Atmospheric and soil moisture controls on ...

HYDROLOGICAL PROCESSESHydrol Process (2013)Published online in Wiley Online Library(wileyonlinelibrarycom) DOI 101002hyp9879

Atmospheric and soil moisture controls on evapotranspirationfrom above and within a Western Boreal Plain aspen forest

S M Brown1 R M Petrone1dagger L Chasmer1 C Mendoza2 M S Lazerjan1 S M Landhaumlusser3

U Silins3 J Leach4 and K J Devito51 Cold Regions Research Centre Wilfrid Laurier University Waterloo Ontario Canada N2L 3C52 Department of Earth and Atmospheric Sciences University of Alberta Edmonton Alberta Canada

3 Department of Renewable Resources University of Alberta Edmonton Alberta Canada4 Department of Geography University of British Columbia Vancouver British Columbia Canada

5 Department of Biological Sciences University of Alberta Edmonton Alberta Canada

CWiE-mdaggerCuUnN2

Co

Abstract

The Western Boreal Plain of North Central Alberta comprises a mosaic of wetlands and aspen (Populus tremuloides) dominateduplands where precipitation (P) is normally exceeded by evapotranspiration (ET) As such these systems are highly susceptible tothe climatic variability that may upset the balance between P and ET Above canopy evapotranspiration (ETC) and understoryevapotranspiration (ETB) were examined using the eddy covariance technique situated at 255 m (75 m above tree crown) and40 m above the ground surface respectively During the peak period of the growing seasons (green periods) ETC averaged 308mm d1 and 345 mm d1 in 2005 and 2006 respectively while ETB averaged 156 mm d1 and 195 mm d1 Early in thegrowing season ETB was equal to or greater than ETC once understory development had occurred However upon tree crowngrowth ETB was lessened due to a reduction in available energy ETB ranged from 42 to 56 of ETC over the remainder of thesnow-free seasons Vapour pressure deficit (VPD) and soil moisture (y) displayed strong controls on both ETC and ETB ETC

responded to precipitation events as the developed tree crown intercepted and held available water which contributed to peakETC following precipitation events gt10 mm While both ETC and ETB were shown to respond to VPD soil moisture in therooting zone is shown to be the strongest control regardless of atmospheric demand Further soil moisture and tension datasuggest that rooting zone soil moisture is controlled by the redistribution of soil water by the aspen root system Copyright copy2013 John Wiley amp Sons Ltd

KEY WORDS evapotranspiration boreal forest Populus tremuloides western boreal plain soil moisture

Received 10 April 2012 Accepted 29 April 2013

INTRODUCTION

The clonal nature and spatial distribution of Populustremuloides Michx (trembling aspen) are drivers that canexert strong controls on evapotranspiration (ET) (Blankenet al 2001) and have the potential to significantly alter thewater balance at scales ranging from the stands tolandscapes (Wullschleger and Hanson 2006 IPCC2007) Total ET is defined as the sum of the canopyevapotranspiration (ETC) and understory evapotranspiration(ETB) (including shrubs ground cover and soil surface)ETC is often the largest component of both the totalevaporative flux to the atmosphere and the overall waterbudget of forested ecosystems (Schafer et al 2002)

orrespondence to Richard Petrone Cold Regions Research Centrelfrid Laurier University Waterloo Ontario Canada N2L 3C5ail rpetronewlucarrent address Department of Geography and EnvironmentalManagementiversity of Waterloo 200 University Ave West Waterloo ON CanadaL 3G1

pyright copy 2013 John Wiley amp Sons Ltd

The sub-humid boreal plain (BP) of CanadarsquosWestern Boreal Forest is especially sensitive to changesin water balance In this region ET often exceedsprecipitation (Devito et al 2005a) thus changes in thespatial and age distribution of aspen forests and inclimate variability could have significant influences onthe local to regional water balance influencing foresthealth and productivityETC is significantly influenced by canopy structure but

also varies with other conditions influencing ET whichinclude radiative transfer of solar energy through the canopyto the understoryground surface (Chen et al 1997 Blankenet al 2001 Ni-Meister and Gao 2011) interceptedprecipitation (Ahrends and Penne 2010) understory devel-opment (Brown et al 2010) and within canopy atmosphericturbulence vapour pressure deficit (VPD) and stomatalconductance (Hogg and Hurdle 1997 Dang et al 1997Blanken and Black 2004) Chen et al (1993) found largedifferences in local meteorology such as air temperatureVPD and soil moisture among recently clear-cut Douglas fir

S M BROWN ET AL

forest at the clear-cut edge and within a mature Douglas firforest sites Gradients between the factors controlling ET cansignificantly affect ET fluxes especially as increasinglylarger areas are harvested or altered in some way Forexample ETC is linearly controlled byVPD until a thresholdis reached (Hogg and Hurdle 1997 Hogg et al 1997Wullschleger et al 1998 Hogg et al 2000 Kurpius et al2003 Bovard et al 2005) and soil moisture (Wullschlegeret al 1998 Cinnirella et al 2002) However while negativefeedbacks occur between VPD and canopy conductance(GC) Stomatal conductance is often unaffected by soilmoisture until a deficit occurs at which point transpiration islimited as trees close stomata in an effort to conserve water(Wullschleger et al 1998 Phillips and Oren 2001Cinnirella et al 2002 Irvine et al 2002) The relationshipbetween stomatal conductance and soil moisture varies fromsite to site andwithin a stands due to differences in vegetationspecies type structure and soil texture as different plantsrespond differently to given soil water states (Roberts 2000)Barbour et al (2005) also found that canopy wetness wasstrongly correlated with ETC Within closed canopies waterexchanges between the understory and atmosphere wererelatively small due to reduced canopy wetness and incidentradiation (Wullschleger et al 1998) whereas the oppositewas true in open canopies which were also prone to spatiallyvariable water interception (Unsworth et al 2004 Barbouret al 2005) Differences in radiation penetration throughcanopies due to differences in leaf area also influence thegrowth and distribution of understory species such as Rosaacicularis and Viburnum edule which are prevalent in aspenstands Proliferation of understory vegetation includingaspen seedlings which are established through a clonal rootsystem also increase water exchanges between the ecosystemand the atmosphere (Peterson and Peterson 1992 Blankenet al 1998)Trembling aspen may also exacerbate limitations to water

balance in theBP because full forests exists over large areas ofwell-drained soils of a region already with a net moisturedeficit (ETgtP) Therefore understanding the spatial distri-bution and heterogeneity of aspen forests how a forest ofaspen is maintained and impacts to ET and water resources isan important question within this region Currently little isknown about the sensitivity of ET to variability in thecontributions to fluxes within the canopy and the understoryin aspen forests of theBP yet the implications for this researchcan be significant given frequent drought periods andinfrequent lsquowetrsquo periods that occur only about every 20 years(Devito et al 2005a) Understanding the natural distributionof ETC and ETB provide important baseline for interpretingthe influences of climate change on forest-stand balances andhow natural (ie fire) or anthropogenic disturbance of forestcanopy influence water resource issues in the regionThe objective of this study is to characterize and quantify the

partitioning of ET fluxes within an aspen-dominated upland

Copyright copy 2013 John Wiley amp Sons Ltd

within the BP of Canadarsquos Western Boreal Forest This studyis part of a larger project examining water sink and sourcedynamics of aspenndashpeatlandndashpond ecosystems typical of theBP a region that persists with a net moisture deficit climatecycles of infrequent wet periods deep soils and large seasonalpatterns in precipitation and antecedent moisture conditionsWe examine the interaction of vegetation structure meteoro-logical and soil controls on the ET ETC and ETB within thisclimate context It is hypothesized that the evaporation ofintercepted water will comprise a significant portion of thedaily canopy ET due to the drier atmosphere Further thevegetation response to atmospheric demand will ultimately bemediated by hydraulic redistribution of soil moisture fromwetareas to the drier aspen stands to meet atmospheric demand Itis hypothesized that driven by diurnal cycles in ET the clonalroots of the aspen standwill bringmoisture to the rooting zonewhich will be evident by examining the interactions of soilsuction soil moisture and stand-scale ET

STUDY SITE

Utikuma region study area

The study site is located within the Utikuma RegionStudy Area (URSA) 320 km north of Edmonton AlbertaCanada (56o4` N 115o28` W Figure 1) URSA is situatedapproximately 150 km south of the discontinuouspermafrost zone (Woo and Winter 1993) within the BPsregion of the Western Boreal Forest The climate withinURSA is characterized by cold winters (146 C) andwarm summers (156 C) (Environment Canada 2005)Annual ET averages 517 mm (Bothe and Abraham 1993)which is slightly higher than the average annual precipita-tion of 481 mm at URSA (Environment Canada 2005)Snowfall in this area averages less than 150 mm yr1 andrepresents less than 30 of the total annual precipitation forthe region (Marshall et al 1999 Devito et al 2005a)Although a water deficit exists during most years in thesub-humidBP annual precipitation surpassesETevery10ndash15years with significant runoff only every 20 years or more(Devito et al 2005a) Approximately 50 of annualprecipitation occurs from June through August coincident withthe period of maximum evaporative demand and is precededand followed by a relatively dry spring and fall periodsrespectively (Devito et al 2005a Petrone et al 2007)URSA is characterized by the presence of rolling

moraines of low relief low lying clay plains and coarse-textured outwash plains where glacial overburdens rangefrom 20 to 240 m in thickness and overlie the UpperCretaceous Smoky Shale Group (Vogwill 1978) Thisstudy was conducted on the moraine portion of theregion (Figure 1) The moraine landscape was chosen asthese systems are dominated by the presence of verticalexchanges of water (Devito et al 2005a Redding and

Hydrol Process (2013)

Figure 1 Schematic diagram of the research catchment location of the instrumentation and dominant vegetation landcover units comprising thecatchment Pond 40 Utikuma Region Study Area (URSA) Alberta Canada

CONTROLS ON ASPEN EVAPOTRANSPIRATION IN THE WESTERN BOREAL PLAIN

Devito 2006 Redding and Devito 2008) making uplandshighly susceptible to fluctuations in climate (Devito et al2005b)

Study catchmentThe dominant overstory species of the study catchment is

trembling aspen (Populus tremuloides) (covering ~68 ofthe catchment determined from remote sensing data) withsporadic clusters of balsam poplar (Populus balsamifera L)in depressions (Figure 1) Fire in 1962 reset much of theecosystem resulting in an average stand age in both studyareas of approximately 45 years The understory vegetationat the site is well developed with a shrub layer dominated byprickly rose (Rosa acicularis) and low bush-cranberry(Viburnum edule) and the herbaceaous layer by twinflower(Linnea borealis) and sarsaparilla (Aralia nudicaulis) Theforest floor surface is dominated by aspen litter withnegligible amounts of mosses and lichens Above canopyand within canopy eddy covariance measurements wereestablished on a permanent tower and a mobile tower inan upland area (~7 m total change in elevation across thestudy catchment Figure 1) consisting of Gray Luvisolicsoils (Soil Classification Working Group 1998) developedfrom disintegration moraine deposits which are typicallysilt-rich but spatially heterogeneous with zones of highclay or sand contents (Fenton et al 2003 Redding andDevito 2008)


METHODSAbove and within canopy ET measurements The eddy

covariance technique was used to measure continuous ETfluxes above an aspen-dominated upland stand during the2005 and 2006 snow-free seasons (DOY 187 to 291 andDOY 123 to 283 respectively) Eddy covarianceinstrumentation was located 70 m above the aspen treeswhich ranged in height from approximately 17 m to 21 mon average The measurement system consisted of a three-dimensional sonic anemometer (CSAT3 CampbellScientific USA) an open-path infrared gas analyzer(LI-7500 Li-Cor USA) and a 25 mm fine-wirethermocouple sampled at a frequency of 10 Hz Meanhorizontal (v) and vertical (w) wind velocities weremathematically rotated to zero following Kaimal andFinnigan (1994) and ET was calculated as described byWebb et al (1980) Mean flux and atmosphericturbulence statistics were computed every 30 min andstored on a CR23X data-logger (Campbell ScientificUSA) Net all-wave radiation (Q) (NR-lite Kipp andZonen Netherlands) air temperature and relative humid-ity (HMP45C Vaisala Oyj Finland) and photosyntheticactive radiation (PAR) (PAR-lite Kipp and ZonenNetherlands) were also measured above the canopy onthe permanent tower 70 m above the tree crownSoil moisture and energy flux data were collected at

three sites located 20 m 25 m and 85 m from the tower


S M BROWN ET AL

(Figure 1) Soil heat fluxes (QG) were measured using twoheat flux transducers (HFT-03 Campbell ScientificUSA) at each site buried 005 m below the litter fallhorizon (LFH)ndashsoil interface Soil temperature and heatstorage in the upper 005 m were measured using athermopile (TCAV-L Campbell Scientific USA) Soilmoisture (y) (CS616 TDR Campbell Scientific USAcalibrated for study site soils) and soil temperatures (107BThermistors Campbell Scientific USA) were recorded ateach site at depths of 001 010 030 050 and 10 mbelow the LFHndashmineral soil interface Soil water matricpotential (cm) was measured in profile at 01 03 and05 m using soil tensiometers (SW-033 Soil MeasurementSystems USA) Precipitation was measured 30 m south ofthe tower over the period using a RM Young TippingBucket (52202 RM Young USA) with a series ofTru-Chek metric rain gauge (Edwards ManufacturingUSA) as backup for any missing data pointsA second (mobile) eddy covariance tower was used to

quantify ETB of the understory and ground surface(Eugster et al 1997 Brown et al 2010) Continuoushalf-hourly ET fluxes were measured at 4 m above theground and aspen understory but below the canopy forboth the 2005 and 2006 snow free seasons The mobiletower was situated 20 m north of the above canopy(permanent) tower for a 25 day period during green up in2005 (DOY 124 to 149) and 33 days in 2006 (DOY 123to 156) before being rotated to new locations every twoweeks from June 3 to October 5 The purpose was tocharacterize fluxes from varying sites within the catch-ment over both seasons at the temporal scale of synopticweather patterns (Eugster et al 1997) The mobile towerhad identical instrumentation and sampling protocolswith the exception of instrument height as that of thepermanent tower system

EC data filtering and gap filling A simple flux footprintparameterization (Kljun et al 2004) was used to estimatetemporally varying (30 min) contribution areas andassociated vegetation and topographical characteristics toeddy covariance instrumentation Analysis of the abovecanopy flux source indicates that the majority of the fluxesoriginated from the north-westerly upwind area (72 and73 of all half hour periods in the 2005 and 2006 studyperiods respectively) encompassing a predominantlyaspen-dominated area (Chasmer et al 2011) The 80probability density function of the total half-hourly fluxextended from between 160 m and 600 m during highlyconvective and near neutral atmospheric conditionsrespectively indicating that fluxes represent aspen-dominated upland (Chasmer et al 2011) The location ofthe maximum probability of flux determined from thefootprint probability density function (Xmax) averaged 78mfrom the tower above the canopy during both study seasons


The aspen area studied was homogeneous to a minimumdistance of 85 m from the tower in the westerly directionand a maximum distance of approximately 250 m in thenorthwesterly and southwesterly directions Results fromthe footprint model considering wind direction and the 80footprint probability density function indicate that fluxesoriginated from the aspen areas approximately 73 of thetime and were used to filter-out data that originated outsidethe area of the aspen stand of interest To quantify the extentof the area capture by the below canopy system a footprintanalysis was done using an analytical solution to thediffusion equations according to Schuepp et al (1990) Thisanalysis showed that the location of the maximumprobability of flux (Xmax) averaged 8 m from the belowcanopy tower during both study seasons To ensure that boththe above and below canopy sensors captured a represen-tative flux spectral density functions were computed usinghigh-frequency EC data to ensure sensor location andsampling intervals were sufficient to capture the full sizerange in eddies (Petrone et al 2001)Eddy covariance data were first corrected for density

effects (Webb et al 1980 Leuning and Judd 1996) andsensor separation (Blanford and Gay 1992 Leuning andJudd 1996) Energy balance closure (including canopyenergy storage) was calculated based on the slope of therelationship Q-QG and the total turbulent (sensible (QH)and latent (QE) energy) flux and canopy energy storage(Jt) (Blanken et al 1997)) For both years ECunderestimates turbulent fluxes by approximately 19for all measurement periods with no significant differ-ences associated with wind direction or between years inthe closure estimates Data were then filtered for periodswhen the atmosphere was highly stable using u (frictionvelocity)lt 02 m s1 (above canopy) and u lt029 m s1

(below canopy) determined using a 99 thresholdcondition for nighttime data (23h00 ndash 5h00) (Petroneet al 2001 Reichstein et al 2005 Papale et al 2006)Additional flux measurements were also removed duringperiods of rainfall early morning flush of CO2 from theecosystem and when rapid and unexpected changes instate variables occurred over half-hour intervals based on15 standard deviations from the mean for that time period(Restrepo and Arain 2005) Finally a de-spiking methodwas used (Papale et al 2006) where night time data wereselected based on 20 W m2 radiation thresholds and az-value of 55 Since this method is mainly used on high-frequency data rather than half-hourly averages a supple-mental de-spiking process was used where data exceeding 2standard deviations from the mean were removed Forreported cumulative flux data short half-hour breaks werefilled by linear interpolation while longer breaks (ie gt12half-hour periods) were filled using the mean diurnalvariation method with a 14-day moving window (Falgeet al 2001) Missing or rejected data occurred for a total of


Table I Summary table of catchment forest characteristics for theupland aspen-dominated forest situated at lsquoPond 40rsquo UtikumaRegion Alberta Canada LAI is leaf area index (m2 m2) PAI isplant area index (m2 m2) and breast height is defined as 13 mMeasurements consisted of three 15 m 15 m plots situated on

the south facing slope (SFS) of the catchment within thecalculated footprint of the EC tower

Max leaf area index LAI 145Max plant area index PAI 092Diameter at breast height DBH (cm) 1297Base diameter (cm) 1545Tree height (m) 1655


19 of all possible time periods during the study mostlyduring major precipitation events and nocturnal periods

Interception measurements Throughfall instrumentationused to estimate canopy wetness contributions to ET differedbetween 2005 and 2006 In 2005 three 15 01mV-shapedplastic troughs draining into tipping bucket rain gauges(52202 RM Young USA) were placed in the gap betweenthe adjacent tree canopies under the full canopy and close tothe trunk of a representative aspen The averages of thethrough fall measured in the gap canopy and near the trunkof the aspen were similar to longer troughs used thefollowing year which integrated spatial difference with inthe aspen stand In 2006 10 m 01 m U-shaped troughswere randomly placed within 30 to 200 m from permanenttower Each trough sampled the integrated precipitationthroughfall among several aspen trunks stems and differingcanopy structural (foliage leaf area) Three 10 m troughswere located in two separate aspen stands The throughfallcoefficient (p) was determined as the slope of the regressionbetween throughfall and cumulative precipitation (P) over3 month period during the two study seasons (Grelle et al1999) Water movement down stems (stemflow) wascollected in a split plastic hose wrapped in a cork-screwfashion three or four times and affixed to the main treestem using galvanized staples and silicone from a height of075 to 25 m of the trunk A tube trained into a collectioncontainer at the ground surface for measurement of volumecollected Average stem flow from three aspen treesassociated with the canopy and with diameter of 15cm(small) 25 cm (medium) and 35 cm (large) were collectedthrough the growing season (cf Crockford and Richardson2000 Toba and Ohta 2005) Storage capacity (S) wasdetermined as the x-intercept of the throughfall-precipitationregression (Grelle et al 1999) Evaporation of intercepted Pwas calculated using a modified Gash Model (Gash 1979McLaren et al 2008)

Mensuration and leaf area measurements Forest men-suration plots were established at the study site in order toassess variations in leaf area index (LAI) and other foreststand characteristics within the catchment during the peakgrowth season (Aug 17 ndash 22) (Chasmer et al 2008)Three 15 m 15 m plots were sampled for tree heightand live crown length using a vertex sonic hypsometerfor crown apices that were not distinct the average ofthree measurements was recorded and diameter at breastheight (DBH) (using a standard DBH caliper at 13 m)Crown diameter was measured along a northndashsouth andeastndashwest crown axes using a survey tape measure Alltrees with a DBH greater than 9 cm were included withinthe survey as trees with a DBH of less than 9 cm weredetermined to not constitute a significant element in theoverall canopy (Table I)


LAI was collected using digital hemisphericalphotography at five sites (one taken at the centre of theplot and four were located 113 m from the centre alongcardinal (N S E and W) directions determined using acompass bearing and measuring tape following Fluxnet-Canada and the Canadian Carbon Program protocol(Fluxnet-Canada 2003) within three representative plotsthroughout the study area (Figure 1) All photographswere taken at a height of 13 m above the ground duringeither diffuse daytime conditions or 30 min beforedawn or dusk and under-exposed by one f-stop to reducethe influence of sun brightness and apparent leafreduction within the photograph (Zhang et al 2005Chen et al 2006)

RESULTS

Environmental conditions

The upland aspen sitewas cooler andwetter on average in2005 than 2006 Total precipitation was 491 mm and 432mm for the two years respectively (Figure 2b) based on thehydrologic year Nov 1 ndash Oct 31 (Figure 2a) Average dailyair temperatures were 22 C (ranging between 348 Cand 191 C) in 2005 and 38 C (ranging between305 Cand 236 C) in 2006 (Figure 2a) However the higherprecipitation in 2005was largely due to snowfall During theJune toAugusts period when historically 50 of the annualprecipitation occurs (Environment Canada 2005) both 2005(231 mm) and 2006 (217 mm) were slightly below normal(248 mm) In both years July rainfall was slightly abovenormal with dry Augusts August rainfall in 2006 (28 mm)was only half that in 2005 (48 mm) and well below normalAugust rainfall of 65mmy in the rooting zone (0ndash50 cm) did not vary greatly

between years ranging between 02 to 03 m3m3 inresponse to precipitation events coupled with evaporativedemand (Figure 2b) However with drier end of summerconditions in 2006 soil moisture at the end of Septemberin the top 05 m dropped to about 011 m3m3 in 2006compared to on 02 m3m3 in 2005


Figure 2 Daily averaged a) air temperature (C) b) rooting zone SoilMoisture ((m3 m3) (0ndash50 cm) and daily precipitation (mm) c) abovecanopy ET (ETC) (mm) and d) within canopy ET (ETB) (mm) for the 2005and 2006 snow-free seasons Pond 40 Utikuma Region Study Area(URSA) Alberta Canada Transitions between defined growth periods aredenoted by vertical lines EG represents early green period G represents

green period and S represents senescence

Table II Total precipitation above canopy ET (ETC) andinterception ( standard deviations) for the study period in 2005

and 2006 (JD 187ndash291 and JD 123ndash283 in 2005 and 2006respectively) Utikuma Region Study Area Alberta Canada

Precipitation (mm) ETC (mm) Interception (mm)

2005 181 13 205 2 27 32006 331 13 442 3 40 2

S M BROWN ET AL

Seasonal partitioning of total stand ET ETC differssignificantly between the 2 years (MannndashWhitneyPlt 0001 U = 5695) and reaches maximum and mini-mum rates during the green and senescence periodsrespectively (Figure 2b) In both the 2005 and 2006snow-free seasons peaks in daily ETC are seen followingprecipitation events (Figure 2b) and the magnitude of theprecipitation events also affects the maximum rate ofdaily ETC across both seasons where ETC is largerfollowing larger events Following times of no precipi-tation ETC is reduced and is presumed to follow thecontrols of transpiration across the upper canopy Asprecipitation is intercepted by the canopy during greenperiods ETC increases suggesting the dominance ofdiurnal evaporation of intercepted water from leaf catchatop the canopy This is supported by the significant


canopy interception rates measured at this site Throughfallmeasurements here show that the aspen intercept as much as25of incoming precipitation for events greater than 5mm(Petrone andDevito unpublished data) and as much as 15on an annual basis (Table II) Precipitation intensity ndashduration and frequency analysis at this site suggests that thisaspen stand intercepted 49 mm and 40 mm during the 2005and 2006 study periods respectively (Table II) Resultsfrom the modified Gash model estimate that evaporation ofthis intercepted precipitation amounted to 43 mm and 35mm in 2005 and 2006 respectively As such peaks in ETCacross both seasons are observed following days whereprecipitation events occurred that were greater than 10 mmETB averages 11 mm d1 during the early green period

and increases at the point of understory vegetationemergence (Figure 2c) (eg DOY 139 in 2005 DOY141 in 2006) ETB also differed significantly between2005 and 2006 (MannndashWhitney Plt 0001 U = 1864)Green period ETB is approximately 45 of ETC overboth study seasons Senescence period ETB in 2005 isapproximately 60 of ETC rates with an average of04 mm d1 and is stable across the late seasonmeasurement period in which no precipitation events occur(Figure 2 3) In 2006 ETB during late green averages09 mm d1 peaking during the early portion of late greenwhen canopy development has just begun to cease

Ecosystem responses to atmospheric demand Figure 3shows average diurnal ET VPD and PAR for three periodsof the snow-free season of 2006 and the differences betweenabove and below canopy fluxes Both ETC and ETB peak inlate afternoon (1400 MST) during the early green periodand at midday (1200 MST) during green and senescenceperiods Diurnal variations of PAR and the phenology ofcanopy versus understory biomass are strongly coupledwithET fluxes (r2 = 095 ndash 099 plt 001) whilst the greatestPAR received by the understory occurs during the green upperiod before canopy leaves have reached their maximumleaf area Figure 3 also indicates that the observed middaypeaks in ET and PAR occur approximately 1 h before peakVPD during the Green period as demonstrated in otherstudies (Blanken et al 1997 Hogg et al 1997 Grelle et al1999) This synchronicity between ET and PAR and offsetwith VPD persists during the Early Green and Senescence


Figure 3 Above and below canopy period averaged diurnal comparisons of a) evapotranspiration (ET) (mm h1) b) vapour pressure deficits (VPD)(kPa) and c) photosynthetic active radiation (PAR) (W m2) Pond 40 Utikuma Region Study Area (URSA) Alberta Canada


periods however the duration of the lag to the peak in VPDis much shorter (Figure 3)The initial VPD for any given day defines the stomatal

response to atmospheric demand such that large atmo-spheric demand early in the day during the onset oftranspiration will cause quicker stomatal restriction tovapour loss under prolonged atmospheric demand andmoisture stress during the course of the day (McLaren et al2008) The relationship between initial VPD (VPDO)defined as the average VPD between 0600 and 0800MSTand ETC for both study periods shows a clustering of highmedium and low ETC as a result of variability in VPD(Figure 4) To illustrate these clustered relationships dataare grouped into three categories based on ranges of VPDwhereby (1) VPDO LOW is defined as VPDO lt005 kPa(2) VPDO MID gt005 kPa and lt015 kPa and (3) VPDO

HIGHgt 015 kPa Although there is some overlap inETndashVPD relationship among the three groups days thatbegin with a high initial VPD clearly result in a greaterprobability of larger rates for the same range in VPD(Figure 4) Coefficients of determination (r2) between low


medium and high VPD and ETC are (adjusted) r2 = 045(plt 001RMSE=0018 mm h1) 044 (plt 001 RMSE=0051 mm h1) and 071(RMSE= 0093 mm h1)indicating that VPD0 influences on ETC are most significantwhen VPD is high and other controlling mechanisms to ETCare not as important Maximum rates of ETC occur duringperiods when VPDO remained high throughout the day (peakdaily maximumsgt 03 mm h1) while days with lowVPDO

had much lower ETC (lt 015 mm h1) These distinctclustering patterns observed between the three categoriesshow that for aspen uplands the use of VPDO as a proxy forETC is an appropriate method for characterizing the dailycatchment ET rates Thus to investigatewhether changes in ycan produce diurnal shifts in maximum ET relative tomaximum VPD data were first grouped by daily initial VPD(0600 and 0800 MST) (VPDO)Given that understory vegetation receives a fraction of

the solar energy received by the canopy (Petrone et al2010) differences in VPD0 due to incident radiation andcanopy shadowing can have significant influences on ETespecially where understory and canopy LAI can vary by


Figure 4 Relationship between above canopy ET (ETC) (mm h1) andVPD (kPa) based on VPDO thresholds (binned in 001 kPa intervals) of a)VPDO LOW lt005 kPa b) VPDO MID gt005 kPa and lt015 kPa and c)VPDO HIGHgt 015 kPa Pond 40 Utikuma Region Study Area (URSA)Alberta Canada Solid lines represent best fit linear regressions through

the respective VPD groups

S M BROWN ET AL

up to 67 m2 m2 depending on canopy closure (Chasmeret al 2011) At the upland aspen site ETC is higher thanETB for a given VPD (Figure 5) Maximum ETB ratesreach 016 mm h1 at VPDO levels less than 01 kPa Anincrease in VPDO from below the canopy does not resultin an increased maximum daily ET above approximately01 kPa The same is observed for the relationshipbetween daily maximum ETC and VPDO with a levelingof ETC across a range of VPDO values above ~010 kPaClear clustering of daily maximum ETC is observed atVPDO levels less than 010 kPa However the range inETC at VPDO values between 01 and 04 kPa is greaterthan observations in ETB (Figure 5) The lower limit ofETC is approximately 015 mm h1 with the exception of

Figure 5 Relationship between daily maximum ET (mm h1) and VPDO

(kPa) for above and within canopy layers of the forest canopy duringdefined green periods of the 2005 and 2006 snow-free seasons Pond 40

Utikuma Region Study Area (URSA) Alberta Canada


minimal outliers which corresponds to the upperthreshold of the daily maximum ETB Maximum GC

rates of approximately 033 mm h1 are observed whenygt 020 m3 m3 (r2 = 061 plt 001 RMSE= 005 mmh1) (Figure 6) Whereas when y is limiting (ylt 020 m3

m3) ET only reaches a maximum of approximately019 mm h1 with the majority of the ET clustering below09 mm h1 (r2 = 053 plt 001 RMSE=003 mm h1)

Rooting zone soil moisture controls on ETC The relation-ships between rooting zone soil moisture (y) (average yover the top 50 cm) and soil water matric potential (cm)with ETC over four selected periods of growth through the2006 season shows the interconnected dependence of thevariables (Figure 7) During the 2006 study season earlygreen period (DOY 135ndash138) y depletion from 0255 to0235 m3 m3 occurs when cm and ETC reach dailymaximums of 82 kPa and 018 mm h1 respectively(Figure 7a) Similar trends are observed for the green periodwith y depletion of 0208 to 0187 m3 m3 coupled with apeak cm of120 kPa and maximum daily ETC rate of 031mm h1 (Figure 7b) Late green period shows a y depletionof 0203 to 0182 m3 m3 and a maximum daily cm of96kPa and ETC of 029 mm h1 (Figure 7c) Senescence in2006 again shows limited y depletion with reduced dailymaximum cm (~ 37 kPa) and limited ETC (012 mm h1)(Figure 7d) This results in limited y depletion within therooting zone

DISCUSSIONSeasonal partitioning of ET and implications for climate

change Aspen in the BP are already forced to make themost use of the limited precipitation the region receives

Figure 6 Relationship between above canopy evapotranspiration (ETC)(mm h1) and VPD (kPa) for periods during the 2005 and 2006 peakgrowth periods when Soil Moisture (Y becomes limited and approachesthe wilting point (02 m3 m3) (Saxton et al 1986) Pond 40 Utikuma

Region Study Area (URSA) Alberta Canada


Figure 7 Precipitation free 4-day periods for the a) early green (EG) b) green (G) c) late green (LG) and d) senescence (S) periods of rooting zone soilmoisture (y (m3 m3) rooting zone soil water matric potential (Ψm) (kPa) and above canopy ET (ETC) (mm) for the 2006 snow-free season Pond 40



(Devito et al 2005b) The clonal nature of aspen(Desrochers and Lieffers 2001) and sparse canopy(Sucoff 1982 Davison et al 1988) are understood tohave the most significant influence on the ability of thespecies to cope to with climatic stress LAI valuesobserved in this study are well within the range ofsimilarly aged sites studied elsewhere (cf Shepperd1993 Constabel and Lieffers 1996 Delong et al 1997)Canopy structure in general and LAI specifically affectsenvironmental factors such as radiation penetration windatmospheric humidity (or VPD) and soil moisture Inboth summers of 2005 and 2006 when aspen LAI was atits maximum a significant decrease within canopy VPDand PAR was observed which suggests less radiationpenetration but also less mixing Both the suppressedVPD and diminished PAR within the canopy translate tomuch a much lower ETB relative to ETC during the


Green period whereas a minimal difference is observedduring Early Green and Senescence periodsInterception is an important component in water

balance of forests as the amount of evaporation frominterception is a significant proportion of precipitation andcan be large in regions with net moisture deficit (Tobaand Ohta 2005 Asdak et al 1998 Scatena 1990Crockford and Richardson 2000) The importance ofinterception and evaporation in BP aspen forests isindicated in both the 2005 and 2006 snow-free seasonswhere peaks in daily ETC occurred following precipita-tion events In both years the magnitude of theprecipitation events also affected the maximum rate ofdaily ETC as ETC increased following larger events(Crockford and Richardson 2000) Following times of noprecipitation ETC is reduced and is presumed to followthe controls of transpiration across the upper canopy


S M BROWN ET AL

(Iacobelli and McCaughey 1993) However as precipita-tion is intercepted by the canopy during green periods ETCincreases suggesting the dominance of diurnal evapo-ration of intercepted water from leaf catch atop thecanopy in maintaining high ETC rates This is supportedby the significant canopy interception rates measured atthis site For example the difference between canopyand canopy + understory LAI varies by as much as 32along the soil evaporation transect and between aboveand below canopy eddy covariance sites (Figure 1)(Chasmer et al 2011) Thus as the BP region isexpected to warm 2 ndash 4 C and experience a 0 ndash 20change in precipitation over the next century (IPCC2007) larger VPDs and greater interception loss havethe potential to increase aspen stand ET and maysignificantly alter the watershed water balance

Atmospheric controls and ET VPD is expected to bethe most important environmental factor governing thetranspiration of boreal deciduous forests when y andradiation are not limiting (Blanken et al 1997 Hogget al 1997) Both ETC and ETB peak in late afternoonduring the green up period but peak at midday (1200MST) during both the green and senescence periodsMaximum and minimum atmospheric demand as indicat-ed by the maximum of VPD from above and below thecanopy can provide insight into the nature of diurnal ETpatterns at both scales Peak VPD corresponds with peakETC and ETB during the early green and senescenceperiods but is delayed compared to peak ET for the greenperiod A delay in peak ET between the canopy layers isbecause eventually the stomata of aspen react to VPDand soil moisture to keep water potential above a criticallevel in water scarce environments (Lieffers et al 2001)In this study the observed midday peaks in ET occurredjust before peak VPD and at approximately the same timeas PAR as demonstrated in other studies (Blanken et al1997 Hogg et al 1997 Grelle et al 1999) Since aspenhas the ability to survive in water stress periods theremust be a mechanism to respond conservatively to thelow soil moisture as well as high VPD which is commonin drought areas like the BP

Table III Soil texture properties for the rooting zone (0ndash50 cm) ofcores and taken during installation (May 2005) soil texture classific

Soil texture properties Sand ()Clay ()Silt ()

Soil moisture properties Wilting poField capaBulk densiSaturationSoil hydraAvailable


Interaction between atmospheric demand and y Athigh transpiration rates trees will experience water stressdue to decreasing water potentials from the roots to theleaves as a result of large VPD andor high productivity(Sperry et al 1998) The stomatal dynamics responsiblefor this transpiration response to VPD variations are afunction of the hydraulic response from the soil to leafwhere stomatal conductance will decrease as the soil dries(Meinzer and Grantz 1991 Iacobelli and McCaughey1993 Hogg and Hurdle 1997) Regulation of leaf waterpotential and transpiration by stomatal conductance inresponse to VPD maintains a homeostasis of water in theleaves for optimal balance between productivity and ysupply (Katul et al 2003 Ewers et al 2005) Howeverwhile a relationship should be observable betweenVPD andcanopy ET under differing y conditions VPD and ymay notalways be independent of each other (Kurpius et al 2003McLaren et al 2008) If VPD is high at the beginning of thediurnal transition cycle (early in the day) stomatal closurewill occur earlier in the day regardless of the magnitude ofVPD later in the cycle (Kurpius et al 2003)The distinct clustering patterns observed between the

three categories (HIGH MID and LOW) of VPDO andmaximum ETC show that for URSA the use of VPDO as aproxy for ETC is an appropriate method for characterizingthe daily catchment ET rates Thus to investigate whetherchanges in y can produce diurnal shifts in maximum ETrelative to maximum VPD data were first grouped by dailyinitial VPD (0600 and 0800 MST) (VPDO)Further given that the vegetation comprising the

understory will differ in the amount of solar radiationreceived and root structure it can be expected that thisrelationship will differ for ETC and ETB The lower limitof ETC is approximately 015 mm h1 with the exceptionof minimal outliers which corresponds to the upperthreshold of the daily maximum ETB This data suggeststhat VPD tends to constrain ET within a much narrowerrange in the understory (Schwartz et al 2006) Despitethe controls of atmospheric demand on stomata openingthe root systems of the aspen are still able to maintaincritical hydraulic gradients from the soil to the leaf(Schwartz et al 2006) Thus it is expected that y will

this silty clay loam soil below the litter fall horizon based on soilations and soil moisture properties based on Saxton et al 1986

70280650

int (m3 m3) 015city (m3 m3) 034ty (g cm3) 129(m3 m3) 051ulic conductivity (cm h1) 072water (m3 m3) 018



influence this relationship between VPD and ET and willlikely be the limiting factor on ET regardless ofatmospheric demand When data was binned into two ycategories ygt 020 m3 m3 and ylt 020 m3 m3roughly corresponding to the wilting point for aspen inthis type of soil (Table III Saxton et al 1986) it can beobserved that ET rates are limited when y approaches thewilting point of the soil (015 m3 m3) regardless of VPD(Saxton et al 1986 Oren et al 1998) During periodswhen y is not limited in the rooting zone ET rates can bemaintained at higher levels even at higher VPDs(Meinzer and Grantz 1991) Further the more gradualoverall decrease in Gc with increasing VPD (Figure 6)even under the drier soil conditions is indicative of thedeeper rooting clonal nature of the aspen stand as theyare collectively capable of progressively drawing waterfrom deeper layers (Breacuteda et al 1995 Oren et al 1998)

Interaction of y and c within the rooting zone and theinfluence on ETC The existence of many plant species isdriven by water availability as this is the key factor inlimiting carbon dioxide fixation and subsequent growth(Horton andHart 1998) This is especially important in sub-humid and arid regions where there are many periods ofwater shortage Therefore in a sub-humid region like theBP it is important to understand where the aspen standsclone is obtaining water to drive ET Under most conditionswithin their range aspen stands are characterized by anextensive clonal root system (Barnes 1966 DesRochersand Lieffers 2001) that could increase the ability of aspen tocope with climatic stress (DesRochers and Lieffers 2001)Thus despite having a relatively high transpiration rate(averaging as high as 35 mmd in this study) aspen mightbe able to succeed on dry sites as a result of its deep andextensive root system (Kemperman and Barnes 1976Sucoff 1982 Strong and La Roi 1983)One of the important factors that could affect the water

balance associated with an extensive clonal root system ishydraulic redistribution (Burgess et al 1998) Duringhydraulic redistribution roots take up water from deepsoil layers or laterally from adjacent wet areas during lowET (at night) and distribute it through the root system andrelease (bleed) the water in areas where soil waterpotentials are much lower than root water potentials(Horton and Hart 1998) Hydraulic lift and hydraulicallyredistributed water have positive implications not only forthose species that are capable of doing so but also forassociated (understory) vegetation (Horton and Hart1998) Thus forest stands comprised of trees capable ofhydraulic redistribution have the potential to maintain ahigher stand transpiration rate and have greater successduring periods of water shortage (Horton and Hart 1998)Studies (cf McLaren et al 2008) have suggested that thediurnal increases in y are evidence of hydraulic lift After


calibration and temperature corrections the error in TDRdata reported here are approximately 0005 While thedownward trend in y over the 3 day period does provideevidence of general decline in rooting zone y the diurnalpatterns in cm provide a clearer indication of theoccurrence of hydraulic lift occurring in response to ETdemand This is especially evident when comparing greenand senescence period data where ET and y diurnalvariation and maximums are decreased but the amplitudeof the daily changes in y remains basically the sameA delay exists in the response of y to morning

decreases in cm throughout the snow-free season of2006 As evaporative demand begins during earlymorning hours a decrease in cm is observed along withET However over all periods the y response is delayeduntil ETC reaches peak daily values at which time cm

increases until daily ETC is limited and a lsquobleeding effectrsquoof soil water demand permits the capillary rise andmoisture release from the roots to occur causing y toincrease gradually (McLaren et al 2008) In our studythese subtle changes in y occur throughout the snow-freeseasons of 2006 but are most pronounced during peakgrowth when water demand from the aspen roots is at itsgreatest (Caldwell et al 1998) While during senescenceperiods when root water uptake is all but removed alimited response is observed in y within the rooting zone

CONCLUSIONS

Understanding the controls on the partitioning of ET withinaspen forests of the BP are essential to interpreting theinfluences of climate change on stand water balances andhow these stands may influence the hydrology of adjacentwetland areas in this region Strong seasonality is observedin ETC across both snow-free seasons with maximum dailyaverages in ETC of 308 and 345 mm d1 occurring duringthe green periods of 2005 and 2006 respectively ETB canbe an important contribution of overall canopy ET althoughETB varies seasonally largely a function of the ability of theaspen canopy to permit radiation to reach the understory andthe aspen root systems influence on the availability of rootzone y For the 2005 and 2006 snow-free seasonsunderstory ET was the dominant flux during the early greenperiod which was a direct result of the timing of understorycanopy shrub layer development For the remainder of thegrowth periods ETB averaged 32 of ETC Further theevaporation of intercepted water comprises a significantportion of the daily canopy ET as a result of the largeatmospheric demand for water vapour in this climateSimilar responses to VPD are observed in both the

above and within canopy ET fluxes However the relativeincrease in ET for a given increase in VPD is greater fromthe understory especially during the green period which


S M BROWN ET AL

is a function of access to available rooting zone y and mutedfluctuations in VPD within the canopy microclimateFurther in all levels of the canopy y exerts the ultimatecontrol on ET That is regardless of the degree ofatmospheric demand ET can be maintained at higher levelsif there is sufficient y in the rooting zone Finally soiltension and moisture data within the rooting zone suggestthat some form of hydraulic redistribution is occurring inaddition to capillary recharge from below at nighttransporting soil moisture from wetter areas to meet thedemands of the transpiring aspen canopy When aspentranspiration decreases this redistributed moisture becomesavailable within the rooting zone for the within canopyvegetation and mitigating the daily decline in soil moisture

ACKNOWLEDGEMENTS

The authors would like to thank the HEAD2-NSERCCollaborative Research and Development grant CanadianOilsands Network for Research and Development ForestProducers Association of Canada Ducks UnlimitedAlberta Pacific Forest Products and Tolko for fundingfor this research We would also like to thank theanonymous reviewers for their insightful comments

REFERENCES

Ahrends B Penne C 2010 Modeling the impact of canopy structure onthe spatial variability of net forest precipitation and interception loss inscots pine stands The Open Geography Journal 3 115ndash124

Asdak C Jarvis PG Van Gardingen P Frazer A 1998 Rainfallinterception loss in unlogged and logged forest areas of centralKalimantan Indonesia Journal of Hydrology 206 237ndash244

Barbour MM Hunt JE Walcroft AS Rogers GND McSeveny TMWhitehead D 2005 Components of ecosystem evaporation in atemperature coniferous rainforest with canopy transpiration scaledusing sapwood density New Pytologist 165 549ndash558

Barnes BV 1966 The clonal growth habitat of American Aspens Ecology47 439ndash447

Blanford JH Gay LW 1992 Tests of a robust eddy correlation system forsensible heat flux Theoretical and Applied Climatology 46 53ndash60

Blanken PD Black TA Yang PC Neumann HH Staebler R Nesic Z denHartog G Novak MD Lee X 1997 The energy balance and canopyconductance of a boreal aspen forest partitioning overstory and understorycomponents Journal of Geophysical Research 102 915ndash927

Blanken PD Black TA Neumann HH den Hartog G Yang PC Nesic ZStaebler R Chen W Novak MD 1998 Turbulent flux measurementsabove and below the overstory of a boreal aspen forest Boundary-LayerMeteorology 89(1) 109ndash140

Blanken PD Black TA Neumann HH den Hartog G Yang PC Nesic ZLee X 2001 The seasonal water and energy exchange above and withina boreal aspen forest Journal of Hydrology 245 118ndash136

Blanken PD Black TA 2004 Canopy conductance of a boreal aspenforest Hydrological Processes 18(9) 1561ndash1578

Bothe RA Abraham C 1993 Evaporation and evapotranspiration inAlberta 1986ndash1992 Addendum Water Resources Services AlbertaEnvironmental Protection Edmonton Canada

Bovard BD Curtis PS Vogel CS Su HB Schmid HP 2005Environmental controls on sap flow in a northern hardwood forestTree Physiology 25 31ndash38


Breacuteda N Granier A Barataud F Moyne C 1995 Soil water dynamics inan oak stand I Soil moisture water potentials and water uptake byroots Plant and Soil 172 17ndash27

Brown SM Petrone RM Mendoza C Devito KJ 2010 Surfacevegetation controls on evapotranspiration from a sub-humid WesternBoreal Plain wetland Hydrological Processes 24 1072ndash1085

Burgess SO Adams MA Turner NC Ong CK 1998 The redistribution ofsoil water by tree root systems Oecologia 115 306ndash311

Caldwell MM Dawson TE Richard JH 1998 Hydraulic liftConsequences of water efflux from the roots of plants Oecologia113 306ndash311

Chasmer L Hopkinson C Treitz P McCaughey H Barr A Black A2008 A lidar-based hierarchical approach for assessing MODIS fPARRemote Sensing of Environment 112 4344ndash4357

Chasmer L Kljun N Hopkinson C Brown S Milne T Giroux K Barr ADevito K Creed I Petrone 2011 Characterizing vegetation structuraland topographic characteristics sampled by eddy covariance within twomature aspen stands using lidar and a flux footprint model Scaling toMODIS Journal of Geophysical Research 116 G02026 doi1010292010JG001567

Chen JM Franklin JF Spies T 1993 Contrasting microclimatesamong clearcut edge and interior of old-growth Douglas-fir forestAgricultural and Forest Meteorology 63(3ndash4) 219ndash237

Chen JM Blanken TA Black M Guilbeault M Chen S 1997 Radiationregime and canopy architecture of a boreal aspen forest Agriculturaland Forest Meteorology 86 107ndash125

Chen JM Govind A Sonnentag O Zhang Y Barr A Amiro B 2006 Leafarea index measurements at Fluxnet-Canada forest sites Agriculturaland Forest Meteorology 140 257ndash268

Cinnirella S Magnani F Saracino A Borghetti M 2002 Response of amature Pinus laricio plantation to a three-year restriction of watersupply structural and function acclimation to drought Tree Physiology22 21ndash30

Constabel AJ Lieffers VJ 1996 Seasonal patterns of light transmissionthrough boreal mixedwood canopies Canadian Journal of ForestResearch 26 1008ndash1014

Crockford RH Richardson DP 2000 Partitioning of rainfall intothroughfall stemflow and interception effect of forest type groundcover and climate Hydrological Processes 14(16ndash17) 2903ndash2920doi1010021099-1085

Dang Q-L Margolis HA Coyea MR Sy M Collatz GJ 1997Regulation of branch-level gas exchange of boreal trees Roles ofshoot water potential and vapour pressure difference Tree Physiology17 521ndash535

Davison RW Atkins RC Fry RD Racey GD Weingartner DH 1988 Asilvicultural guide for the poplar working group in Ontario Ont MinNat Resour Toronto Ont Sci Tech Ser Vol 5

DeLong HB Lieffers VJ Blenis PV 1997 Microsite effects on first yearestablishment and over-winter survival of white spruce in aspen-dominated boreal mixedwoods Canadian Journal of Forest Research27 1452ndash1457

DesRochers A Lieffers VJ 2001 Coarse root system of mature aspen(Populus tremuloides) in declining stands in Alberta Canada Journalof Vegetation Science 12 355ndash360

Devito K Creed I Gan T Mendoza C Petrone R Silins U Smerdon B2005a A framework for broad scale classification of hydrologicresponse units on the Boreal Plains Is topography the last thing to thinkof Hydrological Processes 19(8) 1705ndash1714

Devito KJ Creed IF Fraser CJD 2005b Controls on runoff from apartially harvested aspen-forested headwater catchment Boreal PlainCanada Hydrological Processes 19(1) 3ndash25

Environment Canada 2005 Climate Data Online Slave Lake AlbertahttpwwwclimateweatherofficeecgccaclimateDatacanada_ehtml

Eugster W McFadden JP Chapin III FS 1997 A comparative approachto regional variation in surface fluxes using eddy correlation towersBoundary Layer Meteorology 85 293ndash307

Ewers BE Gower ST Bond-Lamberty B Wang CK 2005 Effects ofstand age and tree species on canopy transpiration and averagestomatal conductance of boreal forests Plant Cell amp Environment28 660ndash678

Falge E Baldocchi D Olson RJ Anthoni P Aubinet M Bernhofer CBurba G Ceulemans R Clement R Dolman H Granier A Gross P



Grunwald T Hollinger D Jensen N-O Katul G Keronen P KowalskiA Ta Lai C Law BE Meyers T Moncrieff J Moors E Munger JWPilegaard K Rannik U Rebmann C Suyker A Tenhunen J Tu KVerma S Vesala T Wilson K Wofsy S 2001 Gap filling strategies fordefensible annual sums of net ecosystem exchange Agricultural andForest Meteorology 107 43ndash69

Fenton MM Paulen RC Pawlowicz JG 2003 Surficial geology of theLubicon Lake area Alberta (NTS 84BSW) Alberta GeologicalSurvey

Fluxnet-Canada 2003 Fluxnet-Canada measurement protocols WorkingDraft Version 13 Laval Quebec Fluxnet-Canada Network Manage-ment Office Available online http wwwfluxnet-canadaca

Gash JHC 1979 An analytical model of rainfall interception in forestsQuarterly Journal of the Royal Meteorological Society 105 43ndash55

Grelle A Lindroth A Moumllder M 1999 Seasonal variation of borealforest surface conductance and evaporation Agricultural and ForestMeteorology 98ndash99 563ndash578

Hogg EH Hurdle PA 1997 Sap flow in trembling aspen implications forstomatal responses to vapour pressure deficits Tree Physiology 17501ndash509

Hogg EH Black TA den Hartog G 1997 A comparison of sap flow andeddy fluxes of water vapour from a boreal deciduous forest Journal ofGeophysical Research-Atmospheres 102 28929ndash28937

Hogg EH Saugier B Pontailler JY Black TA Chen W Hurdle PAWu A 2000 Reponses of trembling aspen and hazelnut to vaporpressure deficit in a boreal deciduous forest Tree Physiology20 725ndash734

Horton JL Hart SC 1998 Hydraulic lift A potentially importantecosystem process Trends in Ecology amp Evolution 13 232ndash235

Iacobelli A McCaughey JH 1993 Stomatal conductance in a northerntemperate deciduous forest temporal and spatial patterns CanadianJournal of Forest Research 2 245ndash252

IPCC 2007 Contribution of Working Group I to the Fourth AssessmentReport of the Intergovernmental Panel on Climate Change 2007Solomon S Qin D Manning M Chen Z Marquis M Averyt KBTignor M Miller HL (eds) Cambridge University Press CambridgeUnited Kingdom and New York NY USA

Irvine J Law BE Anthoni PM Meinzer FC 2002 Water limitations tocarbon exchange in old-growth and young ponderosa pine stands TreePhysiology 22 189ndash196

Kaimal JC Finnigan J 1994 Atmospheric Boundary Layer Flows TheirStructure and Measurement Oxford University Press New York pp255ndash261

Katul G Leuning R Oren R 2003 Relationship between plant hydraulicand biochemical properties derived from a steady-state coupled waterand carbon transport model Plant Cell amp Environment 26 339ndash350

Kemperman JA Barnes BV 1976 Clone size in American aspenCanadian Journal of Botany 54 2603ndash2607

Kljun N Calanca P Rotach M Schmid H 2004 A simple parameter-ization for flux footprint predictions Boundary Layer Meteorology 112503ndash523 DOI 101023BBOUN00000306537103196

Kurpius MR Panek JA Nikolov NT McKay M Goldstein AH 2003Partitioning of water flux in a Sierra Nevada ponderosa pine plantationAgricultural and Forest Meteorology 117 173ndash192

Leuning R Judd MJ 1996 The relative merits of open and closed pathanalysers for measurement of eddy fluxes Global Change Biology 2241ndash253

Lieffers VJ Landhausser SM Hogg EH 2001 Is the wide distribution ofaspen a result of its stress tolerance In Shepperd WD Binkley DBartos DL Stohlgren TJ Eskew LG comps Sustaining aspen inwestern Landscapes symposium proceedings 2000 June 13ndash15 GrandJunction CO

Marshall IB Schut P Ballard M (compilers) 1999 CanadianEcodistrict Climate Normals for Canada 1961ndash1990 A nationalecological framework for Canada Attribute Data EnvironmentalQuality Branch Ecosystems Science Directorate EnvironmentCanada and Research Branch Agriculture and Agri-Food CanadaOttawaHull

McLaren JD Arain MA Khomik M Peichl M Brodeur J 2008 Waterflux components and soil water-atmospheric controls in a temperatepine forest growing in a well-drained sandy soil Journal ofGeophysical Research 113 DOI 1010292007JG000653


Meinzer FC Grantz DA 1991 Coordination of stomatal hydraulic andcanopy boundary layer properties do stomata balance conductances bymeasuring transpiration Physiologia Plantarum 83 324ndash329

Ni-Meister W Gao H 2011 Assessing the impacts of vegetationheterogeneity on energy fluxes and snowmelt in boreal forests Journalof Plant Ecology 4 DOI 101093jpertr004

Oren R Ewers BE Todd P Phillips N Katul G 1998 Water balancedelineates the soil layer in which moisture affects canopy conductanceEcological Applications 8(4) 990ndash1002

Papale D Reichstein M Aubinet M Canfora E Bernhofer C Kutsch WLongdoz B Rambal S Valentini R Vesala T Yakir D 2006 Towardsa standardized processing of Net Ecosystem Exchange measured witheddy covariance technique algorithms and uncertainty estimationBiogeosciences 3 571ndash583

Peterson EB Peterson NM 1992 Ecology management and use of aspenand balsam poplar in the Prairie Provinces Canada Forestry CanadaNorthern Forestry Centre Special Report 1

Petrone RM Waddington JM Price JS 2001 Ecosystem scaleevapotranspiration and net CO2 exchange from a restored peatlandHydrological Processes 15 2839ndash2845

Petrone RM Silins U Devito KJ 2007 Dynamics of evapotranspirationfrom a riparian pond complex in the Western Boreal Forest AlbertaCanada Hydrological Processes 21(11) 1391ndash1401

Petrone RM Solondz DS Macrae ML Gignac LD Devito KJ 2010Microtopographical and canopy cover controls on moss carbon dioxideexchange in a western Boreal Plain peatland Ecohydrology DOI101002eco139

Phillips N Oren R 2001 Intra- and inter-annual variation in transpirationof a pine forest Ecological Applications 11 385ndash396

Redding TE Devito KJ 2006 Particle densities of wetlands soils innorthern Alberta Canadian Journal of Soil Science 86 57ndash60

Redding TE Devito KJ 2008 Lateral flow thresholds for aspen forestedhillslopes on the Western Boreal Plain Alberta Canada HydrologicalProcesses DOI 101002hyp

Reichstein M Katterer T Andren O Ciais P Schulze ED Cramer WPapale D Valentini R 2005 Does temperature sensitivity ofdecomposition vary with soil organic matter quality BiogeosciencesDiscussions 2 737ndash747

Restrepo NC Arain MA 2005 Energy and water exchanges from atemperate pine plantation forest Hydrological Processes 19 27ndash49

Roberts J 2000 The influence of physical and physiological character-istics of vegetation on their hydrological response HydrologicalProcesses 19 2885ndash2901

Saxton KE Rawls WJ Romberger JS Papendick RI 1986 Estimatinggeneralized soil-water characteristics from texture Soil Science Societyof America Journal 50(4) 1031ndash1036

Scatena FN 1990 Watershed scale rainfall interception on two forestedwatersheds in the Liuquillo Mountains of Puerto Rico Journal ofHydrology 113 89ndash102

Schafer KVR Oren R Lai CT Katul GG 2002 Hydrologicbalance in an intact temperate forest ecosystem under ambientand elevated atmospheric CO2 concentration Global ChangeBiology 8 895ndash911

Schuepp PH Leclerc MY MacPherson JI Desjardins RL 1990 Footprintprediction of scalar fluxes from analytical solutions of the diffusionequation Bound-Layer Meteorology 50 355ndash374

Schwartz MD Ahas R Anto A 2006 Onset of spring starting earlieracross the Northern Hemisphere Global Change Biology 12 343ndash351

Shepperd WD 1993 Initial growth development and clonal dynamics ofregenerated aspen in the Rocky Mountains USDA Forest Service ResPap RM-312

Soil Classification Working Group 1998 The Canadian System of SoilClassification 3rd edn Agriculture and Agri-Food Canada OttawaON

Sperry JS Adler FR Campbell GS Comstock JC 1998 Limitation ofplant water use by rhizosphere and xylem conductances results from amodel Plant Cell amp Environment 21 347ndash359

Strong WL La Roi GH 1983 Root-system morphology of commonboreal forest trees in Alberta Canada Canadian Journal of ForestResearch 13 1164ndash1173

Sucoff E 1982 Water relations of aspen Univ Minn Agri Exp Sch StPual Minn Tech Bull 338


S M BROWN ET AL

Toba T Ohta T 2005 Factors affecting rainfall interception determinedby a forest simulator and numerical model Hydrological Processes22 2634ndash2643

Unsworth MH Phillips N Link T 2004 Components and controls ofwater flux in an old-growth Douglas-fir-western hemlock ecosystemEcosystems 7 468ndash481

Vogwill R 1978 Hydrogeology of the Lesser Slave Lake area AlbertaEdmonton AB Alberta Research Council

Webb EK Pearman GI Leuning R 1980 Correction for fluxmeasurements for density effects due to heat and water vapourtransfer Quarterley Journal of the Royal Meteorological Society106 85ndash100


Woo MK Winter TC 1993 The role of permafrost and seasonal frost inthe hydrology of northern wetlands in North America Journal ofHydrology 141(1ndash4) 5ndash31

Wullschleger SD Hanson PJ 2006 Sensitivity of canopy transpiration toaltered precipitation in an upland oak forest evidence from a long-termfield manipulation study Global Change Biology 12 97ndash109

Wullschleger SD Hanson PJ Tschaplinski TJ 1998 Whole-plant waterflux in understory red maple exposed to altered precipitation regimesTree Physiology 18 71ndash79

Zhang Y Chen J Miller J 2005 Determining digital hemisphericalphotograph exposure for leaf area index estimation Agricultural andForest Meteorology 133 166ndash181


S M BROWN ET AL

forest at the clear-cut edge and within a mature Douglas firforest sites Gradients between the factors controlling ET cansignificantly affect ET fluxes especially as increasinglylarger areas are harvested or altered in some way Forexample ETC is linearly controlled byVPD until a thresholdis reached (Hogg and Hurdle 1997 Hogg et al 1997Wullschleger et al 1998 Hogg et al 2000 Kurpius et al2003 Bovard et al 2005) and soil moisture (Wullschlegeret al 1998 Cinnirella et al 2002) However while negativefeedbacks occur between VPD and canopy conductance(GC) Stomatal conductance is often unaffected by soilmoisture until a deficit occurs at which point transpiration islimited as trees close stomata in an effort to conserve water(Wullschleger et al 1998 Phillips and Oren 2001Cinnirella et al 2002 Irvine et al 2002) The relationshipbetween stomatal conductance and soil moisture varies fromsite to site andwithin a stands due to differences in vegetationspecies type structure and soil texture as different plantsrespond differently to given soil water states (Roberts 2000)Barbour et al (2005) also found that canopy wetness wasstrongly correlated with ETC Within closed canopies waterexchanges between the understory and atmosphere wererelatively small due to reduced canopy wetness and incidentradiation (Wullschleger et al 1998) whereas the oppositewas true in open canopies which were also prone to spatiallyvariable water interception (Unsworth et al 2004 Barbouret al 2005) Differences in radiation penetration throughcanopies due to differences in leaf area also influence thegrowth and distribution of understory species such as Rosaacicularis and Viburnum edule which are prevalent in aspenstands Proliferation of understory vegetation includingaspen seedlings which are established through a clonal rootsystem also increase water exchanges between the ecosystemand the atmosphere (Peterson and Peterson 1992 Blankenet al 1998)Trembling aspen may also exacerbate limitations to water

balance in theBP because full forests exists over large areas ofwell-drained soils of a region already with a net moisturedeficit (ETgtP) Therefore understanding the spatial distri-bution and heterogeneity of aspen forests how a forest ofaspen is maintained and impacts to ET and water resources isan important question within this region Currently little isknown about the sensitivity of ET to variability in thecontributions to fluxes within the canopy and the understoryin aspen forests of theBP yet the implications for this researchcan be significant given frequent drought periods andinfrequent lsquowetrsquo periods that occur only about every 20 years(Devito et al 2005a) Understanding the natural distributionof ETC and ETB provide important baseline for interpretingthe influences of climate change on forest-stand balances andhow natural (ie fire) or anthropogenic disturbance of forestcanopy influence water resource issues in the regionThe objective of this study is to characterize and quantify the

partitioning of ET fluxes within an aspen-dominated upland


within the BP of Canadarsquos Western Boreal Forest This studyis part of a larger project examining water sink and sourcedynamics of aspenndashpeatlandndashpond ecosystems typical of theBP a region that persists with a net moisture deficit climatecycles of infrequent wet periods deep soils and large seasonalpatterns in precipitation and antecedent moisture conditionsWe examine the interaction of vegetation structure meteoro-logical and soil controls on the ET ETC and ETB within thisclimate context It is hypothesized that the evaporation ofintercepted water will comprise a significant portion of thedaily canopy ET due to the drier atmosphere Further thevegetation response to atmospheric demand will ultimately bemediated by hydraulic redistribution of soil moisture fromwetareas to the drier aspen stands to meet atmospheric demand Itis hypothesized that driven by diurnal cycles in ET the clonalroots of the aspen standwill bringmoisture to the rooting zonewhich will be evident by examining the interactions of soilsuction soil moisture and stand-scale ET

STUDY SITE

Utikuma region study area

The study site is located within the Utikuma RegionStudy Area (URSA) 320 km north of Edmonton AlbertaCanada (56o4` N 115o28` W Figure 1) URSA is situatedapproximately 150 km south of the discontinuouspermafrost zone (Woo and Winter 1993) within the BPsregion of the Western Boreal Forest The climate withinURSA is characterized by cold winters (146 C) andwarm summers (156 C) (Environment Canada 2005)Annual ET averages 517 mm (Bothe and Abraham 1993)which is slightly higher than the average annual precipita-tion of 481 mm at URSA (Environment Canada 2005)Snowfall in this area averages less than 150 mm yr1 andrepresents less than 30 of the total annual precipitation forthe region (Marshall et al 1999 Devito et al 2005a)Although a water deficit exists during most years in thesub-humidBP annual precipitation surpassesETevery10ndash15years with significant runoff only every 20 years or more(Devito et al 2005a) Approximately 50 of annualprecipitation occurs from June through August coincident withthe period of maximum evaporative demand and is precededand followed by a relatively dry spring and fall periodsrespectively (Devito et al 2005a Petrone et al 2007)URSA is characterized by the presence of rolling

moraines of low relief low lying clay plains and coarse-textured outwash plains where glacial overburdens rangefrom 20 to 240 m in thickness and overlie the UpperCretaceous Smoky Shale Group (Vogwill 1978) Thisstudy was conducted on the moraine portion of theregion (Figure 1) The moraine landscape was chosen asthese systems are dominated by the presence of verticalexchanges of water (Devito et al 2005a Redding and












S M BROWN ET AL


















RESULTS










2005 181 13 205 2 27 32006 331 13 442 3 40 2

S M BROWN ET AL



















S M BROWN ET AL























S M BROWN ET AL











70280650











CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL





















S M BROWN ET AL


















RESULTS










2005 181 13 205 2 27 32006 331 13 442 3 40 2

S M BROWN ET AL



















S M BROWN ET AL























S M BROWN ET AL











70280650











CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL




















RESULTS










2005 181 13 205 2 27 32006 331 13 442 3 40 2

S M BROWN ET AL



















S M BROWN ET AL























S M BROWN ET AL











70280650











CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL
















2005 181 13 205 2 27 32006 331 13 442 3 40 2

S M BROWN ET AL



















S M BROWN ET AL























S M BROWN ET AL











70280650











CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL























S M BROWN ET AL























S M BROWN ET AL











70280650











CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL



















S M BROWN ET AL











70280650











CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL



















CONCLUSIONS




S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL











S M BROWN ET AL


ACKNOWLEDGEMENTS


REFERENCES




















































































S M BROWN ET AL



























































S M BROWN ET AL











Atmospheric and soil moisture controls on ...

Documents

Transcript of Atmospheric and soil moisture controls on ...