Seasonal responses of xylem sap velocity to VPD and solar radiation during drought in a stand of native trees in temperate Australia

Seasonal responses of xylem sap velocity to VPD and solar radiation during drought in a stand of native trees in temperate Australia
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
    © CSIRO 200410.1071/FP032201445-4408/04/050461   Functional Plant Biology   , 2004, 31   , 461–  CSIRO  PUBLISHING  Seasonal responses of xylem sap velocity to VPD and solar radiation during drought in a stand of native trees in temperate Australia   Melanie J. B. Zeppel    A,C   , Brad R. Murray   A   , Craig Barton   B  and Derek Eamus   A  A  Institute for Water and Environmental Resource Management, University of Technology, Sydney, NSW 2065, Australia.  B  State Forests of New South Wales, Research and Development Division, 121–131 Oratava Avenue, West Pennant Hills, NSW 2125, Australia.  C  Corresponding author; email:   Abstract.  Xylem sap velocity of two dominant tree species,  Eucalyptus crebra  F. Muell. and Callitris glaucophylla  J. Thompson & L.A.S. Johnson, in a native remnant forest of eastern Australia was measured in winter and summer during a prolonged (> 12 months) and extensive drought. The influence of vapour pressure deficit (VPD) and solar radiation levels on the velocity of sap was determined. Pronounced hysteresis in sap velocity was observed in bothspecies as a function of VPD and solar radiation. However, the rotation of the hysteresis curve was clockwise for the response of sap velocity to VPD but anti-clockwise in the response of sap velocity to radiation levels. A possiblereason for this difference is discussed.The degree of hysteresis (area bounded by the curve) was larger for the VPD response than the response to solar radiation and also varied with season. A simple linear model was able to predict sap velocity from knowledge of VPD and solar radiation in winter and summer. The consistent presence of hysteresis in the response to sap velocityto VPD and solar radiation suggests that large temporal and spatial models of vegetation water use may require some provision for the different responses of sap velocity, and hence water use, to VPD and solar radiation, betweenmorning and afternoon and between seasons.   Keywords  : drought,  Eucalyptus  , hysteresis, sap velocity, vapour pressure deficit.  Introduction  Tree water use (expressed either as sap velocity or volumeflux) is influenced by several key factors including soilmoisture content (Pataki et al   . 1998), atmospheric water content (vapour pressure deficit, VPD), solar radiation and tree size (O’Grady et al   . 1999). As soil moisture contentdeclines below a threshold, tree water use declines(Whitehead and Jarvis 1981). The pattern of water use inresponse to increasing VPD is more complex, and encom- passes increases and decreases in water use and sap velocitydepending on the range of VPD (Whitehead and Jarvis1981; Benyon et al   . 2001). Generally, increases in solar radiation and tree size cause increased rates of tree water use, although a plateau in water use can be observed whenforest canopies achieve closure (Medhurst et al   . 2002).However, interactions among soil moisture content, VPDand solar radiation as determinants of sap velocity havereceived little attention in Australian native tree species.Daily and seasonal changes in VPD, solar radiation and soil moisture can be large, and may have significant impactson daily and seasonal patterns of water use (Myers et al   .1997; Prior et al   . 1997; Hutley et al   . 2000). Superimposed on these relatively (broadly) predictable patterns of changeis the influence of extreme events such as long-termdrought. In late 2001 and for all of 2002, eastern Australiasuffered a prolonged and extreme drought (Bureau of Meteorology 2003). In order to develop a realistic mechan-istic understanding of patterns of tree water use it isimportant to understand the responses of tree water use notonly to the more predictable daily and seasonal patterns thatcharacterise an ‘average’ year, but also the responses tomore unpredictable events such as prolonged drought.Determinations of stand water use and the water balanceof catchments generally require that rates of water use beextrapolated from measurements on individual trees towhole tree stands and plantations, meaning that water use isscaled spatially. In addition to spatial scaling, temporalscaling is also required. Temporal scaling (extrapolatingdata from short time frames of days or weeks to longer,seasonal or annual, timeframes) can be problematic and    Abbreviations used: ABA, abscic acid; DBH, diameter at breast height; VPD, vapour pressure deficit.    462    Functional Plant Biology   M. J. B. Zeppel et al    .  depend upon how representative the short measurement periods are of the longer time frame (Wullschleger et al   .1998). Thus, it is important to consider daily, seasonal and annual patterns of water use in typical and atypical yearswhen estimating ecosystem water balance.Hysteresis occurs when an increase in a given independ-ent variable,  x  does not cause the same response in adependent variable,  y  , when variable  x  decreases. For example, if the rate of tree water use is different in themorning (when VPD is increasing) compared with theafternoon (when VPD is decreasing), then hysteresis isoccurring. Hysteresis may be a significant factor in hourlyand daily patterns of tree water use and therefore should beinvestigated if a more complete understanding of treefunction and catchment hydrology is desired. Water use isfrequently estimated from models that do not consider theoccurrence of hysteresis (Eamus 2001).Hysteresis has been found in the responses of stomata totranspiration and humidity (Meinzer et al   . 1997). Counter-intuitively, a more marked hysteresis (a larger area inside thehysteresis loop) in the relationship between VPD and transpiration was observed during the dry season than in thewet season in a north Australian savanna (O’Grady et al   .1999). Causes of hysteresis in the response of water use toVPD remain poorly understood and may include a contri- bution of stored water in the tree stem or changes in soil-to-leaf hydraulic conductivity (O’Grady et al   . 1999). A contri- bution to daily water use from stored water is likely, and can be deduced from differences in the time of initiation of water-flow in the stem at the top of the trunk compared withflow at the bottom of the trunk. This time lag may vary fromhalf an hour to several hours, and has a significant effect onthe relationship between diurnal sap flux at the base of thetree and VPD (Wullschleger et al   . 1998; Oren et al   . 1999).Australian native woodlands usually contain several treespecies of various sizes. Each of these species may havedifferent relationships among water use, tree size and VPD.An understanding of these relationships for different speciesis essential when calculating stand water use in mixed stands(Wullschleger et al   . 1998). Kolb and Stone (2000) found that the stomatal conductance of one species,  Pinus ponderosa  Douglas ex Lawson was highly sensitive to dailyand seasonal changes in VPD, while conductance of another co-occurring species, Quercus gambelii   Nutt., was lesssensitive. These different sensitivities to VPD were used toexplain the greater tolerance of Q. gambelii  to atmosphericwater stress.The present study was conducted during the mid and latestages of a prolonged (more than 12 months) drought ineastern Australia. Specifically, we addressed the followingquestions.1.How does sap velocity in two native Australian treespecies vary as a function of VPD and solar radiation inwinter and summer?2.Is hysteresis observed in the response of sap velocity tochanges in VPD and solar radiation?3.Can a simple model be formulated that allows predictionof sap velocity from solar radiation and VPD in winter and summer for the two dominant native tree species?  Materials and methods   Site description   The study site was located in remnant woodlands of the LiverpoolPlains, approximately 90 km south of Tamworth on the north-west plains of New South Wales, Australia (31.5° S, 150.7° E, elevation390 m). Vegetation at the site consisted of open woodland, with anaverage height of 15 m, dominated by  Eucalyptus crebra   F. Muell. and    Callitris glaucophylla   J. Thompson & L.A.S. Johnson. These twospecies account for approximately 75% of the tree basal area at the site.Soils at the site were well-drained acidic lithic bleached earthy sands(Banks 1998) with pockets of clay. Total tree basal area for the site was23.8 ± 3.4 m  2   ha   –1   .    Rainfall, temperature and soil moisture   Rainfall data were obtained from the Bureau of Meteorology from ameteorological station located approximately 5 km west of the studysite. Temperature data (aspirated wet and dry bulb) were obtained froma screened climate station (Environdata Pty Ltd, Warwick, Qld) located approximately 500 m from the study site in a cleared field (approximately 4 ha) while total solar radiation was measured abovethe screen. Vapour pressure deficit (VPD) was calculated from wet and dry bulb temperatures. Soil moisture was measured on site with aneutron moisture metre probe (model ECH2O, Irricrop Technologies,Armidale, NSW) in three centrally located access holes. Soil moisturewas measured at 20, 40 and 80 cm depth.   Sap flow measurement    Sap velocity was measured using the heat pulse technique withcommercial sap-flow sensors (Greenspan Technology Pty Ltd,Warwick, Qld). Two probesets (four sensors) were inserted at 90° toeach other in each tree. A preliminary Monte Carlo simulation wasconducted to determine the number of sap-flow sensors required toobtain an acceptable level of accuracy, and four probes were deemed adequate. A minimum of seven and a maximum of 15 trees wereinstrumented for each species at each sampling time.The sap flow loggers recorded the heat pulse times at 15 minintervals over a 2-week period during July–August 2002 (mid-drought;winter) and January–February 2003 (summer). Sap velocities werecalculated for each sensor for 4 d after allowing 2 d for stabilisation of the wound zone that develops as a result of drilling into the wood (Olbrich 1991). The weighted averages technique of Hatton and Wu(1995) was employed to convert sap velocities to transpired water volume.Volume fractions of wood and water in the sapwood were determined gravimetrically on 5 mm cores taken from 10 trees of each species ontwo occasions. In  E. crebra   the mean wood fraction was 0.56 and 0.50 inwinter and summer, respectively. The mean water fraction was 0.23 and 0.28 in winter and summer, respectively. In C. glaucophylla   the meanwood fraction was 0.33 in winter and 0.34 in summer. The mean water fraction was 0.52 in winter and 0.48 in summer.    Radial sap-flow profiles   Radial profiles of sap flow through the sapwood of each species weredetermined in order to calculate the regions of maximum flow acrossthe sapwood. Sap flow was measured at a minimum of six depthsacross the sapwood, replicated three or four times in different aspects    Responses of sap velocity to VPD and radiation    Functional Plant Biology   463in each tree. Knowledge of the region of maximum sap flow across thesapwood was used to calculate the depth to insert the sap-flow sensors.The full method is described by O’Grady (2000).   Wound estimation   The width of the wound around the holes used to insert the probes wasmeasured twice in seven trees of each species, using the techniquedescribed by O’Grady (2000). A wound diameter of 2.5 mm for    C. glaucophylla   and 3.7 mm for  E. crebra   was used to correct velocityestimates.   Sapwood area and depth   Sapwood cross-sectional area was determined using a 5-mm incrementcorer, twice per tree, at approximately 1.3 m height. Distinct colour changes were observed between bark, sapwood and heartwood in bothspecies.    Data analyses   Two second-order polynomial regressions were fitted to each hysteresisloop, one for the upper curve and one for the lower curve, with MSExcel (2000). Hysteresis area was derived by calculating the area between the upper polynomial and the  x   -axis, and subtracting the area between the lower polynomial and the  x   -axis. Linear regressionanalysis was used to examine the relationship between hysteresis areaand maximum daily VPD for summer and winter separately. Thecombined influence of radiation and VPD on sap velocity wasmodelled using a general linear model (Sigmaplot 7.0, HearneScientific Software Plc, Melbourne). The equation for the model was:sap velocity = a   + b ×   radiation + c ×   VPD. We calculated r   2   values todetermine the amount of variation in sap velocity explained byradiation and VPD, and the three-dimensional plane predicted by themodel was plotted with the measured values.  Results  Climate and soil moisture  Average annual rainfall for the Liverpool Plains is 680 mm,with approximately 50% of this occurring between October and February (5 months) and 50% occurring between Marchand September (7 months; Fig. 1  a  ). However, during thestudy period (winter 2002, summer 2002–2003) rainfall wassubstantially lower than the 20-year average (Fig. 1  a  ). Totalrainfall in 2002 was 366 mm, 60% of the 20-year average. Inwinter 2002 (June–August) rainfall was approximately 25%of the average winter rainfall, while in summer 2002–2003(December–February) rainfall was approximately 60% of the long-term average (Fig. 1  a  ).Volumetric soil moisture at 20 cm depth was low or verylow throughout the study period, and never exceeded 0.14 m  3  m   –3   but was often below 0.08 m  3  m   –3  (Fig. 1  b  ).Rainfall in May and June and also in September, November and December, was sufficient to raise the soil moisturecontent at 20 cm from approximately 0.03 to 0.08 and 0.13 m  3  m   –3  , respectively. Soil moisture at 40 cm depth wastypically 0.07 m  3  m   –3  for much of the study period, butincreased to almost 0.1 m  3  m   –3  and to 0.17 m  3  m   –3   by theMay–June (winter) and September–December (summer)rainfall, respectively. At 80 cm depth soil moisture wasunaffected by any rainfall and was typically 0.15 m  3  m   –3  throughout the study period.Maximum winter temperature in June 2002 was 15°Cand midday solar radiation was 2.5 MJ m   –2  h   –1  (Fig. 2  a  ).Maximum summer temperature during January 2003 was30°C and midday solar radiation was 4.0 MJ m   –2  h   –1  (Fig. 2  b  ). Summer sap velocity was higher than winter sapvelocity (Fig. 2  a  , b  ). a 0102030405060708090J F M A M J J A S O N D J F    R  a   i  n   f  a   l   l   (  m  m   ) b F M A M J J A S O N D J F M A M J Month    V  o   l  u  m  e   t  r   i  c  w  a   t  e  r  c  o  n   t  e  n   t   (  m    3   m   –   3    ) Fig. 1. ( a ) Rainfall at the study site, long-term mean monthly values(closed bars, Bureau of Meteorology 2003) compared with actualrainfall, January 2002 to January 2003 (open bars). ( b ) Volumetric water content of soil from January 2002 to June 2003. Data shown representthe average of three holes at 20 cm ( × ), 40 cm (  ) and 80 cm (  ). a 051015202530350.    S  a  p  v  e   l  o  c   i   t  y   (  c  m    h   –   1    ) b 051015202530355 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Time of day    T  e  m  p  e  r  a   t  u  r  e   (   °   C   )    R  a   d   i  a   t   i  o  n   (   M   J  m   –   2    h   –   1    )   T  e  m  p  e  r  a   t  u  r  e   (   °   C   )   S  a  p  v  e   l  o  c   i   t  y   (  c  m    h   –   1    )   R  a   d   i  a   t   i  o  n   (   M   J  m   –   2    h   –   1    ) Fig. 2. Daily patterns of radiation ( × ), sap velocity (  ) and temperature (  ) for the period from 0600 to 2000 h during June 2002,winter ( a ) and January 2003, summer ( b ). Sap velocity data representthe average of three  Eucalyptus crebra  trees.    464    Functional Plant Biology   M. J. B. Zeppel et al    .  The relationship between sap velocity, VPD and  solar radiation  A distinct hysteresis loop was evident in the relationship between sap velocity and vapour pressure deficit (VPD) for  both species (Fig. 3  a  , b  ) As VPD increased in the morning,sap velocity increased and as VPD decreased in the afternoon,sap velocity declined. At any given VPD, sap velocity washigher in the morning than the afternoon (clockwise rotationof the hysteresis loop) for both species in the summer (Fig. 3  a  , b  ). In the winter, the hysteresis loop was muchreduced in area in C. glaucophylla  and  E. crebra  (Fig. 3  a  , b  ).A distinct hysteresis loop, with a counter clockwiserotation (afternoon values of sap velocity were larger thanmorning sap velocities, at any given value of solar radiation)was observed in the relationship between solar radiation and sap velocity for both species, in the summer (Fig. 3  c  , d   ). Inwinter no hysteresis was observed for C. glaucophylla  (Fig. 3  c  ) but a small hysteresis loop was observed in winter in   E. crebra  (Fig. 3  d   ). Hysteresis in the relationship betweenvolume flux of water through the stem (L h   –1  ) as a function of VPD was observed in summer and winter for both species. Anexample of day-to-day variation in hysteresis for both speciesis presented in Fig. 4 using a normalised plot to remove theinfluence of tree size from the data. During this 4-d period,maximum daytime wet and dry bulb temperatures were11.2 and 25.9°C. Maximum total solar radiation was14.8 MJ m   –2  d    –1  . Maximum PET was 5.4 mm d    –1  , and no rainfell during the 4-d period. From the curves shown in Fig. 4, arelationship between the area within the curve and maximumdaily VPD was derived for both species in winter and summer (Fig. 5). For both species, a linear increase in the area of thenormalised hysteresis curve occurred with increasing VPD.Both species were very similar in winter and a singleregression was applied to the combined data (Fig. 5  a ) but insummer two regressions were applied to the two species(Fig. 5 b ).Sap velocity increased as solar radiation and VPDincreased in both species in winter and summer (Figs 6, 7).For both species, sap velocity was greater in summer thanwinter, reflecting the warmer temperatures and higher insolation in February 2003 (summer) compared withAugust 2002 (winter). Sap velocity could be reasonably predicted in summer and winter, for both species, fromknowledge of solar radiation and VPD, with regressioncoefficients larger than 0.75 being observed in all cases. Discussion The drought of 2001–2003 has been considered one of thelargest (in areal extent), longest and most severe of the past100 years (Bureau of Meteorology 2003). The total annualrainfall for 2002 represents 60% of the 20-year annualaverage. Measurements of sap velocity in winter 2002 were a 05101520253035    S  a  p  v  e   l  o  c   i   t  y   (  c  m    h   –   1    ) b 051015202530350 0.5 1 1.5 2 2.5 Vapour pressure deficit (kPa)    S  a  p  v  e   l  o  c   i   t  y   (  c  m    h   –   1    ) cd 0 1 2 3 4 5 Radiation (MJ m  –2 h  –1 ) Fig. 3. The relationship between vapour pressure deficit (kPa) and sap velocity (cm h  –1 ). Data represent the average of three trees over 24 h inwinter 2002 (  ) and summer 2003 (  ). Data for Callitris glaucophylla  are shown in ( a ) and  Eucalyptus crebra  in ( b ). The relationship betweenradiation (MJ m  –2  hr   –1 ) and sap velocity (cm hr   –1 ). Data represent the average of three trees over 24 h in winter 2002 (  ) and summer 2003 (  ).Data for Callitris glaucophylla  are shown in ( c ) and  Eucalyptus crebra  in ( d  ).  Responses of sap velocity to VPD and radiation  Functional Plant Biology 465 made when rainfall in the preceding 3 months was 25% of the 20-year average. During this period soil moisture in theupper 20 and 40 cm was less than 0.1 m 3  m  –3 , and generallyconsiderably lower (0.025 m 3  m  –3  in May and October 2002). Even at 80 cm depth, soil moisture was low, typicallyat 0.18 m 3  m  –3 . The rain in winter or summer wasinsufficient to increase soil moisture at 80 cm depth.Sap velocities in summer, for both species were signifi-cantly higher (2–10 times) than sap velocities in winter. Threereasons explain this. First, peak solar radiation and total dailysolar radiation levels were approximately 50–100% greater,respectively, in summer compared with winter. Second, air temperatures and VPD were both larger in summer thanwinter, thereby increasing the atmospheric demand on thecanopy. Finally, soil moisture at 40 cm depth was 50% greater in summer than in winter. The influence of higher solar radiation, temperature and soil moisture content within therooting zone on sap velocity or transpiration is well docu-mented (O’Grady 2000; Wullschleger et al  . 2001; Medhurst et al  . 2002). However, the largest increase in solar radiationand soil moisture content (100 and 50%, respectively) between the winter and summer periods appears inadequateto fully explain the 2- to 10-fold increase in sap velocitiesobserved at the same time. However, in the 6 weeks beforethe summer measurements of sap velocity, soil moisturecontent increased 4- to 5-fold because of significant rainevents in December. Therefore, we propose that sap velocityon a given day can be influenced by the water status of thesoil and plant in the preceding 1–6 weeks. Mechanisms for this long-lasting effect may include growth of new leaves,which differ in their stomatal behaviour compared with older leaves; an increase in the surface area of fine roots, therebyincreasing the ability of roots to extract water; emboli repair,thereby increasing the conducting area of sapwood; or decreased abscisic acid content of xylem sap resulting fromincreased soil water availability, thereby increasing stomatalconductance. Clearly all of these mechanism may contributeto the observed increase in sap velocity in summer.Stomatal responses to soil moisture content and VPD areinfluenced by previous exposure to soil or atmosphericdrought (Eamus 1987; Thomas and Eamus 1999; Thomas et al  . 2000). Therefore, it is proposed that stomatal regu-lation of transpiration (and hence sap velocity) in winter wasstronger than in summer because in summer a 3-month period (November–January) of larger soil moisture content preceded the measurement period. This is despite the soilmoisture content at 20 and 40 cm depth at the time of measurement (summer) being increased by only a small a    P  r  o  p  o  r   t   i  o  n  o   f  m  a  x   Q b 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 VPD (kPa)    P  r  o  p  o  r   t   i  o  n  o   f  m  a  x   Q Fig. 4. The daily pattern of tree water use (Q) over four consecutive days during drought, in winter.  ,  , ×  and  represent 2, 3, 4 and 5 August 2003, respectively. Data represent the proportion of maximum tree water use for Callitris glaucophylla  ( a ) and  Eucalyptus crebra  ( b ). H area  was calculated  by multiplying the proportion of maximum Q (dimensionless) from Fig. 3 by VPD (kPa).
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!