High Vapor Pressure Deficit Decreases the Productivityand Water Use Efficiency of Rain‐Induced Pulsesin Semiarid EcosystemsMatthew C. Roby1,2 , Russell L. Scott2 , and David J. P. Moore1
1School of Natural Resources and the Environment, University of Arizona, Tucson, AZ, USA, 2Southwest WatershedResearch Center, USDA‐ARS, Tucson, AZ, USA
Abstract Intermittent rain events drive dynamic pulses of carbon and water exchange in many arid andsemiarid ecosystems. Although soil moisture is known to control these pulses, the effect of atmosphericdryness on pulses is not well documented. Here we hypothesized that vapor pressure deficit (VPD)modulates net ecosystem production (NEP) and ecosystem‐scale water use efficiency (WUE) during pulseevents due to its effects on canopy stomatal conductance and evapotranspiration. We quantifiedrelationships between VPD and carbon and water exchange during growing season rain events and testedtheir generality across four semiarid flux sites with varied vegetation in the southwest United States. Acrossgrassland, shrubland, and savanna sites, we found that high VPD during pulses suppressed ecosystemphotosynthesis and surface conductance to a greater degree than respiration or evapotranspiration,particularly when soil moisture was high. Thus, periods of high VPD were associated with a 13–64%reduction inNEP and an 11–25% decrease inWUE, relative to moderate VPD conditions. Sites dominated byshrubs with the C3 photosynthetic pathway were more sensitive to VPD than sites dominated by C4 grasses.We found that a 1 kPa increase in VPD reduced the average NEP of pulse events by 13–56%, whichillustrates the potential for projected increases in atmospheric demand to reduce the net productivity ofsemiarid ecosystems.
Plain Language Summary In many water‐limited regions, summer storms provide water thatdrives brief periods of high ecosystem activity (photosynthesis, respiration, and evapotranspiration) called“pulse events”. While many studies have focused on soil moisture as a control on pulse events, it is notwell known how changes in the air's evaporative strength, or dryness, influence patterns of carbon and waterexchange during pulses. We analyzed data from four semiarid sites in southern Arizona and found that airdryness modified the widely accepted pulse framework of carbon and water exchange for semiarid areas.Drier air led to larger reductions in photosynthesis than respiration and evapotranspiration, whichdecreased the overall productivity and water use efficiency of plants. These findings indicate the potential fordrier air in a warming climate to reduce the ability of dryland plants to use water. Because pulses areimportant for the carbon balance of these systems, drier during these critical pulse periods could reduce howmuch carbon global drylands remove from the atmosphere. Incorporating the effects of air dryness on pulsepatterns into models may help us better understand global change impacts on semiarid ecosystems.
1. Introduction
Episodic rain events drive pulses of carbon and water exchange in many arid and semiarid ecosystems(Huxman, Snyder, et al., 2004)—regions known to substantially impact interannual variation in the landcarbon sink (Ahlström et al., 2015; Poulter et al., 2014). Pulses are transient periods characterized by highrates of gross ecosystem photosynthesis (GEP), evapotranspiration (ET), and ecosystem respiration (Re) thatoccur when rain events temporarily mitigate water stress effects on plant and microbial activity (Huxman,Snyder, et al., 2004; Noy‐Meir, 1973; Schwinning et al., 2004). With high rates of biogeochemical cycling,pulses have a disproportionately large impact on annualGEP (Kannenberg et al., 2020), which is a key driverof variation in the net ecosystem production (NEP = GEP − Re) and ecosystem‐scale water use efficiency(WUE = GEP/ET) of semiarid regions (Baldocchi et al., 2018; Biederman et al., 2016; Jia et al., 2016; Liuet al., 2019; Scott et al., 2015). Therefore, it is necessary to examine controls on ecosystem carbon and waterfluxes during these important subannual periods (Barnes et al., 2016; Jenerette et al., 2012; Jung et al., 2017).
carbon and water exchange insemiarid ecosystems to includeatmospheric demand for moisture
• High atmospheric demand drovelarger reductions in photosynthesisthan respiration, which decreasedthe net productivity and water useefficiency of rain pulses
• Ecosystem photosynthesis was mostsensitive to high atmosphericdemand in sites dominated by C3shrubs when soil moisture was high
Supporting Information:• Supporting Information S1
Citation:Roby, M. C., Scott, R. L., & Moore,D. J. P. (2020). High vapor pressuredeficit decreases the productivity andwater use efficiency of rain‐inducedpulses in semiarid ecosystems. Journalof Geophysical Research: Biogeosciences,125, e2020JG005665. https://doi.org/10.1029/2020JG005665
Received 29 JAN 2020Accepted 29 JUL 2020Accepted article online 24 AUG 2020
Whereas dryland studies have focused on water availability (soil moisture supply) as a control on ecosystemfunctioning (Chen et al., 2009; Huxman, Cable, et al., 2004; Huxman, Snyder, et al., 2004; Tarin, Nolan,Eamus, et al., 2020; Tarin, Nolan, Medlyn, et al., 2020), syntheses spanning gradients in climate and plantfunctional type highlight the role of atmospheric demand for moisture (vapor pressure deficit; VPD) as animportant, and sometimes dominant, constraint on carbon and water exchange (Novick et al., 2016;Zhang et al., 2019). High VPD can intensify reductions in land carbon uptake during compound (air and soil)drought (Zhou et al., 2019). Because saturation vapor pressure increases nonlinearly with rising tempera-ture, climate warming is projected to cause widespread increases in VPD (Breshears et al., 2013; Ficklin &Novick, 2017; Seager et al., 2015). However, it is not well known how VPD influences pulse responses insemiarid regions.
Biotic and physical responses indicate how high VPD may influence the carbon and water dynamics ofpulses. In response to increasing VPD, plants decrease stomatal conductance to reduce water loss(Buckley, 2019; Collatz et al., 1991; Leuning, 1995). These plant responses can decrease ecosystem‐scalecanopy conductance, GEP, and ET (Anthoni et al., 1999; Ding et al., 2018; Grossiord et al., 2020; Novicket al., 2016; Sulman et al., 2016; Wharton et al., 2009; Zhang et al., 2019). Additionally, physical manifesta-tions of high VPD (warmer temperatures; increased evaporative demand) may influence ET by increasingrates of soil and interception evaporation and may stimulate temperature‐dependent autotrophic and het-erotrophic respiration during periods of sufficient water availability.
We present a modified conceptual model of the pulse response of ecosystem carbon and water exchange torainfall (Huxman, Cable, et al., 2004; Huxman, Snyder, et al., 2004; Yan et al., 2014) and extend it to repre-sent the hypothesized effect of VPD during pulses. We show the expected pattern of canopy‐scale stomatal
Figure 1. Conceptual response patterns for pulse events with moderate (solid) and high (dashed) vapor pressure deficit(VPD) assuming an inactive prepulse state due to soil water stress. Pulse patterns of (a) gross ecosystem photosynthesis(GEP), (b) canopy‐scale stomatal conductance (Gcanopy), (c) ecosystem respiration (Re), (d) evapotranspiration(ET), (e) net ecosystem production (NEP = GEP − Re), and (f) ecosystem water use efficiency (WUE= GEP/ET) expressedas a percentage of maximum rates as a function of days since a large rain event. Adapted from Huxman, Snyder,et al. (2004) to represent hypothesized effects of high VPD using functions described in Table S1 in the supportinginformation.
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 2 of 14
conductance (Gcanopy) alongside fluxes to infer how plant responses to VPD alter GEP and ET (via its tran-spiration component). Assuming a prepulse state of water stress, the pulse event begins when fluxes increasein response to water availability supplied by rainfall (Figure 1; solid lines). Whereas GEP and Gcanopy
increase gradually as plants upregulate in response to wetting of deeper soil layers (Reynolds et al., 2004;Scott et al., 2006), Re, which is dominated by soil respiration during pulses (Sponseller, 2007; Xu et al.,2004), peaks rapidly then declines as changes in shallow soil moisture regulate microbial activity (Barron‐Gafford et al., 2011; Conant et al., 2000; Jenerette et al., 2008; Roby et al., 2019). ET, which rapidly increasesdue to high rates of evaporation from soil and canopy surfaces, remains high as increasing transpiration(with associated increases in Gcanopy) offsets declines in evaporation linked to shallow soil water depletion.Root zone soil drying ultimately causes Gcanopy to decline, driving associated reductions in GEP and ET(dominated by transpiration later in the pulse). Absent VPD stress, these patterns yield two phases of NEPandWUE: (1) a period of negative NEP and lowWUE (caused by low GEP, high Re, and high ET dominatedby evaporation); followed by (2) a prolonged period of high NEP and high WUE (due to high GEP, low Re,and high ET dominated by transpiration).
We predict that physical and biotic responses to high VPD alter pulse patterns (Figure 1; dashed lines).Physical responses to high VPD dominate the early pulse period, during which increased temperature andevaporative demand promote higher evaporation rates from soil and canopy surfaces (Scott et al., 2006)and stimulate respiration processes (Cable et al., 2011; Lloyd & Taylor, 1994). These responses causeincreases in Re and ET, relative to moderate VPD conditions. As events progress, shallow soil dryingdecreases Re (Liu et al., 2009) and reduces the evaporative fraction of ET, permitting biotic processes to dom-inate the pulse response (Szutu & Papuga, 2019). Stomatal responses to high VPD, represented by reducedGcanopy, drive declines in GEP and the transpiration fraction of ET. We hypothesize that high VPD drivesphysical and stomata‐mediated responses that reduce GEP and have a minor impact on total ET, thusdecreasing the NEP and WUE of pulse events. To assess the generality of these responses in pulse‐drivensemiarid regions dominated by summer rainfall, we test these hypotheses with 49 site years of eddy covar-iance data from four sites with varied vegetation.
2. Methods2.1. Site Description
We used data from four AmeriFlux sites that share similar semiarid climates but differ in their vegetationtype and average productivity (Scott et al., 2015). Briefly, the sites have similar annual mean temperatureand total precipitation, about 60% of which falls from July–September during the North AmericanMonsoon (Table 1). Two sites are grasslands and the others are dominated by woody vegetation. On average,the most productive site is Santa Rita Grassland (Grassland‐1; AmeriFlux ID: US‐SRG; Scott, 2008), which isdominated by C4 perennial bunchgrasses (mostly Eragrostis lehmanniana). Kendall Grassland (Grassland‐2;AmeriFlux ID: US‐Wkg; Scott, 2004b) is also composed of the same C4 grasses. The Santa Rita MesquiteSavanna (savanna; AmeriFlux ID: US‐SRM; Scott, 2004a) is a former grassland composed mainly of peren-nial bunchgrasses (Eragrostis lehmanniana; various native species) that is now about 35% covered by velvetmesquite trees (Prosopis velutina). Lucky Hills Shrubland (shrubland; AmeriFlux ID: US‐Whs; Scott, 2007) ischaracterized by Chihuahuan desert shrubs (e.g., Parthenium incanum, Larrea tridentata, Acacia constricta)and is on average the least productive site. Grassland‐1 and the savanna are in the Santa Rita ExperimentalRange, whereas Grassland‐2 and the shrubland are located in the USDA‐ARS Walnut Gulch ExperimentalWatershed (details in Table 1).
2.2. Meteorology and Flux Data
Land‐air fluxes of CO2, water vapor, and energy were measured using the eddy covariance technique.Meteorological variables (including precipitation) were measured at the same locations as fluxes.Instrumentation and data processing have been described previously (Scott et al., 2015). Briefly, net ecosys-tem exchange of CO2 (NEE) was partitioned into gross ecosystem photosynthesis (GEP) and ecosystemrespiration (Re) using the nighttime approach (Reichstein et al., 2005). Here we focus on the net ecosystemproduction of CO2, defined as NEP = − NEE. To minimize correlation between GEP and Re, we analyzedhalf‐hourly and daily mean fluxes, time scales during which these fluxes are relatively decoupled
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 3 of 14
(Figure S2). We tested the partitioning method by comparing the pulse response of Re with independentmeasurements of soil respiration for available years at the savanna (Roby et al., 2019) and found that thedata sets yield similar patterns (data not shown). We calculated bulk surface conductance to water vapor(Gsfc; mol m−2 s−1) by inverting the Penman‐Monteith equation to serve as a proxy for canopy‐scalestomatal conductance (Gcanopy) (Monteith, 1965; Monteith & Unsworth, 2013). Unlike Gcanopy, Gsfc reflectsthe contribution of both stomatal and soil conductances (Shuttleworth & Wallace, 1985). However, thecontribution of soil conductance to Gsfc decreases as shallow soil moisture declines with time since rain.
2.3. Pulse Characterization
Pulse events were identified by examining growing season periods during which daily precipitation exceeded5 mm and was followed by at least 1 day of rain‐free weather. We defined the growing season as 1 July to 30September, during which much of the annual precipitation and 62–68% of annual GEP occurs at thesesites (Table 1). To better isolate the influence of VPD on pulse patterns, we excluded periods of light‐limitation, defined as data during which photosynthetic photon flux density (PPFD) was less than 1,000μmol m−2 s−1, based on the response of GEP to PPFD (Figure S3). Pulse events were analyzed using half‐hourly and daily mean data aggregated into 1‐day bins of time since rain. We tested for differences betweenmean daily fluxes using two‐sample t tests.
To maintain consistency with the conceptual hypothesis (Figure 1), which assumes an inactive prepulsestate caused by soil water stress (Huxman, Snyder, et al., 2004), Figure 2 depicts general pulse patterns fil-tered for events with dry antecedent soil moisture. Subsequent figures were not filtered for dry antecedentconditions. Because pulse events occur sequentially during the summer growing season and have variableantecedent soil water, our decision to not filter for prepulse dryness increases sample size and enables usto make more general inferences about how VPD impacts pulse responses in these systems across the grow-ing season.
Table 1Description of the Study Sites and Growing Season (July–September) Pulse Characteristics
Pulse events 129 193 161 184Avg. events per year 12 14 16 13Pulse length (days) 9 8 7 9Annual GEP(g C m−2 year−1)
173 356 499 273
Growing season GEP(g C m−2 year−1)
114 230 310 186
MODIS LAI averageannual max.
0.57 0.84 1.07 0.84
Woody cover (%) 40 35 11 3Grass/forb cover (%) 3 15 44 37θ depth (cm) 15 5–10 5–10 10Soil type Gravelly sandy loams Deep loamy sands Deep loamy sands Very gravelly, sandy to fine sandy, and clayey
loams
Note. Pulse events defined as periods of daily rainfall exceeding 5 mm followed by no precipitation for at least 1 day.MAT is mean air temperature;MAP is meanannual precipitation for 1971–2010. LAI is leaf area index. Pulse duration is the number of dry days between growing season rain events. Pulse VPD is theaverage vapor pressure deficit during a growing season pulse event. Annual GEP is the average annual gross ecosystem photosynthesis over the study period. θdepth refers to the measurement depth of volumetric soil moisture (m3 m−3).
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 4 of 14
2.4. Modeling the VPD Sensitivity of Pulse Responses
We used empirical models as an initial test of the hypothesis that biotic and physical pulse responses havedifferential sensitivity to VPD (Kimm et al., 2020). All models were fit to half‐hourly pulse data during highlight conditions (PPFD> 1,000 μmol m−2 s−1). To enable inferences about the relative importance of drivers,we applied the same multiple linear regression for each response term:
Y ¼ βint þ βVPDVPD þ βθθ þ βVPDVPD × θ (1)
Where Y is the half‐hourly flux (μmol CO2 m−2 s−1) or surface conductance (mol m−2 s−1) and VPD and θ
(volumetric soil water content; m3 m−3) are half‐hourly predictors. βint is the value of Y when all predictorsare zero. βVPD, βθ, and βVPD × θ are coefficients for the main effects (VPD, θ) and the VPD × θ interaction. Allregression variables were scaled from zero to one to enable us to compare the relative effects size of termswith different magnitude.
2.5. VPD and Soil Moisture Categorization
To isolate the effects of air and soil dryness, we defined moisture categories using frequency distributions ofVPD and soil moisture. First, half‐hourly VPD values <0.6 kPa were excluded (Novick et al., 2016). VPD wasdefined as “low” if <33rd percentile, “moderate” if between the 33rd and 66th percentiles, and “high” if>66th percentile. θ (hereafter soil moisture) was measured from 5–15 cm, depending on site data availabilityand rooting depth (Table 1). Low, moderate, and high θ groups were defined using the same percentileranges as VPD. We can examine the relative importance of changes in VPD and θ on ecosystem responses
Figure 2. General pulse pattern at the savanna averaged over 55 events with dry antecedent soil moisture. Panels showmean (lines) ±1 standard deviation (shading) of (a) gross ecosystem photosynthesis (GEP), (b) surface conductance(Gsfc), (c) ecosystem respiration (Re), (d) evapotranspiration (ET), (e) net ecosystem production (NEP), (f) water useefficiency (WUE), (g) vapor pressure deficit (VPD), and (h) volumetric soil moisture (θ).
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 5 of 14
because these variables have relatively low correlation at hourly todaily time scales (Novick et al., 2016) (Figure S4).
3. Results3.1. Pulse Climatology
Across sites, an average of 12–16 pulse events occurred each summergrowing season (July–September) and the average interstorm dura-tion was 7–9 days (Table 1). Productivity during pulse events was con-siderable, with 62–68% of annual GEP occurring during thesemonths. Annual and growing season GEP were highest atGrassland‐1 (rainiest site) and lowest at the shrubland (driest site).Mean annual temperature and pulse mean VPD were highest at thesavanna and lowest at Grassland‐2.
3.2. General Pulse Patterns at the Savanna
To be consistent with the dry prepulse soil water conditions assumed in Figure 1, we first show the pulseresponse at the savanna filtered for events with low prepulse soil moisture (θ < 20th percentile during theday before a rain event). The pattern averaged over 55 events showed high variation in fluxes and environ-mental drivers (Figure 2; see Figure S5 for a single pulse). Rain events caused increases in θ and declines inVPD, relative to prepulse values, whereas θ decreased and VPD increased as days since rain elapsed. Notably,VPD was highly variable during the pulse event, whereas θ became less variable with time. Rain eventscaused increases in fluxes and Gsfc, relative to prepulse values. Rapid increases in ET, likely due to high ratesof evaporation, outpaced GEP which drove initial declines inWUE. Whereas Re, ET, and Gsfc peaked within1 day since rain, GEP increased more gradually, causing NEP and WUE to peak 2–4 days after rain.
3.3. Drivers of Variation in Carbon and Water Fluxes at the Savanna
We used multiple linear regression to test if changes in VPD explained the high variability in savanna pulseresponses (Figure 2). βVPD was negative for all models, indicating that higher VPD reduced average pulseresponses (Table 2). Conversely, positive values for βθ indicate that higher θ increased pulse responses. Ashypothesized, increased VPD was associated with reductions in GEP but did not strongly impact Re, as indi-cated by the sign and relative magnitudes of βVPD and βθ. The βVPD × θ coefficient was significant for all terms(exceptGsfc), confirming the VPD × θ interaction. Results from empirical models were consistent with that ofmodels based on leaf‐level gas exchange theory (Text S1and Table S2). Output from theory‐based modelsindicated that increasing VPD had a negative effect onGsfc,GEP, and ET, whereas increasing θ had a positiveeffect (Figure S1 and Table S2).
3.4. High VPD Impacts on Pulse Events
We applied a binned‐averaging approach to better isolate the influence of VPD on pulse patterns. At thesavanna, high VPD during periods of moderate θ was associated with reductions in the average GEP, Gsfc,NEP, and WUE of pulse events (p < 0.05), relative to moderate VPD (Figure 3). Because high VPD did notstrongly impact Re or ET (p > 0.05), reductions in NEP and WUE were primarily driven by decreased GEP(p < 0.05), relative to moderate VPD. These general patterns were consistent across sites (Figures S6–S8).To ensure our results were robust to the binning approach, we repeated the analysis using five bins of equalwidth for both θ and VPD and found that VPD effects were not sensitive to the number of bins used(Figures S9–S13).
At all sites, reductions in the average NEP,WUE, and Gsfc of pulse events occurred under high VPD, relativeto moderate VPD (p < 0.05; except for Gsfc andWUE at the shrubland) (Figure 4). Notably, these reductionswere driven by decreased GEP at high VPD (p < 0.05 at all sites except Grassland‐1) instead of changes in Re
or ET. Reductions in Gsfc at high VPD are consistent with stomatal control of water loss, which may explainwhy ET was not greater under conditions of enhanced evaporative demand. Whereas we observed reduc-tions in pulse responses across soil moisture states (low, moderate, and high), adverse effects of high VPDwere often strongest for high θ (Figures S8–S14).
Note. βint is the value of Y when all predictors are zero. βVPD, βθ, and βVPD × θare coefficients from empirical models (Equation 1) fit to half‐hourly surfaceconductance (Gsfc), gross ecosystem photosynthesis (GEP), evapotranspiration(ET), and ecosystem respiration (Re). Explanatory variables are vapor pressuredeficit (VPD) and volumetric soil moisture (θ).*This indicates p < 0.05.
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 6 of 14
3.5. Impact of Air and Soil Water Deficits on Integrated Pulse Responses
To contextualize the influence of high VPD, we quantified the response of pulse integrated fluxes to airand soil water deficits. Compared to moderate conditions (defined as the average value of pulse responseswhen VPD and θ were both moderate), high VPD decreased the NEP (13–64%), WUE (11–25%), and Gsfc
(22–34%) of pulse events at all sites (Figure 5). Because high VPD had minor impacts on Re (−4% to 7%)and ET (−10% to 0.1%), reductions in NEP and WUE were driven by decreased GEP (7–27%). We foundlarger NEP losses at high VPD for the shrubland and savanna than in grasslands. The result that highVPD strongly reduced Gsfc but did not enhance pulse‐integrated ET is consistent with stomatal regulationof transpiration (a major fraction of ET). This contrasts with the consistent negative effects of soil waterdeficits on ET and Gsfc. Pulses with low θ and moderate VPD had less NEP (45–137%), WUE (10–21%),and GEP (37–62%), as well as decreased Gsfc (32–43%), Re (5–25%), and ET (25–35%) compared to eventswith moderate conditions.
Next we quantified the VPD sensitivity of carbon exchange to examine how rising trends in atmosphericdemand may alter semiarid ecosystem productivity. We determined this sensitivity by dividing the averagechange in flux by the average change in VPD between moderate and high VPD pulse events. A 1 kPa increasein VPD above moderate conditions was associated with a 13–56% loss in the average NEP (0.58–2.00 μmolCO2 m
−2 s−1) of pulse events, largely driven by 8–23% reductions in GEP (0.56–1.92 μmol CO2 m−2 s−1)
(Figure 6). Because the majority of annual GEP at these sites occurs during growing season pulse events(Table 1), our findings highlight the potential for rising VPD to reduce the productivity of these ecosystems.
Figure 3. Composite pattern of pulse responses grouped by days since rain for 193 pulse events from 2004–2018 at thesavanna. Points show half‐hourly fluxes aggregated by days since rain for moderate VPD (blue) and high VPD (red)when soil moisture was moderate (defined as θ between 33rd and 66th percentiles). Lines show mean (a) gross ecosystemphotosynthesis (GEP), (b) surface conductance (Gsfc), (c) ecosystem respiration (Re), (d) evapotranspiration (ET), (e) netecosystem production (NEP), and (f) water use efficiency (WUE), which were computed as the average of half‐hourlyvalues for each day since rain and have the same color ordering as points. Shaded region denotes the standard error.
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 7 of 14
Figure 4. Mean patterns of pulse responses for moderate (solid) and high (dashed) vapor pressure deficit (VPD) during moderate soil moisture conditions at thefour sites (defined as θ between 33rd and 66th percentiles). Lines show daily mean net ecosystem production (NEP; black), gross ecosystem photosynthesis(GEP; green), ecosystem respiration (Re; red), evapotranspiration (ET; black), surface conductance (Gsfc; gray), and water use efficiency (WUE; blue) for each daysince rain.
Figure 5. Impact of air and soil water deficits on integrated pulse responses. Bars show changes in mean responsesintegrated over a 10‐day composite pulse due to high vapor pressure deficit (VPD; black) during moderate soilmoisture conditions, or due to low soil moisture (θ; gray) during moderate VPD conditions. Changes refer to percentdifferences in daily mean (a) gross ecosystem photosynthesis (GEP), (b) surface conductance (Gsfc), (c) ecosystemrespiration (Re), (d) evapotranspiration (ET), (e) net ecosystem production (NEP), and (f) water use efficiency (WUE).Error bars show ±1 standard deviation expressed as percentages.
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 8 of 14
4. Discussion
Pulse responses strongly influence annual ecosystem productivityand are a characteristic feature of many drylands (Noy‐Meir, 1973;Schwinning et al., 2004; Scott et al., 2012). Whereas soil moistureis known to regulate pulses, the effect of atmospheric dryness onpulse responses in water‐limited environments may have beenoverlooked. Here we tested the hypothesis that differences in theVPD sensitivity of Gsfc, GEP, Re, and ET modulate pulse patterns ofNEP and WUE (Figure 1). We found that VPD modified the pulseparadigm: high VPD drove larger reductions in GEP and Gsfc
than Re and ET and thus decreased the NEP and WUE of pulseevents (Figures 3–5). Losses in NEP andWUE at high VPD were per-sistent across a range of soil moisture states and vegetation types(Figures 4, 5, and S8–S13), which indicates that atmospheric demandoperates as a semi‐independent control on the pulse dynamics ofwater‐limited ecosystems. Below we discuss three specific aspectsof these findings.
4.1. Decreased GEP at High VPD Drove Losses in NEP and WUE
We found strong evidence that GEP losses at high VPD decreased the net carbon uptake of pulse events.Because GEP peaked several days later than Re, plants were exposed to periods of increasing VPD that sup-pressed ecosystem photosynthesis (Figures 2–5). Specifically, high VPD was associated with reductions inaverage and peak rates of GEP and appeared to shorten pulses by advancing the onset of postpeak declinesin GEP (Table 2; Figures 4 and 5). High sensitivity of GEP and Gsfc to VPD likely derives from leaf‐levelresponses to air dryness. In response to rising VPD, plants decrease stomatal conductance to minimize thewater cost of carbon gain (Medlyn et al., 2011), which reduces intercellular CO2 concentration and ratesof assimilation (Farquhar & Sharkey, 1982; Oren et al., 1999) and can increase photorespiration (Daiet al., 1992). Evidence from empirical (Table 2) and theory‐based models (Table S2) and binned‐averaginganalysis (Figures 3–6) indicates that the negative effects of high VPD on surface conductance (Novicket al., 2016) and ecosystem photosynthesis (Anthoni et al., 1999; Noormets et al., 2008; Sulman et al., 2016;Wagle et al., 2015) decrease the carbon uptake of pulse events in semiarid ecosystems. These findings werevalidated experimentally by Ruehr et al. (2012), who used irrigation to alleviate soil water stress (analogousto a pulse event) in a semiarid pine forest and found that high VPD was associated with declines in canopyconductance that decreased GEP.
Whereas our conceptual hypothesis (Figure 1) assumes thatGEP andGcanopy have similar sensitivity to VPD,previous studies have shown that differences in the VPD sensitivity of assimilation and stomatal conduc-tance cause ecosystem‐scale intrinsic water use efficiency (iWUE = GEP/Gsfc) to vary with environmentalconditions (Beer et al., 2009; Grossiord et al., 2020; Zhang et al., 2019). The relatively high VPD conditionsat these sites constrain both GEP and Gsfc, causing iWUE to become invariant or decline with furtherincreases in VPD (Grossiord et al., 2020; Zhang et al., 2019). Additionally, our comparison of events underhigh VPD (~4.5 kPa) and moderate VPD (~3.5 kPa) spans a relatively narrow range of VPD for which pre-vious research on a site in this study (US‐Wkg) demonstrates that iWUE did not show a clear response toincreasing VPD (Zhang et al., 2019). We find that substantial VPD constraints on photosynthesis and surfaceconductance persist across a range of soil moisture states and are strongest during wet interstorm periods,especially at sites dominated by C3 shrubs (Figures 5 and S8–S13). Our results indicate the potential for highVPD to alter the amount of rainfall required to generate net carbon uptake in semiarid regions (Scottet al., 2015; Tarin, Nolan, Eamus, et al., 2020; Tarin, Nolan, Medlyn, et al., 2020).
The weak relationship between VPD and Re supports our hypothesis that high VPD does not strongly influ-ence Re (Figures 1, 4, and 5 and Table 2). This insensitivity likely derives from the absence of a mechanisticlink between VPD and cellular respiration and may be augmented by opposing responses of Re to VPD‐modulated factors. Whereas high VPDmay indirectly stimulate respiration through associated warming, ele-vated VPD often coincides with low soil moisture, which directly suppresses soil respiration and may cause
Figure 6. Sensitivity of carbon uptake to a 1 kPa increase in vapor pressuredeficit (VPD) averaged over a 10‐day pulse event. Bars show percent changesin net ecosystem production (NEP; black) and gross ecosystem photosynthesis(GEP; gray) for a 1 kPa increase in mean VPD, relative to moderate VPD and soilmoisture conditions. Error bars show ±1 standard deviation expressed aspercentages.
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 9 of 14
reductions in substrate exudation by water‐stressed plants (Anthoni et al., 1999; Baldocchi et al., 2006; Chenet al., 2009; Kuzyakov & Gavrichkova, 2010; Roby et al., 2019; Ruehr et al., 2012). Because autotrophicrespiration generally scales with GEP (Chapin et al., 2011), autotrophic respiration likely decreases duringpulses as high VPD reduces GEP. Re declined monotonically during pulses, indicating that shallow soilmoisture or labile carbon pools were becoming limiting (Barron‐Gafford et al., 2011; Ma et al., 2012; Pottset al., 2006); however, we observed no VPD effect on the decline (Figures 3 and 4). Because Re pulses are gen-erally brief (Huxman, Cable, et al., 2004; Huxman, Snyder, et al., 2004; Sponseller, 2007) and insensitive toVPD (Table 2 and Figures 4 and 5), the lack of a strong respiration pulse signal does not change our conclu-sion that losses in NEP at high VPD are driven by changes in GEP.
ET dynamics and the high sensitivity of GEP to atmospheric demand reduced the water use efficiency ofpulse events when VPDwas high.We report on ecosystemWUE because it considers key abiotic water losses,which can substantially reduce the water efficiency of dryland productivity (Biederman et al., 2017; Scott &Biederman, 2017).Whereas nontranspirational water fluxes are high 0–2 days after rain (Knauer et al., 2018),transpiration (T) dominates ET later in pulse events, during which ecosystem WUE better reflects bioticresponses to hydrologic stress. Rain events caused immediate reductions in WUE (Figures 3 and 4), likelydue to high rates of abiotic evaporation from soil and canopy surfaces that decreased the T/ET ratio(Huxman, Snyder, et al., 2004; Scott et al., 2006; Szutu & Papuga, 2019). As interstorm days elapsed, GEPincreased whereas ET decreased, and T/ET likely increased (Cavanaugh et al., 2011; Scott et al., 2006)—all factors that would contribute to increasing WUE (Figures 3 and 4). Furthermore, WUE decreased moredue to declines in GEP rather than changes in ET when VPD was high (Figures 3–5). While changes in leafarea may influence the magnitude of Gsfc, GEP, and ET, variation in leaf area index (LAI) at these sites gen-erally occurs over time scales longer than pulse events (Kautz et al., 2019). We defined pulses using aggre-gates of data sampled throughout the growing season, which enables us to make more general inferencesabout VPD effects on pulse responses. However, to ensure our results were not sensitive to LAI variation,we repeated the analysis using only data from August, a period of relatively constant LAI (Cavanaughet al., 2011; Kautz et al., 2019). The results were consistent with those obtained for the entire growing season(data not shown).
High VPD reduced GEP and Gsfc but did not increase ET (Figure 5). Often during pulses, ETwas lower whenVPD was high (Figures 3 and 4). Reductions in transpiration linked to stomatal closure may explain why ETdid not increase despite elevated evaporative potential at high VPD. Together, the observed reductions inGEP and Gsfc but not ET (Figures 4 and 5) indicate strong stomatal regulation of water loss under highVPD conditions (Monteith, 1995). While high VPD is known to reduce ecosystem‐scale iWUE in drylands(Zhang et al., 2019), our focus on ecosystem‐scale WUE (GEP/ET) reflects how abiotic water losses impactthe efficiency of carbon gain, and shows that documented differences in the VPD sensitivity of fluxes(Anthoni et al., 1999) influence pulse patterns of NEP and WUE in semiarid systems (Figures 3–5).
4.2. Grass and Shrub‐Dominated Sites Differed in Their Sensitivity to VPD
Site differences in the VPD sensitivity of pulse responses are consistent with leaf‐level physiology. We foundthat high VPD had stronger impacts at sites dominated by C3 trees and shrubs than in C4‐dominated grass-lands (Figures 4–6). High air temperature often accompanies high VPD (Breshears et al., 2013), and previousresearch at the savanna attributed the greater productivity of C4 grasses during warm conditions to a highertemperature optima for photosynthesis than that of C3 shrubs (Barron‐Gafford et al., 2013). Lower VPD sen-sitivity in the grasslands likely stems from the ability of C4 species to maintain high rates of photosynthesisdespite relatively low stomatal conductance in warm and dry conditions, relative to C3 plants (Christin &Osborne, 2014; Long, 1999; Osborne & Sack, 2012; Pearcy & Ehleringer, 1984; Sage & Monson, 1998).High rates of productivity persisted longer in grasslands than in sites dominated by C3 shrubs (Figure 4), per-haps because the higherWUE of C4 plants can reduce hydraulic stress and support higher relative stomatalconductance during air and soil drought (Anthoni et al., 1999; Christin & Osborne, 2014; Osborne &Sack, 2012).
Site differences in fluxes may be also be influenced by canopy structure and plant water use strategies(Barger et al., 2011; van der Molen et al., 2011). Canopy structural differences (e.g., stand density, LAI, sur-face roughness) can cause differences in evapotranspiration between C4 grasses (Roby et al., 2017) and cansupport higher fluxes (Baldocchi, 1994) and biomass accumulation (Dugas et al., 1996) in C3 species despite
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 10 of 14
the higher WUE of the C4 pathway. Additionally, woody‐plant species have deeper roots than grasses andare capable of hydraulic redistribution (Jackson et al., 1996; Nadezhdina et al., 2010), and a study of a nearbyshrubland found that transpiration was primarily sourced from deeper soil moisture (Szutu & Papuga, 2019).These woody‐plant traits that mitigate soil moisture stress may enhance the susceptibility of these ecosys-tems to rising atmospheric demand.
4.3. Implications for Rising VPD
The strong responses of GEP and Gsfc to high VPD indicate that atmospheric demand can limit carbonuptake independent of changes in precipitation. High VPD reduced NEP and WUE most during wet inter-storm periods (Figures 4 and S8–S13). Analogous to pulse events in this study, photosynthesis in C4 plantswas found to be more resilient to dry‐humid transitions than that of C3 species (Kawamitsu et al., 1993). Indrylands, rain events need to be of substantial size (> ~5–10 mm) to generate carbon sink activity (Chenet al., 2009; Huxman, Snyder, et al., 2004; Kurc & Small, 2007). Since GEP has a higher sensitivity to rainevent size than Re, it has been suggested that projected increases in the frequency of large rain events willenhance NEP in semiarid ecosystems (Chen et al., 2009). However, the observed reductions in GEP at highVPD suggest that larger rain events will be required to support carbon sink activity in light of projected risesin VPD (Ficklin & Novick, 2017).
Average summer VPD in the southwest United States is expected to increase by at least 1 kPa by 2065–2099(Ficklin & Novick, 2017). We found that a 1 kPa increase above moderate VPD decreases the average NEP ofpulse events by 13–56% (Figure 6). Because productivity in these ecosystems is pulse dominated, the prospectof frequent high VPD during interstorm periods could substantially reduce the carbon sink capacity of thesesystems. It is important to note that our focus on periods with high light and soil moisture captures VPDimpacts on peak productivity, which may not represent the carbon balance of actual pulse events. Despitethis focus, we still observed strong reductions in NEP at high VPD (Figure 4) that caused brief periods ofcarbon source activity (Figures 3 and S8–S13). High VPD sensitivity in sites dominated by C3 shrubssuggests the potential for rising trends in VPD to counter increases in semiarid ecosystem net productivitylinked to woody plant encroachment (Barger et al., 2011). High sensitivity of GEP to drought has beenshown to accelerate carbon loss and offset historical carbon sequestration in semiarid ecosystems (Luet al., 2019; Scott et al., 2015). Periods of combined air and soil drought have been shown to decreaseecosystem productivity at continental scales and are projected to increase in frequency (Zhou et al., 2019).Here we show that atmospheric dryness can reduce carbon uptake during critical periods of wateravailability, which is likely to become an increasingly important constraint on ecosystem functioning asdrylands expand globally (Feng & Fu, 2013).
5. Conclusions
In summary, here we advance the pulse model of net carbon gain and water use efficiency in semiarid eco-systems by quantifying the influence of VPD on ecosystem fluxes. We found that high VPD after rain sup-presses GEP more than Re, which decreases the productivity of pulses. Notably, the sensitivity of GEP tohigh VPD was most pronounced in C3‐shrub‐dominated sites when soil moisture was high. These findingsindicate the potential for warming‐driven increases in VPD to reduce the ability of plants to use rain pulses,particularly in regions not dominated by C4 plants. Our results likely extend beyond drylands since wide-spread pulse behavior has been documented across biomes (Feldman et al., 2018). Because pulses are impor-tant for the carbon balance of semiarid systems, high VPD during these critical time periods could shift theseregions from carbon sinks to sources. Representing VPD effects on pulse patterns into conceptual and math-ematical models may help us better understand global change impacts on semiarid ecosystems.
Conflict of Interest
The authors declare no conflict of interest. USDA‐ARS is an equal opportunity employer.
Data Availability Statement
Flux data used in this analysis are available on the AmeriFlux repository (http://ameriflux.lbl.gov) andsite pages: US‐SRM (https://doi.org/10.17190/AMF/1246104), US‐SRG (https://doi.org/10.17190/AMF/
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
1246154), US‐Whs (https://doi.org/10.17190/AMF/1246113), and US‐Wkg (https://doi.org/10.17190/AMF/1246112).
ReferencesAhlström, A., Raupach, M. R., Schurgers, G., Smith, B., Arneth, A., Jung, M., et al. (2015). The dominant role of semiarid ecosystems in the
trend and variability of the land CO2 sink. Science, 348(6237), 895–899. https://doi.org/10.1126/science.aaa1668Anthoni, P. M., Law, B. E., & Unsworth, M. H. (1999). Carbon and water vapor exchange of an open‐canopied ponderosa pine ecosystem.
Agricultural and Forest Meteorology, 95(3), 151–168. https://doi.org/10.1016/S0168-1923(99)00029-5Baldocchi, D. (1994). A comparative study of mass and energy exchange rates over a closed C3 (wheat) and an open C4 (corn) crop: II. CO2
exchange and water use efficiency. Agricultural and Forest Meteorology, 67(3–4), 291–321. https://doi.org/10.1016/0168-1923(94)90008-6Baldocchi, D., Chu, H., & Reichstein, M. (2018). Inter‐annual variability of net and gross ecosystem carbon fluxes: A review. Agricultural
and Forest Meteorology, 249, 520–533. https://doi.org/10.1016/j.agrformet.2017.05.015Baldocchi, D., Tang, J., & Xu, L. (2006). How switches and lags in biophysical regulators affect spatial‐temporal variation of soil respiration
in an oak‐grass savanna. Journal of Geophysical Research, 111, G02008. https://doi.org/10.1029/2005JG000063Barger, N. N., Archer, S. R., Campbell, J. L., Huang, C., Morton, J. A., & Knapp, A. K. (2011). Woody plant proliferation in North American
drylands: A synthesis of impacts on ecosystem carbon balance. Journal of Geophysical Research, 116, G00K07. https://doi.org/10.1029/2010JG001506
Barnes, M. L., Moran, M. S., Scott, R. L., Kolb, T. E., Ponce‐Campos, G. E., Moore, D. J. P., et al. (2016). Vegetation productivity responds tosub‐annual climate conditions across semiarid biomes. Ecosphere, 7, e01339. https://doi.org/10.1002/ecs2.1339
Barron‐Gafford, G. A., Scott, R. L., Jenerette, G. D., Hamerlynck, E. P., &Huxman, T. E. (2013). Landscape and environmental controls overleaf and ecosystem carbon dioxide fluxes under woody plant expansion. Journal of Ecology, 101(6), 1471–1483. https://doi.org/10.1111/1365-2745.12161
Barron‐Gafford, G. A., Scott, R. L., Jenerette, G. D., & Huxman, T. E. (2011). The relative controls of temperature, soil moisture, and plantfunctional group on soil CO2 efflux at diel, seasonal, and annual scales. Journal of Geophysical Research, 116, G01023. https://doi.org/10.1029/2010JG001442
Beer, C., Ciais, P., Reichstein, M., Baldocchi, D., Law, B. E., Papale, D., et al. (2009). Temporal and among‐site variability of inherent wateruse efficiency at the ecosystem level. Global Biogeochemical Cycles, 23, GB2018. https://doi.org/10.1029/2008GB003233
Biederman, J. A., Scott, R. L., Bell, T. W., Bowling, D. R., Dore, S., Garatuza‐Payan, J., et al. (2017). CO2 exchange and evapotranspirationacross dryland ecosystems of southwestern North America.Global Change Biology, 23(10), 4204–4221. https://doi.org/10.1111/gcb.13686
Biederman, J. A., Scott, R. L., Goulden, M. L., Vargas, R., Litvak, M. E., Kolb, T. E., et al. (2016). Terrestrial carbon balance in a drier world:The effects of water availability in southwestern North America. Global Change Biology, 22(5), 1867–1879. https://doi.org/10.1111/gcb.13222
Breshears, D. D., Adams, H. D., Eamus, D., McDowell, N. G., Law, D. J., Will, R. E., et al. (2013). The critical amplifying role of increasingatmospheric moisture demand on tree mortality and associated regional die‐off. Frontiers in Plant Science, 4, 266. https://doi.org/10.3389/fpls.2013.00266
Buckley, T. N. (2019). How do stomata respond to water status? New Phytologist, 224(1), 21–36. https://doi.org/10.1111/nph.15899Cable, J. M., Ogle, K., Lucas, R. W., Huxman, T. E., Loik, M. E., Smith, S. D., et al. (2011). The temperature responses of soil respiration in
deserts: A seven desert synthesis. Biogeochemistry, 103(1–3), 71–90. https://doi.org/10.1007/s10533-010-9448-zCavanaugh, M. L., Kurc, S. A., & Scott, R. L. (2011). Evapotranspiration partitioning in semiarid shrubland ecosystems: A two‐site eva-
luation of soil moisture control on transpiration. Ecohydrology, 4(5), 671–681. https://doi.org/10.1002/eco.157Chapin, F. S., Matson, P. A., & Vitousek, P. (2011). Principles of terrestrial ecosystem ecology. New York: Springer Science & Business Media.Chen, S., Lin, G., Huang, J., & Jenerette, G. D. (2009). Dependence of carbon sequestration on the differential responses of ecosystem
photosynthesis and respiration to rain pulses in a semiarid steppe. Global Change Biology, 15(10), 2450–2461. https://doi.org/10.1111/j.1365-2486.2009.01879.x
Christin, P.‐A., & Osborne, C. P. (2014). The evolutionary ecology of C4 plants. New Phytologist, 204(4), 765–781. https://doi.org/10.1111/nph.13033
Collatz, G. J., Ball, J. T., Grivet, C., & Berry, J. A. (1991). Regulation of stomatal conductance and transpiration: A physiological model ofcanopy processes. Agricultural and Forest Meteorology, 54(2–4), 107–136. https://doi.org/10.1016/0168-1923(91)90002-8
Conant, R. T., Klopatek, J. M., & Klopatek, C. C. (2000). Environmental factors controlling soil respiration in three semiarid ecosystems.Soil Science Society of America Journal, 64(1), 383–390. https://doi.org/10.2136/sssaj2000.641383x
Dai, Z., Edwards, G. E., & Ku, M. S. B. (1992). Control of photosynthesis and stomatal conductance in Ricinus communis L. (Castor Bean)by leaf to air vapor pressure deficit. Plant Physiology, 99(4), 1426–1434. https://doi.org/10.1104/pp.99.4.1426
Ding, J., Yang, T., Zhao, Y., Liu, D., Wang, X., Yao, Y., et al. (2018). Increasingly important role of atmospheric aridity on Tibetan Alpinegrasslands. Geophysical Research Letters, 45, 2852–2859. https://doi.org/10.1002/2017GL076803
Dugas, W. A., Hicks, R. A., & Gibbens, R. P. (1996). Structure and function of C3 and C4 Chihuahuan Desert plant communities. Energybalance components. Journal of Arid Environments, 34(1), 63–79. https://doi.org/10.1006/jare.1996.0093
Farquhar, G. D., & Sharkey, T. D. (1982). Stomatal conductance and photosynthesis. Annual Review of Plant Physiology, 33(1), 317–345.https://doi.org/10.1146/annurev.pp.33.060182.001533
Feldman, A. F., Short Gianotti, D. J., Konings, A. G., McColl, K. A., Akbar, R., Salvucci, G. D., & Entekhabi, D. (2018). Moisture pulse‐reserve in the soil‐plant continuum observed across biomes. Nature Plants, 4(12), 1026–1033. https://doi.org/10.1038/s41477-018-0304-9
Feng, S., & Fu, Q. (2013). Expansion of global drylands under a warming climate. Atmospheric Chemistry and Physics, 13(19),10,081–10,094. https://doi.org/10.5194/acp-13-10081-2013
Ficklin, D. L., & Novick, K. A. (2017). Historic and projected changes in vapor pressure deficit suggest a continental‐scale drying of theUnited States atmosphere. Journal of Geophysical Research: Atmospheres, 122, 2061–2079. https://doi.org/10.1002/2016JD025855
Grossiord, C., Buckley, T. N., Cernusak, L. A., Novick, K. A., Poulter, B., Siegwolf, R. T. W., et al. (2020). Plant responses to rising vaporpressure deficit. New Phytologist, 226(6), 1550–1566. https://doi.org/10.1111/nph.16485
Huxman, T. E., Cable, J. M., Ignace, D. D., Eilts, J. A., English, N. B., Weltzin, J., & Williams, D. G. (2004). Response of net ecosystem gasexchange to a simulated precipitation pulse in a semi‐arid grassland: The role of native versus non‐native grasses and soil texture.Oecologia, 141(2), 295–305. https://doi.org/10.1007/s00442-003-1389-y
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
ROBY ET AL. 12 of 14
AcknowledgmentsThis work was supported byUSDA‐ARS and funding for theseAmeriFlux Core Sites was also providedby the U.S. Department of EnergyBerkeley National Laboratory. Wethank D. Breshears for helpful feedbackon early versions of this manuscript. Wethank the thoughtful and constructivereviews provided by K. Novick and oneanonymous reviewer. We also thank R.Bryant for expert technical assistance inmaintaining site instrumentation andM. Bogan for helpful discussion onstatistics.
Huxman, T. E., Snyder, K. A., Tissue, D., Leffler, A. J., Ogle, K., Pockman, W. T., et al. (2004). Precipitation pulses and carbon fluxes insemiarid and arid ecosystems. Oecologia, 141(2), 254–268. https://doi.org/10.1007/s00442-004-1682-4
Jackson, R. B., Canadell, J., Ehleringer, J. R., Mooney, H. A., Sala, O. E., & Schulze, E. D. (1996). A global analysis of root distributions forterrestrial biomes. Oecologia, 108(3), 389–411. https://doi.org/10.1007/BF00333714
Jenerette, G. D., Barron‐Gafford, G. A., Guswa, A. J., McDonnell, J. J., & Villegas, J. C. (2012). Organization of complexity in water limitedecohydrology. Ecohydrology, 5(2), 184–199. https://doi.org/10.1002/eco.217
Jenerette, G. D., Scott, R. L., & Huxman, T. E. (2008). Whole ecosystemmetabolic pulses following precipitation events. Functional Ecology,22(5), 924–930. https://doi.org/10.1111/j.1365-2435.2008.01450.x
Jia, X., Zha, T., Gong, J., Wang, B., Zhang, Y., Wu, B., et al. (2016). Carbon and water exchange over a temperate semi‐arid shrubland duringthree years of contrasting precipitation and soil moisture patterns. Agricultural and Forest Meteorology, 228‐229, 120–129. https://doi.org/10.1016/j.agrformet.2016.07.007
Jung, M., Reichstein, M., Schwalm, C. R., Huntingford, C., Sitch, S., Ahlström, A., et al. (2017). Compensatory water effects link yearlyglobal land CO2 sink changes to temperature. Nature, 541(7638), 516–520. https://doi.org/10.1038/nature20780
Kannenberg, S., Bowling, D., & Anderegg, W. R. L. (2020). Hot moments in ecosystem fluxes: High GPP anomalies exert outsized influenceon the carbon cycle and are differentially driven by moisture availability across biomes. Environmental Research Letters, 15, 054004.https://doi.org/10.1088/1748-9326/ab7b97
Kautz, M. A., Holifield Collins, C. D., Guertin, D. P., Goodrich, D. C., van Leeuwen, W. J., & Williams, C. J. (2019). Hydrologic modelparameterization using dynamic Landsat‐based vegetative estimates within a semiarid grassland. Journal of Hydrology, 575, 1073–1086.https://doi.org/10.1016/j.jhydrol.2019.05.044
Kawamitsu, Y., Yoda, S., & Agata, W. (1993). Humidity pretreatment affects the responses of stomata and CO2 assimilation to vaporpressure difference in C3 and C4 plants. Plant and Cell Physiology, 34, 113–119. https://doi.org/10.1093/oxfordjournals.pcp.a078384
Kimm, H., Guan, K., Gentine, P., Wu, J., Bernacchi, C. J., Sulman, B. N., et al. (2020). Redefining droughts for the U.S. Corn Belt: Thedominant role of atmospheric vapor pressure deficit over soil moisture in regulating stomatal behavior of maize and soybean.Agricultural and Forest Meteorology, 287, 107930. https://doi.org/10.1016/j.agrformet.2020.107930
Knauer, J., Zaehle, S., Medlyn, B. E., Reichstein, M., Williams, C. A., Migliavacca, M., et al. (2018). Towards physiologically meaningfulwater‐use efficiency estimates from eddy covariance data. Global Change Biology, 24(2), 694–710. https://doi.org/10.1111/gcb.13893
Kurc, S. A., & Small, E. E. (2007). Soil moisture variations and ecosystem‐scale fluxes of water and carbon in semiarid grassland andshrubland. Water Resources Research, 43, W06416. https://doi.org/10.1029/2006WR005011
Kuzyakov, Y., & Gavrichkova, O. (2010). Time lag between photosynthesis and carbon dioxide efflux from soil: A review of mechanismsand controls. Global Change Biology, 16, 3386–3406. https://doi.org/10.1111/j.1365-2486.2010.02179.x
Leuning, R. (1995). A critical appraisal of a combined stomatal‐photosynthesis model for C3 plants. Plant, Cell & Environment, 18, 339–355.https://doi.org/10.1111/j.1365-3040.1995.tb00370.x
Liu, W., Zhang, Z., & Wan, S. (2009). Predominant role of water in regulating soil and microbial respiration and their responses to climatechange in a semiarid grassland. Global Change Biology, 15, 184–195. https://doi.org/10.1111/j.1365-2486.2008.01728.x
Liu, Z. H., Kimball, J. S., Parazoo, N. C., Ballantyne, A. P., Wang, W. J., Madani, N., et al. (2019). Increased high‐latitude photosyntheticcarbon gain offset by respiration carbon loss during an anomalous warm winter to spring transition. Global Change Biology, 26, 682–696.https://doi.org/10.1111/gcb.14863
Lloyd, J., & Taylor, J. A. (1994). On the temperature dependence of soil respiration. Source: Functional Ecology, 8(3), 315–323.Long, S. P. (1999). Environmental responses. In C4 plant biology (pp. 215–249).Lu, C., Tian, H., Zhang, J., Yu, Z., Pan, S., Dangal, S., et al. (2019). Severe long‐lasting drought accelerated carbon depletion in the
Mongolian Plateau. Geophysical Research Letters, 46, 5303–5312. https://doi.org/10.1029/2018GL081418Ma, S., Baldocchi, D. D., Hatala, J. A., Detto, M., & Curiel Yuste, J. (2012). Are rain‐induced ecosystem respiration pulses enhanced by
legacies of antecedent photodegradation in semi‐arid environments? Agricultural and Forest Meteorology, 154–155, 203–213. https://doi.org/10.1016/j.agrformet.2011.11.007
Medlyn, B. E., Duursma, R. A., Eamus, D., Ellsworth, D. S., Prentice, I. C., Barton, C. V. M., et al. (2011). Reconciling the optimal andempirical approaches to modelling stomatal conductance. Global Change Biology, 17, 2134–2144. https://doi.org/10.1111/j.1365-2486.2010.02375.x
Monteith, J., & Unsworth, M. (2013). Principles of environmental physics: Plants, animals, and the atmosphere. Academic Press.Monteith, J. L. (1965). Evaporation and environment. Symposia of the Society for Experimental Biology, 19, 205–234. https://doi.org/
10.1613/jair.301Monteith, J. L. (1995). A reinterpretation of stomatal responses to humidity. Plant, Cell & Environment, 18(4), 357–364. https://doi.org/
10.1111/j.1365-3040.1995.tb00371.xNadezhdina, N., David, T. S., David, J. S., Ferreira, M. I., Dohnal, M., Tesař, M., et al. (2010). Trees never rest: The multiple facets of
hydraulic redistribution. Ecohydrology, 3(4), 431–444. https://doi.org/10.1002/eco.148Noormets, A., McNulty, S. G., DeForest, J. L., Sun, G., Li, Q., & Chen, J. (2008). Drought during canopy development has lasting effect on
annual carbon balance in a deciduous temperate forest. New Phytologist, 179(3), 818–828. https://doi.org/10.1111/j.1469-8137.2008.02501.x
Novick, K. A., Ficklin, D. L., Stoy, P. C., Williams, C. A., Bohrer, G., Oishi, A. C., et al. (2016). The increasing importance of atmosphericdemand for ecosystem water and carbon fluxes. Nature Climate Change, 6(11), 1023–1027. https://doi.org/10.1038/nclimate3114
Noy‐Meir, I. (1973). Desert ecosystems: Environment and producers. Annual Review of Ecology and Systematics, 4(1), 25–51. https://doi.org/10.1146/annurev.es.04.110173.000325
Oren, R., Sperry, J., Katul, G., Pataki, D., Ewers, B., Phillips, N., & Schäfer, K. (1999). Survey and synthesis of intra‐and interspecific var-iation in stomatal sensitivity to vapour pressure deficit. Plant, Cell & Environment, 22(12), 1515–1526. https://doi.org/10.1046/j.1365-3040.1999.00513.x
Osborne, C. P., & Sack, L. (2012). Evolution of C4 plants: A new hypothesis for an interaction of CO2 and water relations mediated by planthydraulics. Philosophical Transactions of the Royal Society, B: Biological Sciences, 367(1588), 583–600. https://doi.org/10.1098/rstb.2011.0261
Pearcy, R. W., & Ehleringer, J. (1984). Comparative ecophysiology of C3 and C4 plants. Plant, Cell and Environment, 7(1), 1–13. https://doi.org/10.1111/j.1365-3040.1984.tb01194.x
Potts, D. L., Huxman, T. E., Cable, J. M., English, N. B., Ignace, D. D., Eilts, J. A., et al. (2006). Antecedent moisture and seasonal preci-pitation influence the response of canopy‐scale carbon and water exchange to rainfall pulses in a semi‐arid grassland. New Phytologist,170(4), 849–860. https://doi.org/10.1111/j.1469-8137.2006.01732.x
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
Poulter, B., Frank, D., Ciais, P., Myneni, R. B., Andela, N., Bi, J., et al. (2014). Contribution of semi‐arid ecosystems to interannual varia-bility of the global carbon cycle. Nature, 509(7502), 600–603. https://doi.org/10.1038/nature13376
Reichstein, M., Falge, E., Baldocchi, D., Papale, D., Aubinet, M., Berbigier, P., et al. (2005). On the separation of net ecosystem exchangeinto assimilation and ecosystem respiration: Review and improved algorithm. Global Change Biology, 11(9), 1424–1439. https://doi.org/10.1111/j.1365-2486.2005.001002.x
Reynolds, J. F., Kemp, P. R., Ogle, K., & Fernández, R. J. (2004). Modifying the ‘pulse–reserve’ paradigm for deserts of North America:Precipitation pulses, soil water, and plant responses. Oecologia, 141(2), 194–210. https://doi.org/10.1007/s00442-004-1524-4
Roby, M. C., Fernandez, M. G. S., Heaton, E. A., Miguez, F. E., & VanLoocke, A. (2017). Biomass sorghum and maize have similar water‐use‐efficiency under non‐drought conditions in the rain‐fed Midwest US. Agricultural and Forest Meteorology, 247, 434–444. https://doi.org/10.1016/j.agrformet.2017.08.019
Roby, M. C., Scott, R. L., Barron‐Gafford, G. A., Hamerlynck, E. P., & Moore, D. J. P. (2019). Environmental and vegetative controls on soilCO2 efflux in three semiarid ecosystems. Soil Systems, 3, 6. https://doi.org/10.3390/soilsystems3010006
Ruehr, N. K., Martin, J. G., & Law, B. E. (2012). Effects of water availability on carbon and water exchange in a young ponderosa pine forest:Above‐ and belowground responses. Agricultural and Forest Meteorology, 164, 136–148. https://doi.org/10.1016/j.agrformet.2012.05.015
Sage, R. F., & Monson, R. K. (1998). C4 Plant Biology. Elsevier.Schwinning, S., Sala, O. E., Loik, M. E., & Ehleringer, J. R. (2004). Thresholds, memory, and seasonality: Understanding pulse dynamics in
arid/semi‐arid ecosystems. Oecologia, 141(2), 191–193. https://doi.org/10.1007/s00442-004-1683-3Scott, R. (2004a). AmeriFlux US‐SRM Santa Rita Mesquite, Dataset. https://doi.org/10.17190/AMF/1246104Scott, R. (2004b). AmeriFlux US‐Wkg Walnut Gulch Kendall Grasslands, Dataset. https://doi.org/10.17190/AMF/1246112Scott, R. (2007). AmeriFlux US‐Whs Walnut Gulch Lucky Hills shrub, Dataset. https://doi.org/10.17190/AMF/1246113Scott, R. 2008). AmeriFlux US‐SRG Santa Rita Grassland, Dataset. https://doi.org/10.17190/AMF/1246154Scott, R. L., & Biederman, J. A. (2017). Partitioning evapotranspiration using long‐term carbon dioxide and water vapor fluxes. Geophysical
Research Letters, 44, 6833–6840. https://doi.org/10.1002/2017GL074324Scott, R. L., Biederman, J. A., Hamerlynck, E. P., & Barron‐Gafford, G. A. (2015). The carbon balance pivot point of southwestern U.S.
semiarid ecosystems: Insights from the 21st century drought. Journal of Geophysical Research: Biogeosciences, 120, 2612–2624. https://doi.org/10.1002/2015JG003181
Scott, R. L., Huxman, T. E., Cable, W. L., & Emmerich, W. E. (2006). Partitioning of evapotranspiration and its relation to carbon dioxideexchange in a Chihuahuan Desert shrubland. Hydrological Processes, 20(15), 3227–3243. https://doi.org/10.1002/hyp.6329
Scott, R. L., Serrano‐Ortiz, P., Domingo, F., Hamerlynck, E. P., & Kowalski, A. S. (2012). Commonalities of carbon dioxide exchange insemiarid regions with monsoon and Mediterranean climates. Journal of Arid Environments, 84, 71–79. https://doi.org/10.1016/j.jaridenv.2012.03.017
Seager, R., Hooks, A., Williams, A. P., Cook, B., Nakamura, J., & Henderson, N. (2015). Climatology, variability, and trends in the U.S.vapor pressure deficit, an important fire‐related meteorological quantity. Journal of Applied Meteorology and Climatology, 54(6),1121–1141. https://doi.org/10.1175/JAMC-D-14-0321.1
Shuttleworth, W. J., & Wallace, J. S. (1985). Evaporation from sparse crops‐an energy combination theory. Quarterly Journal of the RoyalMeteorological Society, 111(469), 839–855. https://doi.org/10.1002/qj.49711146910
Sponseller, R. A. (2007). Precipitation pulses and soil CO2 flux in a Sonoran Desert ecosystem. Global Change Biology, 13(2), 426–436.https://doi.org/10.1111/j.1365-2486.2006.01307.x
Sulman, B. N., Roman, D. T., Yi, K., Wang, L., Phillips, R. P., & Novick, K. A. (2016). High atmospheric demand for water can limit forestcarbon uptake and transpiration as severely as dry soil. Geophysical Research Letters, 43, 9686–9695. https://doi.org/10.1002/2016GL069416
Szutu, D. J., & Papuga, S. A. (2019). Year‐round transpiration dynamics linked with deep soil moisture in a warm desert shrubland.WaterResources Research, 55, 5679–5695. https://doi.org/10.1029/2018WR023990
Tarin, T., Nolan, R. H., Eamus, D., & Cleverly, J. (2020). Carbon and water fluxes in two adjacent Australian semi‐arid ecosystems.Agricultural and Forest Meteorology, 281, 107853. https://doi.org/10.1016/j.agrformet.2019.107853
Tarin, T., Nolan, R. H., Medlyn, B. E., Cleverly, J., & Eamus, D. (2020). Water‐use efficiency in a semi‐arid woodland with high rainfallvariability. Global Change Biology, 26, 496–508. https://doi.org/10.1111/gcb.14866
van der Molen, M. K., Dolman, A. J., Ciais, P., Eglin, T., Gobron, N., Law, B. E., et al. (2011). Drought and ecosystem carbon cycling.Agricultural and Forest Meteorology, 151(7), 765–773. https://doi.org/10.1016/j.agrformet.2011.01.018
Wagle, P., Xiao, X., Scott, R. L., Kolb, T. E., Cook, D. R., Brunsell, N., et al. (2015). Biophysical controls on carbon and water vapor fluxesacross a grassland climatic gradient in the United States. Agricultural and Forest Meteorology, 214‐215, 293–305. https://doi.org/10.1016/j.agrformet.2015.08.265
Wharton, S., Schroeder, M., Bible, K., Falk, M., & Paw, U. K. T. (2009). Stand‐level gas‐exchange responses to seasonal drought in veryyoung versus old Douglas‐fir forests of the Pacific Northwest, USA. Tree Physiology, 29(8), 959–974. https://doi.org/10.1093/treephys/tpp039
Xu, L., Baldocchi, D. D., & Tang, J. (2004). How soil moisture, rain pulses, and growth alter the response of ecosystem respiration totemperature. Global Biogeochemical Cycles, 18, GB4002. https://doi.org/10.1029/2004GB002281
Yan, L., Chen, S., Xia, J., & Luo, Y. (2014). Precipitation regime shift enhanced the rain pulse effect on soil respiration in a semi‐arid steppe.PLoS ONE, 9, e104217. https://doi.org/10.1371/journal.pone.0104217
Zhang, Q., Ficklin, D. L., Manzoni, S., Wang, L., Way, D., Phillips, R. P., & Novick, K. A. (2019). Response of ecosystem intrinsic water useefficiency and gross primary productivity to rising vapor pressure deficit. Environmental Research Letters, 14, 074023. https://doi.org/10.1088/1748-9326/ab2603
Zhou, S., Zhang, Y., Park Williams, A., & Gentine, P. (2019). Projected increases in intensity, frequency, and terrestrial carbon costs ofcompound drought and aridity events. Science Advances, 5, eaau5740. https://doi.org/10.1126/sciadv.aau5740
10.1029/2020JG005665Journal of Geophysical Research: Biogeosciences
