Trees evolved strategies in persevering moderate drought episodes through physiological and morphological adaptations which aid in balancing cooling mechanisms in the crown while preventing excessive water loss and carbon starvation. Such adaptions involve the regulation of stomatal conductance, smaller leaf surface area and solar tracking (Mocko et al., 2017; Nobel, 2020), as well as wood traits which enable xylem to be resilient to hydraulic failure (Choat et al., 2012). Drought periods, even in a moderate form, increase the possibility for mortality (van Mantgem & Stephenson, 2007) and account for at least a reduction in growth (Brzostek et al., 2014; Nathalie et al., 2006) even in the absence of climate change (Allen et al., 2010). Recent extreme drought occurrences such as experienced in Europe in 2018 and 2019 (Leuschner, 2020; Senf & Seidl, 2021) due to climate change induced warming and shifts in precipitation patterns (Moran et al., 2017) are of raising concern in terms of amplified tree mortality and die-off regardless of climate zone (Schuldt et al., 2020).
European Beech (Fagus sylvatica), due to its extremely high shade tolerance, can under optimal conditions outcompete every other tree species in many parts of Europe (Kutsch et al., 2009). In recent years however, beech has been declining in growth throughout Europe (Delaporte et al., 2016; Martinez del Castillo et al., 2022; Tomasella et al., 2018; van der Werf et al., 2007) yet could gradually acclimatize to drought over time (Pretzsch et al., 2020).
Typical drought episodes can result in a reduction in carbon uptake due to a decrease in stomatal conductance, premature leaf senescence (Hartmann et al., 2021; Raspe et al., 2004) as well as a decrease in foliage the following year due to a reduction of available buds (Roloff, 1988). Prolonged drought episodes, however, can cause irreversible damage especially for anisohydric plant species in terms of xylem embolism (Tomasella et al., 2018) where permanent damage occurs to the hydraulic system (Garcia-Forner et al., 2017; Hartmann, 2011).
An approach in assessing drought-tolerant species is through the classification of a trees’ hydraulic strategy in terms of the anisohydric and isohydric spectrum (Berger-Landefeldt, 1936; Hartmann et al., 2021). Despite numerous studies in the hydric behaviour of trees, there is no mathematical model describing this plant trait and is typically categorized in reference to the relationship of stomatal conductance gs and leaf water potential Ѱ1 (Klein, 2014). The strategy of isohydric plants is typically known for a reduction in transpiration with closing of the stomata during water shortage with the consequence of reduced CO2 assimilation (Sade et al., 2012). On the other hand, risk-taking anisohydric plants leave the stomata open for longer periods despite water shortage making them more vulnerable to hydraulic failure yet maintain a higher CO2 uptake during drought episodes (Burkhardt & Pariyar, 2015; Leuschner, 2020). The anisohydric strategy would in effect require more water to keep leaves cool during extreme heat and rely on substantial fluctuations of tree stem (xylem) water content with a reliance on nocturnal refilling (Yi et al., 2017). Within each species, variations in hydric behaviour can also occur depending on genetic variation in terms of drought stress tolerance (Leuschner, 2020; Moran et al., 2017) where species with high phenotypic plasticity could allow for individuals to acclimatise to changing climate conditions (Tomasella et al., 2018). Categorizing tree species and even provenances (Moran et al., 2017) into hydric behavioural classes by means of the quantification of stomatal conductance with gas exchange measurements and leaf temperature could aid in the assessment of drought stress tolerance to climate change. Such a classification should however, not be assumed, and a comprehensive holistic approach (Leuschner, 2020) in terms of whole-tree carbon balance is recommended (Garcia-Forner et al., 2017). Practically speaking however, central European species are rarely either aniso- or isohydric, but rather estimated in reference to other species. For example, Quercus species would be typically more anisohydric than Fagus, and Pinus would be more isohydric than Fagus. A better understanding of hydric behaviour among species and even individuals mapped at the regional scale could aid in focussing forest management goals.
An increased awareness of the effects of drought on tree productivity and survival (Pretzsch et al., 2020) is required if we are to select appropriate species and provenances for the purpose of improving silvicultural practices (Bolte et al., 2009) in terms of drought stress adaptation. The effects of prolonged and extreme drought conditions on forests in the future is however relatively unknown, and the ability for trees to acclimate is underestimated (Lapenis et al., 2005; Pretzsch et al., 2020; Reich et al., 2016). Pretzsch et al. (2020) showed that over a 5-year experiment of induced drought, beech acclimated faster than spruce, while spruce acclimatised faster when mixed with beech. This suggests that some species could acclimatise to extended drought stress over time within a generation providing hydraulic failure does not occur. The detection of hydric behaviour could aid in determining tree species mixing strategies as in which species co-exist well during drought as well as an assessment of hydric variability within species.
The use of thermal infrared (TIR) sensors have proven useful for non-destructive water content retrieval and stomata closure detection in plants (Cohen et al., 2005; Feller, 2016; Grant et al., 2006). Recent developments in UAV (Unmanned Aerial Vehicle)-mounted sensors provide the opportunity to acquire thermal imagery from above the crop or forest canopy. Gómez-Candón et al. (2016) used UAV-based thermal imagery to detect higher canopy temperatures in non-irrigated trees while using reference ground targets for temperature accuracy validation. For the purpose of mapping surface energy and water fluxes, Simpson et al. (2022) implemented UAS thermography and multispectral data to produce evapotranspiration maps for oak trees. The accuracy of low-cost thermal imagers can however prove challenging and various studies have been carried out to assess and develop thermal imaging acquisition methods (Acorsi et al., 2020; Aragon et al., 2020; Kelly et al., 2019; Perich et al., 2020; Ribeiro-Gomes et al., 2017; Smigaj et al., 2017; Wan et al., 2021; Zakrzewska et al., 2022). Challenges affecting thermal imaging accuracy can arise from the influence of meteorological variables such as air temperature (AT), relative humidity (RH), solar radiation (SR), and wind speed (WS) which will not only affect the tree canopy temperature but also influence the sensor itself. Furthermore, other issues exist which can affect thermal sensors such as sensor drift, internal calibration, “bad pixels”, “spot-size”, leaf angle due to solar tracking and the exclusion of relevant pixels through masking.
The aim of this study is to explore the possibility to acquire accurate thermal imagery at the individual tree level as would be of interest for intensive forest monitoring plots (i.e. ICP forests Level II). Here we developed a single shot method with the Micasense Altum sensor for the acquisition of tree crown temperature for the purpose of calculating the Leaf-to-air vapor press deficit (LVPD) and modelling TWD. Additionally, we test various acquisition and pixel extraction methods with the aim of minimizing error propagation. Using indoor and outdoor experiments we assess the dispersion of acquired thermal data while implemented leaf temperature sensors for validation. The specific aims of the study are as follows:
-
Determine the minimum number of pixels required to obtain accurate temperature and whether dry vegetation with lower emissivity will affect accuracy.
-
Test the accuracy of tree crown temperature to upper canopy mounted leaf temperature sensors over repeated missions under varying weather conditions throughout the growth season.
-
Evaluate the dispersion of TIR values acquired with grid type against single shot acquisition methods.
-
Assess the possibility to model the TWD with TIR tree crown data and meteorological data.