Drought-sensitivity ranking of deciduous tree species based on thermal imaging of forest canopies

https://doi.org/10.1016/j.agrformet.2011.06.019Get rights and content

Abstract

Most climate change projections for Central Europe predict higher mean summer temperatures and prolonged summer drought periods. However, in diverse mixed forest stands we expect tree species specific responses to water shortage, as tree species are highly variable in rooting depth and physiological traits related to the water balance. Here, we assessed the drought sensitivity of the water relations of six deciduous forest tree species at four sites with contrasting water availability by airborne thermal imagery of canopy foliage temperature, sap flow and soil water potential. Canopy architecture had a consistent influence on canopy foliage temperature with ‘dense canopy’ species (Acer pseudoplatanus, Fagus sylvatica and Tilia platyphyllos) being warmer (0.5–1.5 K) than ‘open canopy’ species (Fraxinus excelsior, Prunus avium and Quercus petraea). While the canopy foliage was close to air temperature at the beginning of the drought period (ΔTC−A = −0.1 to 0.7 K) it strongly warmed up with ongoing drought, especially at the two ‘dry’ sites with a ΔTC−A of 3.5–5 K. The pronounced canopy foliage warming at the ‘dry’ sites after 22 days of drought was associated with reduced transpiration rates as sap flow was curtailed by 20–35% in all species except F. excelsior and Q. petraea. Based on canopy foliage temperature and sap flow data, we considered A. pseudoplatanus the most drought sensitive species followed by F. sylvatica, T. platyphyllos and P. avium and the two ring-porous species F. excelsior and Q. petraea being clearly the least sensitive ones. At drier sites, increasing summer temperatures and drought might change the competitive abilities of tree species in favour of those that are able to maintain transpirational fluxes and cooler canopies such as F. excelsior and Q. petraea.

Highlights

► We show infrared thermometry and sap flow data of six deciduous tree species. ► Species with dense canopies were considerably warmer especially at dry sites. ► Under drought, diffuse-porous-wood species show a strong reduction in sap flow. ► The sap flow of the two ring-porous species was not affected by prolonged drought. ► Based on the data we ranked the species according to their drought sensitivity.

Introduction

Plant species composition of communities and productivity strongly depend on water availability anywhere on the globe (Schulze et al., 1996, Kreft and Jetz, 2007). European temperate forests are composed of species that differ in their moisture requirements (Ellenberg, 1974). As most climate change projections for Central Europe predict lower summer precipitation with prolonged drought periods (Christensen et al., 2007) water availability will become critical for some taxa. Regional projections for Switzerland indicate a 20% decrease in summer precipitation and 4 K higher mean summer temperatures in 2070 (Frei et al., 2006, Meehl et al., 2007). The higher summer temperatures in combination with high atmospheric vapour presser deficits (vpd) will lead to higher evaporative demand and therefore will exert higher water demand and, thus, a greater likelihood of water shortage. Forest tree species are highly variable in rooting depth, canopy architecture and physiological traits related to the water balance (Canadell et al., 1996, Dawson, 1996, Jackson et al., 1996, Schulze et al., 1996, Schenk and Jackson, 2002, Körner, 2005). The species specific responses to the expected climate change might therefore alter the competition among species (Leuzinger et al., 2005, Zweifel et al., 2009) and lead to new community compositions (Gessler et al., 2007, van Mantgem and Stephenson, 2007). As most forests in Central Europe are actively managed (timber production) these changes are not only biologically important (Adams et al., 2009, Allen et al., 2010) but will also have socio-economic consequences (Millennium Ecosystem Assessment, 2005). Even if the species composition might stay the same, species specific changes in fitness directly influence the timber harvest. As the seedlings of today will be the forest of 2100, information about tree responses to climate change are essential for both biological and economic (forestry) planning.

Most studies dealing with drought effects on plant species communities either work with single extreme events such as for example the hot and dry summer 2003 (e.g., Weaver et al., 1935, Bollinger et al., 1991, Stämpfli and Zeiter, 2004, Czajkowski et al., 2005, Leuzinger et al., 2005, Archaux and Wolters, 2006, Low et al., 2006, Reichstein et al., 2007), experimentally with rainout shelters (e.g., Sternberg et al., 1999, Grime et al., 2000, Köchy and Wilson, 2004, Morecroft et al., 2004, Gilgen and Buchmann, 2009) or throughfall exclusion experiments (e.g., da Costa et al., 2010, Markewitz et al., 2010). While rainout shelters allow exerting different controlled conditions over several years they are only practicable for low stature vegetation such as grasslands, dwarf shrubs and small areas. Studies working on mature forest stands with 50–100 year old trees and 30 m canopy height, as the one presented here, are mostly limited to observations during drought periods or confined to relatively small areas (given the ca. 100 m2 of a single tree crown). Broad airborne screening offers the replication and spatial coverage needed to allow conclusion on an ecosystem level.

A non invasive method to gain information about the water status of a plant is thermal imaging of leaves or canopy foliage as a proxy for the energy balance and, thus, for concurrent transpiration (Fuchs, 1990, Jones, 1999a, Jones and Leinonen, 2003, Leinonen et al., 2006). Canopy foliage temperature is a resultant of the energy balance (net radiation  latent heat flux  sensible heat flux) at leaf level and is dominated by plant architecture (canopy shape, canopy density, leaf size, boundary layer resistance, etc.), environmental conditions (incoming radiative energy, air temperature, wind) and plant-controlled transpiration (Monteith and Unsworth, 1990). Adequate information on these essential parameters is imperative to evaluate the actual water status of a tree, especially when ‘dry’ and ‘wet’ reference surfaces are missing (Jones et al., 2002). However, as the architecture of an individual tree only changes marginally during a few weeks of drought in mid-summer, relative comparisons (in time) are possible. On bright, warm days the canopy foliage temperature to air temperature difference (ΔTC−A) is mainly the result of transpiration (Fuchs, 1990, Jones, 1999b). By revisiting selected trees or forest sites under similar weather conditions (incoming radiative energy, wind) the ΔTC−A yields direct estimates of the relative changes in transpiration. To ensure that the changes in ΔTC−A are connected to changes in transpiration, and therefore water status, direct measurements such as stomatal conductance or sap flow are essential (Jones and Leinonen, 2003, Leuzinger et al., 2005, Leuzinger and Körner, 2007b). Unfortunately, these direct methods are expensive or too time consuming to cover many trees, whereas thermal imagery allows to collect huge amounts of data in a short time over very large areas. There remains an elusive problem: trees that are highly sensitive to vpd reduce their transpiration by diminishing stomatal conductance even under favourable soil water conditions. Hence, neither flux nor canopy temperature can reveal with certainty whether a tree suffers from water shortage (Körner, 1985). In the current study, we can identify such responses during the earliest part of the hot summer period when soils were still wet, but vpd quite high. A species performing such pronounced vpd response should show a flux reduction and canopy warming relative to others at this early stage.

In this study we combined airborne thermal imagery of canopy foliage temperature with ground based measurements of sap flow and soil water potential in diverse mixed forest patches with contrasting water availability, (1) to analyse species specific differences in canopy foliage temperature due to leaf and canopy architecture, (2) to establish and compare a ‘ranking’ of tree species according to their canopy foliage temperature at sites and periods with contrasting water availability, and (3) to explore the relevance of thermal image data with regards to tree-specific actual sap flow and soil moisture.

Section snippets

Study sites and species

In total, we had two different temperate-lowland test regions, namely: Hofstetten-Flüh (47°28′N, 7°30′E; elevation 550 m a.s.l.) and Münchenstein (47°30′N, 7°37′E; elevation 350 m a.s.l.) both in the NW part of Switzerland. Within each test region we searched for diverse mixed forest stands on flat terrain (no exposition) that were at least 50 years old and had a closed canopy. Out of these potential study sites we then chose for each test region one site we expected to dry out fast during a

Weather conditions and soil water potential

In 2010, the average air temperature in July was 20.9 °C, which is 1.8 K higher than the long term average, indicating a warm and dry July with no rain fall from 24 June to 22 July at the study sites in Münchenstein and only a small rain event (10.6 mm on 4 July) at the study sites in Hofstetten-Flüh (Fig. 2). During our three thermal imaging campaigns the concurrent air temperatures were ca. 30.5, 31 and 31 °C. The combination of high air temperatures and dry weather conditions led to high vpd

Species specific canopy foliage temperatures

Our results showed constant differences among tree species, with A. pseudoplatanus, F. sylvatica and T. platyphyllos having higher relative canopy foliage temperatures than F. excelsior, P. avium and Q. petraea. As these differences in ΔTC−A were consistent across all sites and thermal imaging campaigns, canopy architecture and leaf morphology are the most likely explanations. The findings that ‘dense canopies’ (A. pseudoplatanus, F. sylvatica, T. platyphyllos) are warmer than ‘open canopies’ (

Conclusion

Our process based analysis of the drought sensitivity of six major European deciduous tree species revealed clear species and micro-site specific responses with F. excelsior and Q. petraea being the most drought tolerant species in terms of maintenance of transpirational flux followed by P. avium and F. sylvatica and the two most drought sensitive species T. platyphyllos and A. pseudoplatanus. Providing the first combined evidence for canopy responses, sap flux and soil moisture, this ranking

Acknowledgements

We thank Reinhard Bischoff, Jonas Meyer and Roman Zweifel for technical and scientific support and the forest authorities of Basel-Land and Solothurn for making the forest patches available, and for their collaboration in the field. This work was funded by the Swiss Science Foundation through NCCR Climate (P3.3 ECOWAT), and by the research program ‘Wald und Klimawandel’ (WKW) of the Federal office of the Environment (BAFU in cooperation with WSL).

References (61)

  • E.K. Bollinger et al.

    Effects of seasonal drought on old-field plant-communities

    Am. Midl. Nat.

    (1991)
  • J. Canadell et al.

    Maximum rooting depth of vegetation types at the global scale

    Oecologia

    (1996)
  • J.H. Christensen et al.

    Regional climate projections

  • K.L. Clawson et al.

    Infrared thermometry for scheduling irrigation of corn

    Agron. J.

    (1982)
  • E.R. Cook et al.

    Methods of Dendrochronology: Applications in the Environmental Sciences

    (1990)
  • T. Czajkowski et al.

    Impact of the 2003 summer drought on growth of beech sapling natural regeneration (Fagus sylvatica L.) in north-eastern Central Europe

    Allg. Forst Jagdztg.

    (2005)
  • A.C.L. da Costa et al.

    Effect of 7 yr of experimental drought on vegetation dynamics and biomass storage of an eastern Amazonian rainforest

    New Phytol.

    (2010)
  • T.E. Dawson

    Determining water use by trees and forests from isotopic, energy balance and transpiration analyses: the roles of tree size and hydraulic lift

    Tree Physiol.

    (1996)
  • H. Ellenberg

    Zeigerwerte der Gefässpflanzen Mitteleuropas

    (1974)
  • H. Ellenberg

    Vegetation Mitteleuropas mit den Alpen in ökologischer, dynamischer und historischer Sicht

    (2010)
  • C. Frei et al.

    Future change of precipitation extremes in Europe: intercomparison of scenarios from regional climate models

    J. Geophys. Res. Atmos.

    (2006)
  • M. Fuchs

    Infrared measurement of canopy temperature and detection of plant water-stress

    Theor. Appl. Climatol.

    (1990)
  • A. Gessler et al.

    Potential risks for European beech (Fagus sylvatica L.) in a changing climate

    Trees Struct. Funct.

    (2007)
  • A.K. Gilgen et al.

    Response of temperate grasslands at different altitudes to simulated summer drought differed but scaled with annual precipitation

    Biogeosciences

    (2009)
  • A. Granier

    A new method of sap flow measurement in tree stems

    Ann. Sci. Forest.

    (1985)
  • J.P. Grime et al.

    The response of two contrasting limestone grasslands to simulated climate change

    Science

    (2000)
  • R.B. Jackson et al.

    A global analysis of root distributions for terrestrial biomes

    Oecologia

    (1996)
  • H.G. Jones

    Plants and Microclimate

    (1992)
  • H.G. Jones

    Use of thermography for quantitative studies of spatial and temporal variation of stomatal conductance over leaf surfaces

    Plant Cell Environ.

    (1999)
  • H.G. Jones et al.

    Thermal imaging for the study of plant water relations

    J. Agric. Meteorol.

    (2003)
  • Cited by (121)

    • Applicability of the crop water stress index based on canopy–air temperature differences for monitoring water status in a cork oak plantation, northern China

      2022, Agricultural and Forest Meteorology
      Citation Excerpt :

      Xue et al. (2004) reported similar results for wheat under high soil-water-potential conditions. Scherrer et al. (2011) and Romero-Trigueros et al. (2019) believed that an increase in VPD caused stomatal closure and Tr to decrease even under well-watered conditions, resulting in a positive correlation between dTll and VPD in trees. However, this explanation is not suitable for our results on NWSB failure establishment, as dTll was positively correlated with both VPD and Tr during the daytime, around noon, and in the afternoon (Fig. 4b–d).

    View all citing articles on Scopus
    View full text