Download Manuscript - CSIRO Research Publications Repository

NOTICE: This is the author’s version of a work that was accepted for publication in Nature. Changes
resulting from the publishing process, such as peer review, editing, corrections or structural
formatting may not be reflected in this document. A definitive version was subsequently published in
NATURE, VOL 491, (29 NOVEMBER 2012). DOI: http://dx.doi.org/10.1038/nature11688
Title: Global convergence in the vulnerability of forests to drought
Choat B, Jansen S, Brodribb TJ, Cochard H, Bhaskar R, Bucci S, Delzon S, Feild TS,
Gleason SM, Jacobsen AL, Lens F, Maherali H, Martinez-Vilalta J, Mayr S, Mencuccini M,
Mitchell PJ, Nardini A, Pittermann J, Pratt RB, Sperry JS, Westoby M, Wright IJ, Zanne A.
Summary
Shifts in rainfall patterns and increasing temperatures associated with climate change
are likely to cause widespread mortality of forest plants in regions where the duration
and severity of droughts increase1. One primary cause of drought-induced mortality is
hydraulic failure of the plant water transport system2-4. Water stress creates trapped
gas emboli in this transport system, which reduces the ability of plants to supply water
to leaves for photosynthetic gas exchange and can ultimately result in desiccation and
death. However, at present we lack a clear picture of how thresholds to hydraulic
failure vary across a broad range of species and forest environments. Using a new data
synthesis of woody plants (478 species from 185 sites), we show that the majority of
forest species operate with narrow hydraulic safety margins against injurious levels of
water stress and therefore face a high risk of mortality if significant shifts in rainfall
accompany increasing temperatures. Safety margins were largely independent of mean
annual precipitation, with many species highly vulnerable to hydraulic failure
regardless of their current rainfall environment. These findings provide insight into
why climate induced mortality is occurring not only in arid regions but also in mesic
forests not normally considered to be at risk. This is particularly relevant to projections
of mortality in mesic forest species, which have constitutively low embolism resistance
and are predicted to face reduced rainfall in many areas5.
1
Text
Sensitivity to drought is of fundamental importance in shaping the distribution of plant
species, community structure, and biogeography6-8. Drought, often in combination with other
abiotic and biotic factors, underpinned many large scale forest mortality events over the past
century1,9. Recent evidence suggests that rising global temperatures are already amplifying
drought-induced forest dieback and affecting terrestrial net primary productivity10-14. The
consequences of this mortality are potentially dramatic. For example, rapid forest collapse via
drought could convert the world’s tropical forests from a net carbon sink into a large carbon
source during this century5,15. Predicting how forests will respond to future climate change
hinges on a quantitative understanding of the physiological mechanisms governing droughtinduced mortality at the species level. The most promising avenue for characterizing the
sensitivity of plants to water stress is by quantifying the strength of the liquid (hydraulic)
connection between soil and leaves through the water-transporting xylem tissue.
Cavitation, a phase change from liquid water to vapour, occurs in plants because sap
transported through the xylem is held under negative pressure16. The resultant air emboli
block xylem conduits and reduce the ability of the plant to move water from the soil to the
sites of photosynthesis17. Recent evidence indicates that the ability of woody plants to survive
and recover from periods of sustained drought is strongly related to their embolism
resistance2,3. This property varies widely among species and is largely determined by
differences in the structure of the xylem18. Though xylem structure can acclimate to
environmental variation during growth and development, subsequent acclimation of
embolism resistance to environmental stress is not possible because xylem conduits are dead
at maturity. Embolism resistance therefore represents a critically important trait in defining
the limits of drought tolerance across woody species and for predictions of drought-induced
mortality at regional and global scales.
2
The resistance of a plant to embolism is described by the relationship between xylem
pressure and the loss of hydraulic conductivity due to conduit occlusion by gas emboli
(Supplementary Fig. 1). The xylem pressure at which a 50% loss of conductivity occurs (Ψ50)
is the most commonly used index of embolism resistance. In order to examine the
vulnerability of forest biomes to drought-induced hydraulic failure we assembled a database
of Ψ50 containing 478 woody species. In these analyses we broadly defined forests to include
Mediterranean and savanna woodland biomes. Site climate varied widely, e.g. mean annual
precipitation (MAP) ranged from 300 to 4500 mm, and mean annual temperature from -4º to
27 ºC. Data for angiosperms (flowering plants) and gymnosperms (mainly conifers) were
analysed separately because of fundamental differences in xylem structure between these two
groups.
In drying soil, stomata initially regulate water loss from the leaves to maintain xylem
pressure within a range that will protect the plant from extensive embolism and hydraulic
failure in the xylem17,19. As drought continues, stomatal closure slows but does not halt the
decline of xylem pressure and hydraulic capacity. If soil water is not replenished before
complete hydraulic failure occurs then the plant will desiccate and die. It is therefore
expected that the minimum xylem pressure observed in plants under natural conditions (Ψmin)
will be strongly related to Ψ50. We observed a significant linear relationship between Ψmin
and Ψ50 for both angiosperms and gymnosperms -- demonstrating that embolism resistance is
tightly linked with the level of water stress experienced by plants across a broad range of
environments (Fig. 1). The difference between Ψmin and Ψ50 represents the ‘safety margin’
with which the plant operates in a given environment20,21. A Ψmin equal to Ψ50 indicates an
impaired vascular system that limits photosynthesis and increases the risk of catastrophic
hydraulic failure.
3
Across all forest biomes, angiosperm species operated at narrow safety margins (Fig.
2). The low safety margin of many species indicates that all forest biomes would be
vulnerable to loss of taxa via drought induced mortality, particularly in light of predicted
increases in extreme drought events associated with global climate change22,23. This finding
provides insight into why climate-induced mortality is occurring not only in arid regions but
also in mesic forests not normally considered to be at risk1,5,24. Species from mesic forests are
particularly vulnerable because of their constitutively low embolism resistance compared to
species from drier environments. In addition, the slope of the relationship between xylem
pressure and the loss of hydraulic conductivity is steeper for mesic taxa, meaning that a
smaller absolute increase in water stress is required to push these plants to hydraulic failure
(Supplementary Fig. 2). While angiosperms have a well documented capacity to recover from
high levels of embolism by refilling xylem vessels, it is clear that this recovery can only
occur if periods of drought are followed by sufficient precipitation and a return to favourable
water status25. Therefore, refilling does not represent an effective strategy for mitigating the
effects of severe and persistent drought.
Gymnosperms maintained greater safety margins between Ψmin and Ψ50 than
angiosperms across woodland and temperate forest biomes, with safety margins increasing in
species with higher embolism resistance. However, this does not indicate that gymnosperms
are generally immune to the possibility of hydraulic failure. In fact, species in the Pinaceae
have significantly lower embolism resistance and safety margins than species in the
Cupressaceae26,27. This taxonomic difference is reflected in the greater frequency of dieback
events involving Pinaceae species compared with Cupressaceae in Asia, Europe, and North
America1,28. A clear illustration of how species differences in embolism resistance and safety
margin play out during drought is provided by the Piñon-Juniper woodlands of the south
western USA. In the severe 2002-2003 drought, Pinus (Pinaceae) species suffered
4
widespread mortality while co-existing Juniperus (Cupressaceae) species with greater
embolism resistance and safety margins survived12.
In our dataset, Ψ50 was significantly related to MAP such that, on average, species at
higher rainfall sites exhibited lower resistance to embolism (Fig. 3). Similar relationships
were found between Ψ50 and climate variables that account for both the variation in potential
evapotranspiration (PET) and seasonality of precipitation: Aridity Index (MAP / PET) and
mean precipitation of the driest quarter (Supplementary Fig. 3). That said, a wide range of
hydraulic strategies occur within any given climate region, with the greatest variation in Ψ50
occurring at sites with MAP between 300-1000 mm. These sites are characterised by high
seasonality of rainfall and include savannas, seasonally dry tropical forests and
Mediterranean environments. In high MAP sites, represented by tropical rainforests in our
dataset, variation is compressed to less negative Ψ50. This suggests that the lower transport
efficiency and higher structural ‘costs’ associated with achieving high embolism resistance
render these dry forest species less competitive in highly productive, wet tropical
environments18.
It is clear that Ψmin and MAP are decoupled in certain cases, implying that some species
growing in drier environments escape from water stress, therefore alleviating the need for
high embolism resistance. There are numerous examples in the literature of species with low
embolism resistance growing in areas of low rainfall or which are prone to seasonal drought.
This includes riparian and ground water dependent vegetation occurring in drier regions8 and
drought deciduous behaviour in dry tropical forest trees29. These plants manage to avoid very
negative xylem pressures by some combination of predictable access to ground water (deep
rooting phreatophytes), reduction in leaf area, and water storage in stems. However, it
appears that the majority of species operate close to their functional limits regardless of these
adjustments, a finding that is largely independent of rainfall environment and biome.
5
Our analyses of Ψmin and safety margin demonstrate the fundamental vulnerability of
woody plant species to reduction in rainfall and increasing temperatures. The risky hydraulic
strategy exhibited by many species is most likely the result of a trade-off that balances
growth against risk of mortality in a given environment17,19. However, the limited genetic
diversity of embolism resistance within species suggests that it is unlikely that most species
will be able to adapt hydraulically to accelerated climate change30. If the tight link between
embolism resistance and water availability is the product of an evolutionary response to
natural selection over many generations, then the rapid pace of climate change may outstrip
the capacity of populations to evolve increased embolism resistance. Such limits to
adaptation could result in changes to the composition of forest and woodland communities,
with potentially large loss of biodiversity in many areas. While it is evident that a number of
different mechanisms (hydraulic failure, carbohydrate depletion, insect attack) are involved in
drought induced mortality, these mechanisms are highly interdependent9. Embolism
formation is a key mechanism of mortality because it sets the thresholds for stomatal closure,
leading to limitations on photosynthesis, increased heat and light damage, and a rundown of
carbohydrate reserves over longer time scales. Our results indicate that knowledge of
embolism resistance is essential to predict how community structure and the distribution of
species will change in response to increased aridity. Incorporating embolism resistance and
safety margins into dynamic global vegetation models is not only desirable, but also now
within reach, using this broad hydraulic dataset.
Methods
Xylem traits were compiled from published and unpublished sources. The full dataset used
for these analyses contained 478 species from 185 sites, with 384 angiosperm and 94
6
gymnosperm species. Data presented for Ψ50 are from stem xylem only and therefore
represent a conservative estimate of embolism resistance since roots are normally less
embolism resistant than stems within a species. Ψmin data are the minimum water potential
recorded for each species in the field, indicating a seasonal, rather than daily minimum value.
‘Lethal’ Ψmin values reported in manipulative trials where plants were droughted to death
were not included in the database. Safety margins were calculated as the margin by which
Ψmin differs from the xylem pressure causing 50% loss of hydraulic conductivity (Ψmin – Ψ50).
Data presented in Fig. 1 were binned by 1.0 MPa increments of Ψ50; bins with only one data
point were pooled with the next highest bin. Regression lines presented are for raw data to
avoid bias associated with uneven bin size. Differences in biome means for safety margin
were analysed using a general linear model, with differences between angiosperms and
gymnosperms considered separately. Climate data were taken from (a) the paper in which Ψ50
data were published or (b) the WorldClim or CRU Climate Database depending on which
value matched better with elevation. Relationships between Ψ50 and climate variables (MAP,
AI, mean precipitation of the driest quarter) were analysed using a generalised model
assessed by restricted maximum likelihood where variance was simultaneously modelled as a
power function of the same climate variables. Forest biomes were assigned based on site
descriptions contained in the primary sources. We defined ‘forest’ broadly to include
Mediterranean, savanna and woodland environments that are not commonly classified as
forests. The dataset therefore encompasses tree, shrub and liana species from vegetation
communities with a significant component of woody plants.
7
References
1
Allen, C. D. et al. A global overview of drought and heat-induced tree mortality
reveals emerging climate change risks for forests. Forest Ecology and Management
259, 660-684 (2010).
2
Brodribb, T. J. & Cochard, H. Hydraulic failure defines the recovery and point of
death in water-stressed conifers. Plant Physiol. 149, 575-584 (2009).
3
Kursar, T. A. et al. Tolerance to low leaf water status of tropical tree seedlings is
related to drought performance and distribution. Functional Ecology 23, 93-102
(2009).
4
McDowell, N. et al. Mechanisms of plant survival and mortality during drought: Why
do some plants survive while others succumb to drought? New Phytol. 178, 719-739
(2008).
5
Phillips, O. L. et al. Drought sensitivity of the amazon rainforest. Science 323, 13441347 (2009).
6
Brodribb, T. & Hill, R. S. The importance of xylem constraints in the distribution of
conifer species. New Phytol. 143, 365-372 (1999).
7
Engelbrecht, B. M. J. et al. Drought sensitivity shapes species distribution patterns in
tropical forests. Nature 447, 80-82 (2007).
8
Pockman, W. T. & Sperry, J. S. Vulnerability to xylem cavitation and the distribution
of Sonoran desert vegetation. Am. J. Bot. 87, 1287-1299 (2000).
9
McDowell, N. G. et al. The interdependence of mechanisms underlying climatedriven vegetation mortality. Trends in Ecology and Evolution 26, 523-532 (2011).
10
Anderegga, W. R. L. et al. The roles of hydraulic and carbon stress in a widespread
climate-induced forest die-off. Proceedings of the National Academy of Sciences of
the United States of America 109, 233–237 (2012).
8
11
Bréda, N., Huc, R., Granier, A. & Dreyer, E. Temperate forest trees and stands under
severe drought: A review of ecophysiological responses, adaptation processes and
long-term consequences. Annals of Forest Science 63, 625-644 (2006).
12
Breshears, D. D. et al. Regional vegetation die-off in response to global-change-type
drought. Proceedings of the National Academy of Sciences of the United States of
America 102, 15144-15148 (2005).
13
Ciais, P. et al. Europe-wide reduction in primary productivity caused by the heat and
drought in 2003. Nature 437, 529-533 (2005).
14
Zhao, M. & Running, S. W. Drought-induced reduction in global terrestrial net
primary production from 2000 through 2009. Science 329, 940-943 (2010).
15
Lewis, S. L. Tropical forests and the changing earth system. Philosophical
Transactions of the Royal Society B: Biological Sciences 361, 195-210 (2006).
16
Pockman, W. T., Sperry, J. S. & O'Leary, J. W. Sustained and significant negative
water pressure in xylem. Nature 378, 715-716 (1995).
17
Tyree, M. T. & Sperry, J. S. Vulnerability of xylem to cavitation and embolism. Annu.
Rev. Plant Physiol. Plant Molec. Biol. 40, 19-38 (1989).
18
Sperry, J. S., Hacke, U. G. & Pittermann, J. Size and function in conifer tracheids and
angiosperm vessels. Am. J. Bot. 93, 1490-1500 (2006).
19
Sperry, J. S., Adler, F. R., Campbell, G. S. & Comstock, J. P. Limitation of plant
water use by rhizosphere and xylem conductance: results from a model. Plant Cell
Environ. 21, 347-359 (1998).
20
Alder, N. N., Sperry, J. S. & Pockman, W. T. Root and stem xylem embolism,
stomatal conductance, and leaf turgor in Acer grandidentatum populations along a soil
moisture gradient. Oecologia 105, 293-301 (1996).
9
21
Meinzer, F. C., Johnson, D. M., Lachenbruch, B., McCulloh, K. A. & Woodruff, D.
R. Xylem hydraulic safety margins in woody plants: Coordination of stomatal control
of xylem tension with hydraulic capacitance. Functional Ecology 23, 922-930 (2009).
22
Allison, I. et al. The Copenhagen Diagnosis: updating the world on the latest climate
science. (The University of New South Wales Climate Research Centre, 2009).
23
Zhang, X. et al. Detection of human influence on twentieth-century precipitation
trends. Nature 448, 461-465 (2007).
24
Meir, P. & Woodward, F. I. Amazonian rain forests and drought: Response and
vulnerability. New Phytol. 187, 553-557 (2010).
25
Brodersen, C. R., McElrone, A. J., Choat, B., Matthews, M. A. & Shackel, K. A. The
dynamics of embolism repair in xylem: In vivo visualizations using high-resolution
computed tomography. Plant Physiol. 154, 1088-1095 (2010).
26
Delzon, S., Douthe, C., Sala, A. & Cochard, H. Mechanism of water-stress induced
cavitation in conifers: Bordered pit structure and function support the hypothesis of
seal capillary-seeding. Plant, Cell and Environment 33, 2101-2111 (2010).
27
Maherali, H., Pockman, W. T. & Jackson, R. B. Adaptive variation in the
vulnerability of woody plants to xylem cavitation. Ecology 85, 2184-2199 (2004).
28
Bigler, C., Bräker, O. U., Bugmann, H., Dobbertin, M. & Rigling, A. Drought as an
inciting mortality factor in scots pine stands of the Valais, Switzerland. Ecosystems 9,
330-343 (2006).
29
Choat, B., Ball, M. C., Luly, J. G. & Holtum, J. A. M. Hydraulic architecture of
deciduous and evergreen dry rainforest tree species from north-eastern Australia.
Trees - Structure and Function 19, 305-311 (2005).
30
Lamy, J. B. et al. Uniform selection as a primary force reducing population genetic
differentiation of cavitation resistance across a species range. PLoS ONE 6 (2011).
10
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank the ARC-NZ Vegetation Function Network for hosting the
original working group from which the data set was compiled.
Author Contributions BC and SJ led the initial working group and coordinated analysis and
write up of the work. All authors contributed to compilation and organization of the dataset.
SG and IJW extracted climate data from WorldClim and CRU Climate Database. HM, MM
and JMV assisted in statistical analyses of the dataset.
Author Information The authors declare that they have no competing financial interests.
Figure legends
Figure 1. Minimum xylem pressure (Ψmin) as a function of embolism resistance (Ψ50) for
angiosperm and gymnosperm species. The dashed line indicates the xylem pressure at which
a 50% loss of hydraulic conductivity occurs. The safety margin is the distance between each
point and this line. Error bars show standard deviation. Points were binned in 1.0 MPa
increments for Ψ50 with a total sample size of 209 species for angiosperms and 36 species for
gymnosperms. Bins were pooled with the next lowest bin if they contained only one sample.
Regression lines shown were fitted to raw data (angiosperms r2 = 0.57, P < 0.0001,
gymnosperms r2 = 0.59, P < 0.0001).
Figure 2. Box plot of hydraulic safety margins (Ψmin – Ψ50) for angiosperm and gymnosperm
species across major forest biomes, with the dashed line indicating the pressure causing a
11
50% loss of hydraulic conductivity in the stem xylem. Boxes show the median, 25th and 75th
percentiles, error bars show 10th and 90th percentiles, filled symbols show outliers.
Gymnosperm species were not represented in tropical forests. Significant differences (P <
0.01) between biome means are indicated by different letters above boxes with angiosperms
and gymnosperms considered separately.
Figure 3. Embolism resistance (Ψ50) as a function of mean annual precipitation (MAP) for
angiosperm and gymnosperm species with each point representing one species (N = 384 for
angiosperms and 94 for gymnosperms). A generalised model indicated Ψ50 was significantly
related (P < 0.00001) to MAP for angiosperms and gymnosperms (see methods summary for
details).
12