SEARCH

SEARCH BY CITATION

Keywords:

  • alpine meadow;
  • eddy covariance;
  • evapotranspiration;
  • decoupling coefficient;
  • soil water content;
  • canopy conductance

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[1] To characterize evapotranspiration (ET) over grasslands on the Qinghai-Tibetan Plateau, we examined ET and its relevant environmental variables in a Kobresia meadow from 2002 to 2004 using the eddy covariance method. The annual precipitation changed greatly, with 554, 706, and 666 mm a−1 for the three consecutive calendar years. The annual ET varied correspondingly to the annual precipitation with 341, 407, and 426 mm a−1. The annual ET was, however, constant at about 60% of the annual precipitation. About 85% annual ET occurred during the growing season from May to September, and the averaged ET for this period was 1.90, 2.23, and 2.22 mm/d, respectively for the three consecutive years. The averaged ET was, however, very low (<0.40 mm/d) during the nongrowing season from October to April. The annual canopy conductance (gc) and the Priestley-Taylor coefficient (α) showed the lowest values in the year with the lowest precipitation. This study first demonstrates that the alpine meadow ecosystem is characterized by a low ratio of annual ET to precipitation and that the interannual variation of ET is determined by annual precipitation.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[2] Evapotranspiration (ET) in terrestrial ecosystems plays a key role in climate change [Dirmeyer, 1994; Pielke et al., 1998; Betts et al., 1999], because the soil moisture, vegetation productivity, carbon cycle, and water budgets are all affected by ET. Grasslands are one of the most widespread ecosystem types, and natural grasslands account for about 32% of the Earth's natural vegetation [Adams et al., 1990]. Information regarding ET in grassland ecosystems is thus necessary to better understand global climate change and water budgets, which have received increasing attention recently [Baldocchi et al., 2000; Petrone et al., 2001; Toda et al., 2002].

[3] The Qinghai-Tibetan Plateau, with a mean altitude of higher than 4000 m, is the world's highest alpine ecosystem with extensive grasslands that cover more than 60% of the plateau area. The large expanses of alpine meadows on the plateau are of great significance for biodiversity conservation and carbon storage. With its unique topographical and landscape features, the Qinghai-Tibetan Plateau plays an important role in the global atmospheric circulation system and exerts a great influence on both the physical environments and ecosystem functions of adjacent regions [Sun and Zheng, 1996; Yabuki et al., 1998a; Li et al., 1999; Zheng, 2000]. Moreover, the plateau is very sensitive and vulnerable to global climate change because of its high elevation [Klein et al., 2004]. Recent studies indicate that the ground temperature of the Qinghai-Tibetan Plateau has significantly increased over last decades [Liu and Chen, 2000; Yao et al., 2000]. Thus, assessing the ET in the alpine ecosystem on the Qinghai-Tibetan Plateau may provide insights into not only the water cycle in the alpine ecosystem, but also the local and regional climate changes in the highest plateau of the world.

[4] The alpine meadow, which is characterized by relatively high precipitation, strong solar radiation and low temperatures, offers a unique opportunity to examine ecosystem ET in response to changes of environmental and biological conditions. In particular, precipitation showed large interannual variation [Li et al., 2004], which may result in large changes in evapotranspiration, other components of water cycle and energy balance. In this study, we assessed the water vapor exchange between the atmosphere and a Kobresia meadow ecosystem, which is typical of grassland ecosystems on the Qinghai-Tibetan Plateau. The period includes one year with normal annual precipitation that closes to the average from 1980 to 2000, and other two years with much higher precipitation than the average. Our major objective is to characterize the seasonal pattern and interannual variation in the ET for an alpine meadow that coves vast area on the Qinghai-Tibetan Plateau. We intend also to examine the influences of physical and biological environmental variables on the ET in this alpine meadow.

2. Materials and Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

2.1. Study Site Description

[5] The study site (lat. 37°36′N, long. 101°18′E, alt. 3250 m) is an alpine meadow on the northern Qinghai-Tibetan Plateau. The alpine meadow experiences low air temperatures: the daily temperatures for January and July are −15.0°C and 10.0°C, respectively, and the annual mean value was −1.7°C from 1980 to 2000. The mean annual precipitation is 567 mm, of which 80% falls within the growing season from May to September. More detailed information about the study site was described elsewhere [Gu et al., 2003; Kato et al., 2004].

[6] The alpine meadow is dominated by Kobresia humilis (C. A. Mey. Ex Trautv) Serg., K. pygmaea C. B. Clarke, and K. tibetica Maximowicz. The subdominants include Stipa aliena Keng, Elymus nutans Griseb, Festuca ovina Linn, Poa spp. Linn, Gentiana straminea Maxim, and Polygonum viviparum Linn. The grassland around the study site has not been grazed by yaks and sheep during the growing season during the last 20 years. In the alpine ecosystem, plants start to grow in May and began to senesce in September. Leaf area index (LAI) reaches its maximum value of about 3 around mid-August when the average height of vegetation is about 30 cm and the vegetation coverage is over 90% [Shi et al., 2001], and more than 90% of the grass roots are distributed in the 10-cm topsoil layer.

2.2. Measurements

[7] The open-path eddy covariance system was conducted in a flat Kobresia meadow at 220 cm above the ground, with a fetch of more than 250 m from all directions. A three-dimensional sonic anemometer (CSAT3, CSI, Utah, United States) was used to measure turbulence. Fluctuations in heat and water vapor density were measured using the anemometer's temperature and open-path CO2/H2O analyzer set at 10 Hz (Li-7500, Li-Cor, Lincoln, Nebraska, United States).We measured the following environmental variables: above-canopy temperature and humidity at two heights of 110 and 220 cm, respectively (HMP45C, CSI), soil temperature at 0, 5, 10, 20, 30, 50 cm (thermocouple), solar radiation, net radiation at 150 cm (CNR-1, Kipp and Zonen, Netherlands), wind velocity at 110 and 220 cm (014A and 034A-L, CSI), precipitation at 70 cm (TE525MM, CSI), soil water content at 5 and 50 cm depths (TDR soil moisture sensor; CS615, CSI), soil heat flux at 2 cm depth (HFT-3, CSI). All these data were recorded with a data logger (CR23X, CSI) at 15-min intervals. We started our measurements in August 2001, and the data for 2002–2004 are presented in this study. Our study applied the WPL adjustment [Webb et al., 1980] to correct the effect of air density variation on the measured water vapor flux data. It is recommended to make corrections of coordinate rotation, trend removal and water vapor in the eddy covariance method. We did not perform these corrections because we found that the difference between the corrected and uncorrected fluxes was very small (less than 4%, see details given by Kato et al. [2004]), and thus concluded that the bias due to noncorrections can be negligible in the study.

[8] In this study, the aboveground biomass was measured consecutively at approximately 2-week intervals by clipping aboveground vegetation within a 50 × 50 cm quadrat during the growing season for the three years. Biomass was determined gravimetrically after samples had been dried for 72 h at 65°C.

2.3. Data Analysis

[9] The energy budget for a vegetation surface can be expressed by using the following equation [e.g., Oke, 1987; Wilson et al., 2002]:

  • equation image

where Rn is the net radiation flux, H is the sensible heat flux, G is the soil heat flux, L is the latent heat of vaporization, and LET is the latent heat flux. The energy balance ratio (EBR) was calculated using the following equation [Wilson et al., 2002]:

  • equation image

The term (LET + H) measured by the eddy covariance method seemed to be underestimated because the average value of EBR was 0.70 in this study site [Gu et al., 2005].

[10] The measured data of water vapor flux could not be used from May to early July 2003 because of an error of H2O calibration in the Li-7500 sensor. The missing data of daily ET for this period were estimated using the measured Rn, H, and G from equation (1) and EBR from equation (2).

[11] Canopy conductance was estimated by inverting the Penman-Monteith equation [Monteith and Unsworth, 1990]:

  • equation image

where ρ is air density, Cp the specific heat of air at constant pressure, VPD the vapor pressure deficit, ga the air conductance, γ the psychrometric constant, and gc the canopy conductance. β (= H/LET) is the Bowen ratio, Δ the slope of the saturation vapor pressure curve at the mean wet-bulb temperature of the air, and ga is calculated by the following equation [Monteith and Unsworth, 1990]:

  • equation image

where u* is friction velocity and u is wind speed.

[12] The decoupling coefficient (Ω) represents the magnitude of the coupling effect of the canopy and the aerodynamic conductance in controlling rates of canopy transpiration [Jarvis and McNaughton, 1986]. Ω was calculated using the following equation [Jarvis and McNaughton, 1986]:

  • equation image

[13] The equilibrium evapotranspiration (ETeq), which represents the minimum possible evaporation rate from a moist surface, was calculated as follows [Priestley and Taylor, 1972]:

  • equation image

ETeq depends only on the temperature and available energy.

[14] Potential evapotranspiration (PET) is a measure of the ability of the atmosphere to remove water from the surface through the processes of evaporation and transpiration assuming no limitation on water supply. The PET was calculated based on the method as follows [Priestley and Taylor, 1972; Qiu et al., 1999]:

  • equation image

in which the value of 1.26 was given as a constant by Priestley and Taylor.

[15] To assess the influence of weather conditions and vegetation on the ET, the climatic moisture index (CMI) was calculated from the difference between monthly precipitation and PET using the following equation [Hogg, 1994]:

  • equation image

where Pm is monthly precipitation and PETm is monthly PET.

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

3.1. Temporal Variation of Environmental Variables

[16] The total solar radiation averaged 6300 MJ m−2 a−1 with its 48% within the growing season from May to September (Figure 1a). The daily mean solar radiation was about 17 MJ m−2 d−1 for the three years, but reached 20 MJ m−2 d−1 for the growing season. There was no significant difference in the observed values of the annual total solar radiation and the annual mean temperature among the three years (Figures 1a and 1c). The daily mean air temperature was around 7°C during the growing season (Table 1). Annual mean soil temperature at 5-cm depth was around 5°C for the three years, which was higher than annual mean air temperature (−1.7°C) (Table 1).

image

Figure 1. Annual variation in (a) solar radiation (Rs), (b) precipitation, (c) air temperature (Ta), (d) soil temperature at 5-cm depth (Ts), (e) soil water content at 5-cm depth, (f) soil water content at 50-cm depth, and (g) air vapor pressure deficit (VPD) from 2002 to 2004.

Download figure to PowerPoint

Table 1. Comparison of Environmental Conditions During the Growing and Nongrowing Season in Study Site on the Qinghai-Tibetan Plateau
YearGrowing Season (May–September)Nongrowing Season (January–April and October–DecemberAnnual (January–December)
Solar Radiation, MJ m−2d−1
200220.914.917.5
200320.114.416.9
200420.115.017.3
 
Air Temperature, °C
20027.9−7.6−1.2
20037.4−6.5−0.7
20047.1−7.8−1.6
1980–20007.2−7.8−1.7
 
Soil Temperature at 5-cm Depth, °C
200213.4−0.45.3
200312.6−0.15.1
200412.2−0.54.7
 
Total Precipitation, mm
2002476.177.9554.0
2003607.398.3705.6
2004579.786.5666.2
1980–2000460.4106.5566.9
 
VPC, kPa
20020.730.480.58
20030.710.480.58
20040.710.430.55
 
CMI, mm
2002−89.9−185.3−275.2
200320.0−188.5−168.5
200433.8−208.4−174.6

[17] In comparison with radiation and temperatures, the precipitation on the meadow showed a larger difference during the growing season among the three years (Figure 1b). The precipitation during the 2002 growing season was 476 mm, which was comparable with the average value of 460 mm for the period 1980–2000 (Table 1). In contrast, the precipitation during the 2003 and 2004 growing seasons was 607 and 580 mm, respectively, more than 100 mm, i.e., 32% and 26% higher than the normal year (Table 1). In particular, about 85% of annual precipitation fell within the growing season for the three years (Table 1). Rainy days with precipitation less than 1 mm d−1 had the highest frequency (34%), and about 70% of rainy days had precipitation less than 5 mm d−1.

[18] With the high precipitation, the volumetric soil water content was also high during the growing season (Figures 1e and 1f). Soil water content at 5-cm depth during the growing season varied between about 0.22 and 0.56 cm3 cm−3 (Figure 1e). Soil water was significantly lower during the period between June and August than the earlier growing season despite that precipitation was higher in the late growing season. There were no major differences observed for soil water content at 50-cm depth among the three years (Figure 1f), with the value consistently higher than 0.40 cm3 cm−3 during the growing season. The highest soil water content recorded at depths of both 5 and 50 cm occurred in May. The soil water content from August to November 2002 was lower than that in the same period 2003 and 2004. The field capacity for the alpine meadow soil was reported to be 0.37 and 0.40 cm3 cm−3 at depths of 0–20 cm and 50 cm, respectively [Cao et al., 1998]. We found that the soil water content at 5-cm depth was often lower than the field capacity from July to October 2002, but it remained relatively high during the same period in 2003 and 2004 (Figure 1e).

3.2. Evapotranspiration and Its Environmental Controls

[19] ET was relatively low during early spring but quickly increased beginning in May and reached its peak value in July [data not shown]. After July, ET started to decline, accompanied by decreases in both available energy and plant senescence. The daily average ET values were 0.98, 1.11, and 1.17 mm for the three years. The daily average ET values during the growing season were 1.90, 2.23, and 2.22 mm d−1 in 2002, 2003, and 2004, respectively. The ET was always less than 1 mm d−1 during the nongrowing season and became extremely low (<0.2 mm d−1) during the period from December to February.

[20] Precipitation was significantly greater than ET during the growing season, but approximately equal to ET from November to March (Figure 2). The average annual sum of ET was about 60% of the annual precipitation for the three years. There was no significant difference in cumulative ETeq over the three years, with values of 667, 665, and 645 mm, respectively. Both the cumulative precipitation and actual ET in 2002 were markedly lower than those in 2003 and 2004 (Figure 2).

image

Figure 2. Cumulative precipitation, evapotranspiration (ET), and equilibrium evapotranspiration (ETeq) from 2002 to 2004.

Download figure to PowerPoint

[21] CMI during the 2002 growing season was −89.9, indicating that precipitation was lower than PET in this period (Table 1). However, the CMI during the growing season 2003 and 2004 was greater than zero, indicating that the precipitation exceeded PET in this period. CMI was similar among the three nongrowing seasons.

[22] VPD, calculated in midday from 11:00 to 15:00 at Beijing Standard Time (BST, which is about 1 h 15 min earlier than the solar time at the study site), showed similar seasonal patterns for the three years (Figure 1g), with high values during the growing seasons (Table 1). The 2002 growing season had a slightly high VPD value than that observed in 2003 and 2004; the VPD during the nongrowing season 2004 was lowest (Table 1).

3.3. Albedo and Bowen Ratio

[23] Daily albedo was calculated by taking the ratio of daily sums of radiation reflected by the meadow to the incoming solar radiation. In the three years the daily albedo was the lowest in September (Figure 3a). There was a small but significant difference in albedo during the growing season, with the albedo being lower in 2004 than in 2002 and 2003. During the nongrowing season, the albedo varied greatly and showed higher values than that during the growing season. The albedo during snow covered period reached a maximum of 0.92.

image

Figure 3. Seasonal and interannual variation in (a) daily albedo and (b) Bowen ratio from 2002 to 2004. Data points are daily values averaged over five consecutive days.

Download figure to PowerPoint

[24] The Bowen ratio (β) ranged from below 0.2 to more than 10 from July to December (Figure 3b). The curve of β versus time for the alpine meadow was U-shaped, with the low values during the growing season. There was almost no difference in β from June to August among the three years. However, the β in September and October 2002 was significantly higher than that in the same months of 2003 and 2004.

3.4. Canopy Conductance and Decoupling Coefficient

[25] To assess physiological control over water losses, we examined the canopy conductance (gc). The gc exhibited a similar pattern during the growing seasons of the three years (Figure 4), and the annual maximum gc occurred in July. The diurnal maximum gc occurred between 12:00 and 13:00 BST from June to August and at around 9:00 in May and September. In addition, the gc was markedly lower throughout the growing season in 2002 compared to that in 2003 and 2004.

image

Figure 4. Seasonal and interannual variation in the mean diurnal course of canopy conductance (gc). Error bars indicate standard error.

Download figure to PowerPoint

[26] The decoupling coefficient (Ω) has a range between 0 and 1; a low Ω indicates a relatively high influence of VPD on ET, whereas a high Ω suggests that Rn is the dominant influence. In this study, the value of Ω normally was more than 0.5 for the three growing seasons (Figure 5). The highest monthly average Ω occurred in July for three study years. There was no distinct difference in Ω values from May to August among the three years, although the Ω in September 2002 was lower than that in September 2003 and 2004. The diurnal variation showed a similar pattern among the three years, with higher values in the morning and lower values in the evening.

image

Figure 5. Seasonal and interannual variation in the mean diurnal course of the decoupling coefficient (Ω). Error bars indicate standard error.

Download figure to PowerPoint

[27] The highest Priestley-Taylor coefficient (α = ET/ETeq) was recorded in July, and low values occurred during the nongrowing seasons of the three years (data not shown). The α increased quickly with the increase of canopy conductance in the low-conductance range (gc < 15 mm−1), but the increase of α was slow when gc was more than 15 mm−1 (Figure 6). There was little difference between 2003 and 2004, but the averaged α in 2002 growing season was 0.66, which was markedly lower than that in 2003 and 2004 (about 0.77).

image

Figure 6. Relationship between canopy conductance (gc) and the daily average Priestley-Taylor coefficient (ET/ETeq). The lines represent a fitted regression for 2002, ET/ETeq = 0.1582ln(gc) + 0.2316 (r2 = 0.83); 2003, ET/ETeq = 0.1972ln(gc) + 0.2355 (r2 = 0.85); and 2004, ET/ETeq = 0.1915ln(gc) + 0.2486 (r2 = 0.85).

Download figure to PowerPoint

4. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

4.1. Characteristics of Evapotranspiration in the Alpine Meadow

4.1.1. Annual Variation of Evapotranspiration

[28] The significant seasonal variation of ET was determined by the temporal variation of precipitation, availability of soil water and vegetation coverage. In the study site, plants started growing in May and began to senesce from September. The high ET during the growth season was due to high precipitation and available energy. During this period, the fraction of transpiration increased gradually, while the fraction of evaporation from soil surface decreased with the increase of plant coverage. The low evaporation during the nongrowing season was due to the low soil water content and available energy. In particular, the lack of liquid water in frozen soil and the low energy input greatly reduce the evaporation from November to March.

[29] In terms of the terrestrial water cycle, ET is the largest component of water loss from ecosystems: about 70% of precipitation returns to the atmosphere through evaporation and transpiration processes [Rosenberg et al., 1983]. In this alpine meadow, about 60% of precipitation was returned through ET on an annual basis for the three years (Figure 2). The ratio of actual ET to precipitation was also significantly lower than that reported for tussock grassland [Hunt et al., 2002] and northern temperate grassland [Wever et al., 2002], but it was comparable to wet temperate grassland in Japan [Li et al., 2005].

[30] Precipitation inputting into an ecosystem is balanced by ET, infiltration, storage, runoff and interception. In this meadow, it is reasonable to neglect the runoff because of the flat study site and/or small amount of precipitation for each rain event. The lower ratio of ET to precipitation in the alpine meadow as compared with other grasslands was perhaps due to the high soil water infiltration during the growing season (Figure 2). Numerous studies have shown that ecosystem ET is mainly dependent on the soil water content [Lafleur et al., 1992; Betts and Ball, 1997; McFadden et al., 1998; Ridolfi et al., 2000]. In this alpine meadow, the soil water content at 5-cm depth is likely to be the most important for vegetation growth because most of the root system is distributed within 0- to 10-cm surface layer. In this study, the large amplitude variations in daily mean soil water content at 5-cm depth were observed during the growing season for three years (between 0.21 and 0.56 m3 m−3) (Figure 7a), indicating that the soil water limitation at 5-cm depth was highly correlated with ET and precipitation events (daily precipitation data not shown). However, the soil water content at 50-cm depth was almost above the field capacity during this season (Figure 7b). These results indicate that there was a higher infiltration in the upper layer of soil, while soil water at deeper layers was almost not influenced by ET. Moreover, the ET from the alpine meadow was much lower than PET, indicating further that water supply might restrict the evapotranspiration.

image

Figure 7. Seasonal and interannual variation in (a) daily soil water content at 5-cm depth and (b) soil water content at 50-cm depth from 2002 to 2004.

Download figure to PowerPoint

4.1.2. Evapotranspiration and Radiation Regime

[31] Assuming that moisture is available, ET depends primarily on the solar energy available to vaporize the water [Hanson, 1991]. While available energy is largely dependant on surface albedo, which in turn is determined by many factors including vegetation coverage, vegetation height, species composition, soil moisture, and incident solar radiation [Song, 1999]. The mean daily albedo during the growing season in the alpine meadow was higher than that reported for many other grasslands [Gao et al., 2005] and Arctic tundra [Eugster et al., 2000]. On the other hand, the incoming long-wave radiation from the atmosphere was lower in the alpine ecosystems, only about 50–70% that of lowland ecosystems [Ji et al., 1995], but the net long-wave radiation is much higher than that of lowlands [Ji et al., 1995] owing to the larger difference between the meadow surface and air temperatures.

[32] The high albedo combined with the high net long-wave radiation resulted in the low Rn despite of the high incident solar radiation on the alpine meadow. The average annual Rn obtained for the alpine meadow is comparable to those reported for a grassland in Japan [Li et al., 2005], though Rs was about 1000 MJ m−2 a−1 lower in the grassland than in the meadow. The Rn/Rs about 0.54 for the growing seasons and about 0.43 for the three years in the meadow are also much lower compared with those reported from other grasslands [Shaw, 1956; Rosenberg, 1969; Li et al., 2005] and boreal vegetation [Baldocchi et al., 2000; Eugster et al., 2000]. These results indicate that despite of high Rs, in the alpine meadow, the energy available for driving the ET is not high. Such the energy partitioning may also result in the low ET in the alpine meadow. The ET from the meadow was lower than other grassland ecosystems [Kelliher et al., 1993; Li et al., 2005; Wever et al., 2002].

4.1.3. Effects of Vapor Pressure Deficit on Evapotranspiration

[33] Vapor pressure deficit (VPD) is one of the principal weather variables affecting ET because VPD affects canopy conductance. In this alpine meadow, the value of measured VPD varied within a very narrow range, from 0.02 to 1.7 kPa, and it was significantly lower than many other grassland values with the maximum VPD ranged from about 2 to 5 kPa [Verhoef et al., 1996; Kellner, 2001; Hunt et al., 2002; Wever et al., 2002]. As shown in equation (3), a low VPD tends to decrease ET in this ecosystem.

4.1.4. Evapotranspiration and Plant Biomass, Canopy Conductance and Decoupling Coefficient

[34] Vegetation has a significant impact on ET [Wever et al., 2002]. An increase in leaf area will initially increase gc (and ET) when soil water content is high, this response will weaken at high LAI [Ripley and Saugier, 1978; Obrist et al., 2003; Li et al., 2005]. In the alpine meadow ecosystem on the Tibetan Plateau, aboveground biomass is highly related to LAI because of low biomass proportion of shoots [Shi et al., 2001]. We arbitrarily divided the growing season into two phases at the point of maximum aboveground biomass, i.e., the preceding phase with increasing aboveground biomass and the late phase with decreasing aboveground biomass. ET increased gradually with the increase of aboveground biomass before July, but decreased slowly thereafter despite that the aboveground biomass still increased until August because of growth of reproduction organs (Figure 8). In the late growing season, ET decreased rapidly with the decline of aboveground biomass in late phase when leaves were dying rapidly. This confirms the importance of leave biomass in controlling ET of the alpine meadow.

image

Figure 8. Relationship between ET and aboveground biomass during the growing seasons of 2002–2004. Solid circles indicate increasing aboveground biomass; the curve is a smoothing fit (r2 = 0.77); open circles indicate decreasing aboveground biomass. The linear relationship is described by ET = 0.0119biomass − 1.481 (r2 = 0.54).

Download figure to PowerPoint

[35] Ecosystem ET is linked with canopy conductance (gc) of the upper canopy leaves when averaged across the growing seasons [Bernacchi et al., 2007]. Higher ET is often associated with larger gc, although the effect of gc and environment on ET varies widely among vegetation types and environmental conditions. gc is often strongly controlled by environmental conditions, such as soil moisture and VPD [Morison and Gifford 1983; Sperry et al., 2002; Wever et al., 2002]. Other studies indicated that gc tends to be higher at low VPD [Kellner, 2001; Law et al., 2002; Whitehead et al., 2004]. The value of gc in this alpine meadow was higher than that reported for wet meadow, arctic tundra, and northern temperate grassland [McFadden et al., 1998; Wever et al., 2002], which may in part be due to the low VPD. The alpine meadow gc was smaller compared to maize canopy [Steduto and Hsiao, 1998] and wet temperate grassland [Li et al., 2005] during the growing season, perhaps due to their higher LAI. In this study, gc in 2002 was lower than that in 2003 and 2004 throughout the growing season (Figure 4), one possible reason for this is that 2002 had the relatively high VPD compared with 2003 and 2004 during the growing season (Table 1).

[36] The decoupling coefficient (Ω) provides a tool for separating the effect of VPD on the ET from that of Rn. Ω during growing season is much higher than northern temperate grassland [Wever et al., 2002], but is comparable to the wet temperate grassland [Li et al., 2005]. Usually Ω was large 0.5 throughout growing season, the high value of Ω during the growing season suggested that ET was mainly determined by the available energy. The Ω value decreased in the late afternoon, indicating growing coupling of the canopy to the atmosphere because the low gc and high VPD usually occurred in the afternoon, indicating that the contribution of VPD to ET increased in the afternoon.

4.2. Energy Partitioning of the Alpine Meadow

[37] The Bowen ratio (β) is often used to describe the contribution of LET and H to the surface energy balance, which implicitly reflects the physiological activities of the canopy. The lowest β (below 0.2) in July and August was comparable to that of wet temperate grassland in Japan [Li et al., 2005], indicating that the contribution of ET to energy balance was very important during this period. However, the Bowen ratio during the nongrowing season was very high and even higher than that of semiarid regions, which have a β of 2–6 [Nobel, 1999]. The high β values are mainly because the loss of soil water through evapotranspiration is greatly reduced; in particular, frozen soil may completely block the movement of excess moisture through the soil profile in the cold winter. It should also be noted that the Bowen ratio in September and October 2002 was markedly higher than that in the same months of 2003 and 2004. Soil water content at 5-cm depth in September and October 2002 was markedly lower than that in the same period 2003 and 2004 (Figure 1e), indicating that soil water might have a limitation on ET in September 2002 in comparison with the other two years, which may have caused the high β.

5. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[38] We examined the ET over a Kobresia meadow, one of the most broadly distributed vegetations on the Qinghai-Tibetan Plateau from 2002 to 2004. It is evident that either the interannual ET or the seasonal variation of ET was highly correlated with precipitation. About 60% of annual precipitation was evapotranspirated from the alpine ecosystem during the three years. The proportion of ET to precipitation is smaller than other temperature grassland ecosystems. Despite of the high solar radiation input into the alpine meadow, the radiation available for ET was similar to that in lowland grasslands reported so far. These results indicate that both the low Rn/Rs and low available water contributed to the low ET in the alpine ecosystem. The canopy conductance (gs) of the alpine meadow was higher than many dry and/or cold grassland ecosystems, but lower than wet temperate grasslands. Calculation of the decoupling coefficient (Ω) indicated that the diurnal and seasonal variability of ET was primarily dependent on the variations of available energy. An understanding of the ET from this ecosystem is critical for any further assessment of ET in this region.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[39] This study was supported by the One Hundred Talent Project (0429091211) and by the Global Environment Research Fund of the Ministry of the Environment, Japan (S-1: Integrated Study for Terrestrial Carbon Management of Asia in the 21st Century Based on Scientific Advancements). The study was conducted under a Knowledge Innovation Research Project of the Chinese Academy of Science (KZCX1-SW-01-01A).

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Materials and Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
jgrd14267-sup-0001-t01.txtplain text document1KTab-delimited Table 1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.