Understanding tree-response to extreme drought events is imperative for maintaining forest ecosystem services under climate change. While tree-ring derived secondary growth measurements are often used to estimate direct and lagging drought impacts, so-called drought legacies, underlying physiological responses remain difficult to constrain across species and site conditions. As extreme droughts may alter the functioning of plants in terms of resource allocation being shifted towards repair and physiological adjustments, climate control on growth may consequently be altered until physiological recovery is completed. In this context, we here advance the concept of drought legacy effects by quantifying “functional legacies” as deviations in climate sensitivity of secondary growth (CSD) after droughts, i.e. temporary alterations of climate-growth relations. We quantified climate sensitivity deviations after extreme drought events by applying linear mixed-effects models to a global-scale, multi-species tree-ring dataset and differentiated responses by clades, site aridity and hydraulic safety margins. We found that while direct secondary growth legacies were common across these groups, responses in post-drought climate sensitivity deviations were nuanced. Gymnosperms showed weaker coupling between secondary growth and the dominant climatic driver after droughts, a response that was narrowed down to gymnosperms with a small hydraulic safety margin, i.e. risky hydraulic strategy. In comparison, angiosperms instead showed stronger coupling between secondary growth and the dominant climatic driver following droughts, which was narrowed down to the angiosperms growing in arid sites. These results are consistent with current understanding of physiological impairment and carbon reallocation mechanisms, and the distinct functional responses suggest functional legacies quantified by climate sensitivity deviations is a promising avenue for detecting and thus studying physiological mechanisms underlying drought-responses in tree growth on large scales.

<p>In the course of climate change, forests around the globe will be exposed to more frequent and more severe extreme drought events. Direct and lagging impacts of drought events on forests, so-called drought legacies, are often estimated from tree-ring derived secondary growth measurements which easily translate into biomass and are available globally. However, secondary growth is a result from multiple internal mechanisms and therefore does not reveal potential impacts on the underlying physiology, such as hydraulic dysfunction, repair mechanisms and altered carbon allocation. Instead, the carbon demand of these impacts translates to less carbon being available for secondary growth which therefore results in temporarily altered relationships between climate and growth.</p> <p>Here, we advance the concept of drought legacies by additionally quantifying simultaneous “functional legacies” as climate sensitivity deviations (CSD) of secondary growth. We quantified legacies in both growth and climate sensitivity after extreme drought events using linear mixed-effects models on a global-scale, multi-species tree-ring dataset. We further differentiated the responses by clade, site aridity and hydraulic safety margins in order to determine common factors which convey heightened or lessened vulnerability to extreme drought events.</p> <p>Our results show that while depressed growth was common after droughts across most of the analysed groups, although with varying legacy lengths, post-drought climate sensitivity deviations were more nuanced. The climate sensitivity of growth was decoupled for gymnosperms with small hydraulic safety margin, i.e. those with a more risk-prone hydraulic growth strategy. In comparison, the climate sensitivity of growth tightened for angiosperms growing in arid sites, a response which occurred in conjunction with a post-drought growth overshoot. These responses are consistent with current understanding of impaired hydraulic function and increased carbon allocation towards xylogenesis, respectively. In conclusion, climate sensitivity deviations reveal physiological responses not discernible from growth legacies alone and therefore serves as a promising avenue for a more comprehensive identification of drought impacts on tree growth at large scales.</p>

<p>Hotter droughts will have an increasingly influential role in shaping forest ecosystems in the future. Risks include decreases in species richness, altered species distributions, forest dieback and changed function as carbon sink. A common method to study the impacts of droughts on forests is the quantification of reductions in biomass productivity via secondary growth – approximated by ring-width measurements –, including duration until growth rates return to pre-drought levels, so-called legacy periods. However, while these metrics are practical and relatively easy to measure, the underlying governing mechanisms are not, and thus poorly understood. Consequently, it is uncertain if drought-induced reductions in secondary growth are due to corresponding decreases in total physiological function or high plasticity, and if recovery times are due to lasting damage or adaptation with more carbon allocated to drought-mitigating structures.</p><p>The principle of the most limiting factor for tree-growth can be used to track temporal variations in climate-growth relationships. Similarly, the considerable strain hotter drought constitutes for tree-growth, and the need to repair damaged structures or alter carbon allocation, may imply temporary climate sensitivity deviations during legacy periods. Identifying their existence and quantifying subsequent differences in these deviations can help to shed light on strategies used by trees to respond to droughts.</p><p>Here, we detect and quantify deviations in climate-growth relationships during hotter drought legacy periods and assess how they differ according to clade (angiosperms – gymnosperms), site aridity and hydraulic safety margin. We do this by applying a linear mixed model on all ring-width indices (RWI) in the global-scale International Tree-Ring Data Bank (ITRDB) which exhibit a positive correlation with Standardized Precipitation-Evapotranspiration Index (SPEI). We apply a combined climatological and ecological definition for drought events and use site-dependent SPEI time-scales to allow for specific climate dependencies.</p><p>Results show heterogeneous post-drought climate sensitivity deviations, which are broadly categorized in three groups: 1) angiosperms growing in arid sites become increasingly sensitive to climate for 2 – 4 years; 2) angiosperms in mesic sites and or with high hydraulic safety margin show abrupt and complete disruption of the climate-growth relationship for the first year after droughts, which turn into a decrease in climate sensitivity for an additional 1 – 3 years; 3) gymnosperms in arid sites become less sensitive to climate for 2 – 4 years, although without the abrupt disruption seen in group 2. We discuss these results and their implications in an ecophysiological context, including future research avenues.</p><p>In conclusion, the results clearly show a functional legacy effect that is not detected through measurements of reductions in biomass accumulation alone, hinting at differential strategies employed by trees to cope with hotter droughts. This is a first step towards a better understanding of the mechanisms underlying hotter drought legacies which may help to improve ecosystem models and better predict how trees will respond to drought in a warming future climate.</p>

<p>Extreme drought events will have an increasing influence on forest ecosystems and services in the course of climate change. It is thus pivotal to understand their direct impacts and subsequent recovery patterns. Drought impacts are often reported as changes in net primary productivity (NPP) related to impact severity. However, because NPP integrates over all physiological processes, variability in NPP alone cannot explain internal mechanisms.</p><p>Due to temporally variable growing conditions, climate-growth relationships are naturally non-stationary on longer time scale. On shorter time scales, extreme drought events are considerable perturbations that likewise alter the climate sensitivity of growth, corresponding to physiological impacts caused by drought. Therefore, post-drought changes in the climate sensitivity of growth serves as a potential avenue of studying physiological impacts. Decoupled climate-growth relationships would be expected in the case of damage on the hydraulic system or reallocation of carbon to rebuild foliage. Conversely, tightened coupling would be expected in the case of stricter growing conditions as per the law of the minimum or carbon reallocation towards increased xylogenesis. </p><p>Because experimental ecophysiological studies are labour and cost intensive, they are typically limited in space and time. Meanwhile, climate-growth relationships derived from tree-ring widths (as an approximation for variability in NPP) and high-resolution climate products are easily accessible on regional to global scales. By finding common post-drought responses in climate sensitivity of tree-growth for trees grouped by abiotic and biotic factors it is possible to analyse physiological impacts on large scales, thus effectively enhancing our understanding of the underlying mechanisms that result in quantified impacts on NPP.</p><p>Here, we aim to find intraspecific differences in post-drought climate-growth relations for the ecologically and economically important European tree species European beech (<em>Fagus sylvatica</em> L.). Using a European-wide dataset of tree-ring widths, the European Beech Tree-Ring Network (EBTRN), we compute post-drought changes in climate-growth relationships – climate sensitivity deviations – in addition to direct and lagging impacts on absolute tree-ring derived growth. Preliminary analyses indicate a complex connection between growth recovery rates and diverging post-drought climate sensitivity deviations, in turn shaped by growing condition factors.</p>