus sylvatica L.) trees from 2118 sites, we applied a linear mixed‐effects modeling framework to (i) explain variation in climate‐dependent growth and (ii) project growth for the near future (2021–2050) across the entire distribution of beech. We modeled the spatial pattern of radial growth responses to annually varying climate as a function of mean climate conditions (mean annual temperature, mean annual climatic water balance, and continentality). Over the calibration period (1952–2011), the model yielded high regional explanatory power (R² = 0.38–0.72). Considering a moderate climate change scenario (CMIP6 SSP2‐4.5), beech growth is projected to decrease in the future across most of its distribution range. In particular, projected growth decreases by 12%–18% (interquartile range) in northwestern Central Europe and by 11%–21% in the Mediterranean region. In contrast, climate‐driven growth increases are limited to around 13% of the current occurrence, where the historical mean annual temperature was below ~6°C. More specifically, the model predicts a 3%–24% growth increase in the high‐elevation clusters of the Alps and Carpathian Arc. Notably, we find little potential for future growth increases (−10 to +2%) at the poleward leading edge in southern Scandinavia. Because in this region beech growth is found to be primarily water‐limited, a northward shift in its distributional range will be constrained by water availability.