Climate and Soil Water

Water Freedom System

Survive Global Water Shortages

Get Instant Access

Precipitation or water supply is fundamental to terrestrial ecosystem evaporation. In the boreal zone, it is relatively sparse. Precipitation cannot generally be exceeded by evaporation in the long term. Ground water collection areas are exceptional in that water supply for evaporation is greater than the precipitation. An example is the type of wetland known as a fen. However, precipitation is the primary source of water in the boreal zone and besides the orographic effect mentioned earlier, long-term records suggest that average annual rates (computed as annual totals divided by 365 days) decrease in Siberia and North America from around 2 mm day-1 in the south down to only 0.3 mm day-1 in the north (Table 1). In Scandinavia, there seems to be little or no latitudinal gradient in the 1.5 mm day-1 annual average precipitation rate, probably reflecting its general proximity to the sea. The boreal zone's short summers are generally wetter than other times of the year with rainfall averaging 0.6-3.5 mm day-1 for June-August, except for northern Scandinavia.

To examine the summer rainfall gradient more closely, we utilize a scaled-up 1° latitude by 1° longtitude grid-cell data set of average monthly rainfall (Leemans and Cramer, 1991). The data are divided into four boreal regions: North America (48.5-68.5°N, 58.5-163.5°W, 776 cells), European Russia (52.5-68.5°N, 22.5-59.5°E, 161 cells), Western Siberia (48.5-69.5°N, 60.5-89.5°E, 331 cells), and Eastern Siberia (49.5-72.5°N, 90.5-178.5°E, 1168 cells). For North America, which is the driest of the regions except on the windward/western side of the cordillera, regression shows that the inverse linear relation between daily summer rainfall and latitude has a slope of — 0.09 mm per degree and offset of 7.2 mm accounting for 76% of the variation (Fig. 2A). Results obtained for European Russia are — 0.05 mm per degree,

TABLE 1 Latitudinal Transects of Climate Data from Müller (1982) for the Boreal Zone

Site

Latitude/Longitude (degrees)

Average Temperature

Jim —Aug Year

Average Precipitation Rate (mm day-1)

Jun-Aug Year

Drought Coefficient Jun-Aug

Western North America

Barrow

71.3 N/136.8 W

2.6

- 12.4

0.6

0.3

0.24

Whitehorse

60.7 N/135.0 W

13.1

-0.7

1.1

0.7

0.33

Fort Nelson

58.8 N/122.6 W

15.3

- 1.1

2.0

1.2

0.38

Prince George

53.9 N/122.7 W

13.8

3.3

2.1

1.7

0.45

Central Siberia

Dudinka

69.4 N/86.2 E

8.7

- 10.7

1.2

0.7

0.48

Yeniseysk

58.3 N/92.2 E

15.6

-2.2

1.9

1.3

0.52

Minusinsk

53.7 N/91.7 E

17.8

-0.1

1.8

0.9

0.42

Eastern Siberia

Bulun

70.8 N/127.8 E

8.1

- 14.5

0.8

0.3

0.36

Yakutsk

62.0 N/1 29.8 H

16.3

- 10.2

1.2

0.6

0.32

Bornnak

54.7 N/129.0 E

16.2

-4.9

3.5

1.5

0.42

Scandanavia

Vardo

70.4 N/31.1 E

8.4

1.6

1.4

1.5

0.47

Stockhom

59.4 N/18.0 E

16.4

6.6

2.0

1.5

0.43

The drought coefficient (£) for a period is equal to the number of days with precipitation divided by the total number of days. It is used to estimate the probability of a day being dry (<J>d) for a given time (t) since rainfall in the integrated Poisson Process model d>j = 1 — e-t'.

5.3 mm, and 42%, respectively, suggesting a more variable and wetter regime (Fig. 2B). For Western Siberia at — 0.11 mm per degree, 9.0 mm, and 80%, respectively, the latitudinal gradient is similar and generally as tight as in North America but, as in European Russia, rainfall intensity is 0.5-0.9 mm day-1 greater (Fig.

2C). North and south of 60°N, there are different rainfall regimes in Eastern Siberia (Fig. 2D). The north is much drier. Rainfall is also less variable with a latitudinal gradient and offset nearly equal to those of European Russia although regression only accounts for 14% of the variation. South of 60°N the latitudinal gradient and

Wlhilliiiii.

0 sIIIIflIIB

□ au □□

50 55 60 65 70

50 55 60 65 70

Latitude [degrees North] FIGURE 2 The relation between average rainfall rate during June through August 1988 (mm day-1) computed as total rainfall for the 3 months (mm) divided by 92 days) and latitude (degrees North) in boreal North America (A), European Russia (B), Western Siberia (C) and Eastern Siberia (D). The data and regions are described in the text.

offset: were greater, —0.15 mm per degree and 11.3 mm, respectively (r2 = 0.40).

Longitudinal variation in Siberian rain during summer includes a generally decreasing fall going eastward from the Ural Mountains, as noted by Schulzc et al. (1999), with some influence of orography and proximity to the major rivers and the Sea of Okhotsk at the eastern frontier. For example, at latitude 66°N, average rainfall declines by nearly a factor of 2 from 2.5 to 1.3 mm day 1 going from 62 to 168°E. The relation is described by a line of slope —0.007 mm per degree and offset 2.65 mm; regression accounts for 42% of the variation (data not shown).

Rainfall frequency can be important because surface drying during fine weather greatly reduces the ground evaporation rate, which can be a significant fraction of the total evaporation from boreal ecosystems with sparse vegetation (e.g., Kelliher et al., 1998). For the summer months of June-August, rainfall frequency may be considered a Poisson process with the probability of each event being independent (although, in a continental climate, terrestrial surface-energy partitioning may contribute to rainfall occurrence.) It is represented by a simple exponential function of time since the last event it) (e.g., <bd = 1 — e where <()d and L, are the probability of a day being dry and a drought coefficient, respectively; Kelliher et al., 1997). The coefficient £is computed simply as an inverse of the average time between rainfalls. In drier eastern Siberia and western North America, like rainfall, £ tends to decrease significantly from south to north, but it is relatively constant throughout central Siberia and Scandinavia (Table 1). fable 1 shows that the probability of a weeklong dry period after rainfall varies from 0.03 (Yeniseysk with the highest £ tabulated) to 0.19 (Barrow with the lowest £). Considering the number of rainy days, or chances to "reset the clock," the Barrow site is thus relatively certain to have such a weeklong dry spell each summer, while the chance is 57% for Yeniseysk.

Besides rainfall intensity and frequency, soil storage is another important determinant of water supply for evaporation. The relation between volumetric soil water content (6, m3 of water per nr; of soil) and suction (S) reflects the pore-size distribution and quantifies the water storage characteristics of a soil (Fig. 3). The curve's shape can be complex, but for the range of values encountered under most field conditions in natural boreal ecosystems, where S is rarely near 0, a relatively simple function is sufficient (Campbell, 1985); S(6) = Se{0/6sy1', where Se is the air entry value of S at which the largest water-filled pores just drain, 0S is the saturation value of 0 that is equal to the soil's porosity, and the power coefficient b is empirically determined. In practice, both Se and b are determined by plotting S and 0 on logarithmic scales and fitting a straight line to the data, the regression slope and offset being b and Se, respectively.

Values of water-release curve parameters Se, 8S, and b have been related to texture, the most permanent feature of a soil, based on the percentages of sand-, silt-, and clay-sized particles (1448 samples from 35 locations; Cosby et al., 1984). This information may be combined with the 1° latitude by 1° longitude grid-cell texture database of Xobler (1986) to estimate soil water storage capacities

FIGURE 3 The relation between volumetric water content (6, m! m--') and suction (S, kPa) for sand and silt loam mineral soils (solid squares and line and open squares and dashed line, respectively) and humus (i.e., organic matter; solid triangles and line). The sand came from the upper 0.1 -m depth of soil beneath a stand of Pinus sylvestris trees located in central Siberia (Kelliher et al, 1998), the silt loam from the 0.2-0.3-m depth of soil beneath a stand of Larix gmelinii trees in eastern Siberia (Kelliher et al, 1997), and the humus from depth 0.14 m in a 0.3-m- deep forest floor beneath a stand of Tsuga heterophylla and Thuja plicata trees near Vancouver, Canada (Plamandon et al, 1975).

Suction [kPa]

FIGURE 3 The relation between volumetric water content (6, m! m--') and suction (S, kPa) for sand and silt loam mineral soils (solid squares and line and open squares and dashed line, respectively) and humus (i.e., organic matter; solid triangles and line). The sand came from the upper 0.1 -m depth of soil beneath a stand of Pinus sylvestris trees located in central Siberia (Kelliher et al, 1998), the silt loam from the 0.2-0.3-m depth of soil beneath a stand of Larix gmelinii trees in eastern Siberia (Kelliher et al, 1997), and the humus from depth 0.14 m in a 0.3-m- deep forest floor beneath a stand of Tsuga heterophylla and Thuja plicata trees near Vancouver, Canada (Plamandon et al, 1975).

throughout the world. Caution is thus required and the reader is directed to an insightful critique by Nielsen et al (1996). The boreal zone includes about half the world's wetland area (2.6 Tm2 for 50-70°N; Matthews and Fung, 1987) that has organic soils. Our areal analysis excluded these areas. This necessitated estimation of organic soil areas within cells whose vegetation is classified as forest, or cells adjacent to nonforested wetlands that are forested, because these cells are part of the boreal zone's forest/wetland mosaic that also contains mineral soils. Organic soils of this mosaic cover 9% of the total area analyzed (i.e., 1.3 of 14.2 million km2). About half the organic soil area of the mosaic is in Siberia with the greatest percentage cover in Western Siberia at 24%. For all of boreal North America and Russia, our mineral soil analysis suggests that there are effectively only four textural classes ranging from coarse or loamy sand to medium fine or loam (Table 2). Mostly though, the soils are either coarse or loamy. Large areas of boreal North America and western Siberia have coarse soils, but Russia predominantly has loamy soils.

The water storage capacity of soil depends variably on the two values of S (i.e., the lower or wet limit and the upper or dry limit) chosen for the computation (Table 2). Our purpose is to estimate the quantity of water available for evaporation over a range where it is not supply limited. Thus, we wish to quantify the supply variable. We advocate a lower S limit of 10 kPa. This value is analogous to the so-called field capacity. In terms of the associated 6, it is generally the wettest a soil can become because drainage then

TABLE 2 Boreal Zone Regional Soil Water Storage Capacities for 1°

Latitude by 1° Longitude Grid Cells Based

on the

Global Soil Texture Database of Zobler (1986)

with Hydraulic Parameter Values from Cosby et al. (1984)

Region (degrees of latitude (N)/longitude)

Total Area(Tm2)

LS SCL

SL

L

Fraction of total area minus wetland

area in each

mineral soil texture class

North America (48.5-68.5/58.5-163.5 W)

4.55

0.40 0.14

0.08

0.38

European Russia (52.5-68.5/22.5-59.5 E)

0.83

0.16 0.02

0.14

0.68

Western Siberia (48.5-69.5/60.5-89.5 E)

1.52

0.29 0.06

0.00

0.65

Eastern Siberia (49.5-72.5/90.5-178.5 E)

7.25

0.01 0.09

0.00

0.90

Water storage capacity,

Suctions

mm water per 0.1 m depth of soil

10 and 100 kPa

8 9

11

13

10 and 1500 kPa

13 16

19

22

Wetland areas with organic soils were excluded using the global database of Matthews (1989) as described by Matthews and Fung (1987) and in the text. Water storage capacities were derived from the difference between water contents determined in the laboratory on mineral soil cores subject to 10 and 100 or 10 and 1500 kPa of applied suction. From left to right, the mineral soil texture classes of Zobler (1986)/Cosby et ai (1984)(abbreviation, percentages of sand:silt:clay) are coarse/loamy sand (LS, 82:12:6), medium line/sandy clay loam (SCL, 58:15:27), medium coarse/sandy loam (SL, 58:32:10), and medium/loam (L, 43:39:18).

Wetland areas with organic soils were excluded using the global database of Matthews (1989) as described by Matthews and Fung (1987) and in the text. Water storage capacities were derived from the difference between water contents determined in the laboratory on mineral soil cores subject to 10 and 100 or 10 and 1500 kPa of applied suction. From left to right, the mineral soil texture classes of Zobler (1986)/Cosby et ai (1984)(abbreviation, percentages of sand:silt:clay) are coarse/loamy sand (LS, 82:12:6), medium line/sandy clay loam (SCL, 58:15:27), medium coarse/sandy loam (SL, 58:32:10), and medium/loam (L, 43:39:18).

becomes minimal. In exceptional cases, lower values of S (and higher values of 0) result from impeded drainage, rainfall intensity in excess of the soil's infiltration capacity, and groundwater intrusion. Next, we turn to the upper limit of S. Kelliher ct al. (1998) synthesized the scarce forest literature identifying how evaporation rate was limited by the available energy (see later discussion) from when S and 0 were equal to the field capacity until a drier critical value of 0 was reached. This critical value seems to be surprisingly conservative and it is equal to 0 when about half the root zone depth of soil water is depleted. (See also Choudhury and DiGirolamo (1998) who argue that the fraction is smaller at 0.4 based on a review of all available data, with about half from agricultural studies.) Thereafter, as the soil further dries, the forest evaporation rate declines sharply in response to the declining water supply. In the coarse soils they examined, the critical value of 0 corresponded to S~ 100 kPa. We adopt this value as our upper limit of S. However, for comparative purposes, we also report another, much higher value of S (and drier soil): 1500 kPa. This represents the so-called permanent wilting point, a long-advocated concept based on experiments conducted with potted sunflowers. It is essentially the driest a soil can become in terms of S and 0.

The mineral soils analyzed in Table 2 can store 8 to 13 mm of water per 0.1 m depth for S between 10 and 100 kPa. For S between 10 and 1500 kPa, these soils store about twice as much, or 13 to 22 mm of water per 0.1 m depth. Water storage capacity grades with texture from the lowest value for coarse soils to the greatest for loams. Boreal zone water-release data are available for sand and silt loam mineral soils beneath Siberian Pinas sylvcstris and Larix gmelinii forests, respectively (Fig. 3; Kelliher ct al, 1997; 1998). The well-sorted sand (i.e., the particles are mostly of a similar size) holds only 4 and 8 mm of water per 0.1 m depth for S of 10-100 and 10-1500 kPa, respectively. For the silt loam, the respective storage capacities are 7 and 14 mm. These soils are thus at the low end of the data given in Table 2. Comparable organic matter data are available for humus from the floor of a Tsuga het-erophylla and Thuja plicata forest near Vancouver, Canada (such measurements were not made for humus from the Siberian forests because of its very shallow depth at those two sites). The respective storage capacities of the humus are 7 and 9 mm (note how the humus water-release curve flattens for S between 100 and 1500 kPa; Plamandon et al, 1975). This illustrates the limited nature of organic matter or peaty soil water storage capacity for S > 10 and especially > 100 kPa. This reflects the predominance of larger pores that are mostly emptied by little or low suction in humified organic matter.

The nagging question emerges of what depth of soil is relevant to evaporation. For example, a dry surface-layer of relatively shallow depth, about 1-10 mm, dramatically reduces soil evaporation compared to the rate obtained when the surface is wet and the weather fine, especially for coarse-textured material (e.g., Kelliher et al, 1998). In terms of vegetation transpiration, plant root density declines strongly with depth as quantified by a recent global synthesis of available data (Jackson et al, 1997). Jackson et al represent the vertical distribution of fine root biomass by the power function (Y= 1 — cd), where Vis the cumulative fraction varying between 0 and 1, d the depth (cm), and c an extinction coefficient. For boreal forest and tundra vegetation, c is 0.943 and 0.909, respectively. This suggests that forest and tundra have 44 and 61%, respectively, of their fine roots in the upper 0.1 m of soil. The percentages are 69 and 85, and 83 and 93% for the upper 0.2 and 0.3 m of soil, respectively. For a depth of 0.3 m and S values of 10 and 100 kPa, the mineral soils shown in Table 2 can store 25-40 mm of water. For comparison, the corresponding value for the so-called average soil of the world is 41 mm (Nielsen et al, 1983).

Related to this limited water storage capacity of soils in the boreal zone is the climatological feature, especially in Siberia, of an abrupt transition from frozen winter to warm summer. Snow melt thus occurs relatively quickly and it involves relatively large quantities of water. Much of this water is generally lost, however, as a pulse contributor to extreme river flow rates in spring. For example, peak flow rate at the northern mouth of the Yenesei River in central Siberia was once a remarkable 70,000 m3 s-1 (Beckinsale, 1969). This loss of winter precipitation in spring may result from surplus water unable to seep into the still-frozen soil (Walter, 1985). If the soil is not frozen, high-volume drainage may bypass the matrix through macropores (Clothier et al. (1998) review the pertinent physics involved.)

Summer days in the boreal zone are long, the daily period of illumination increasing virtually to 24 h in the north. Radiation is a primary driving variable of earth surface-atmosphere energy exchange including evaporation. On a clear day, for an unpolluted atmosphere, shortwave irradiance at the earth's surface obtains a maximum of around 70-80% of the extraterrestrial value due to

TO T3

TO V

TO T3

TO V

180 200 220 Day of the year

FIGURE 4 (A) Relations between daily clear-sky shortwave irradiance at the earth's surface, computed with the atmospheric transmissivity = 0.7, and latitude from 50 to 70°N on 1 June (dotted line), 1 July (short dashes line), 1 August (long dashes line) and 31 August (solid line). (B) Relation between the ratio of daily clear-sky shortwave irradiance and latitude ( i.e., slopes of lines like those shown in (A)) and day of the year from 1 June (day 152) through 31 August (day 243).

attenuation by water vapor and dust (i.e., atmospheric transmis-sivity = 0.7-0.8). For a clear sky during summer (June-August) in the boreal zone, daily shortwave irradiance varies significantly with a peak on 21 June and a minimum on 31 August (e.g., 22.0 and 12.3 MJ m-2 day-1 at 60°N, respectively, for these two days with atmospheric transmissivity = 0.7.) Irradiance decreases with increasing latitude but the rate of change is not constant in the boreal zone (Fig. 4A). Differences in irradiance across the boreal latitudes are least in June around the summer solstice and greatest in terms of decline during August. From linear regression of irradiance and latitude, the offset is relatively constant at 34-36 MJ m-2 day-1 but the slope is least on 21 June at —0.21 MJ m-2 day-1 degree-1, declining linearly thereafter to —0.39 MJ m-2 day-1 degree-1 on 31 August (Fig. 4B). The boreal zone, of course, does not always have a completely clear sky. During fine summer weather in Northern Canada, an additional 15% attenuation of shortwave radiation by smoke aerosols from forest fires has been reported (Miller and O'Neill, 1997). On a completely overcast day, the atmospheric transmissivity would typically be only around 0.25. As an example of the integrated effect of atmospheric radiation attenuation, for June-August 1997 at a relatively sunny aspen forest site in Northern Canada (53.7°N, 106.2°W ), the measured shortwave irradiance averaged 80% of the clear-sky value computed for an atmospheric transmissivity of 0.8 (T.A. Black, personal communication).

3. Evaporation Theory

The evaporation rate (E) obtained for an extensive, wet surface in dynamic equilibrium with the atmosphere and in the absence of advection (£eq) may be written (Slatyer and Mcllroy, 1961; Mc-Naughton, 1976) as

180 200 220 Day of the year

FIGURE 4 (A) Relations between daily clear-sky shortwave irradiance at the earth's surface, computed with the atmospheric transmissivity = 0.7, and latitude from 50 to 70°N on 1 June (dotted line), 1 July (short dashes line), 1 August (long dashes line) and 31 August (solid line). (B) Relation between the ratio of daily clear-sky shortwave irradiance and latitude ( i.e., slopes of lines like those shown in (A)) and day of the year from 1 June (day 152) through 31 August (day 243).

where e is the rate of change of latent heat content of saturated air with change in sensible heat content, A the latent heat of vaporization, and Ra the available energy flux density. The partitioning of Ra into E according to Eq. (1) is strongly temperature-dependent, especially through term e (Fig. 1). All of the available energy may thus be dissipated by evaporation if E = £eq at the so-called partitioning temperature of 33°C (Priestley, 1966; Priestley and Taylor, 1972; Calder et al, 1986) including consideration of the entrain-ment of dry air from aloft into the convective boundary layer (De Bruin, 1983). In addition to temperature and R.d, it also takes time for E —» Eeq (McNaughton and Jarvis, 1983; Finnigan and Raupach, 1987), but daily or longer periods are sufficient.

To illustrate the application of Eq. (1), we can conduct a global average annual computation. The global average surface temperature is 15°C according to Graedel and Crutzen (1993). At this temperature, A = 2.465 J g-1 and e/(e + 1) = 0.63. For the earth's continents, Baumgartner and Reichel (1975) estimate RJ\ = 850 mm year-1 so that Eeq = 535 mm year-1. This may be compared to their continental £ = 480 mm year-1 independently determined and constrained by conservation of mass according to a global water balance including the oceans. Consequently, for the terrestrial biosphere as a whole, £nj is within 10% of E and therefore not significantly different, given a reasonable error associated with the computations.

Equation (1) may seem a purely meteorological model but it does not ignore the evaporating surface, particularly in its application to the terrestrial biosphere including vegetation and soil (Priestley and Taylor, 1972). First and foremost, the surface affects Rit by determining the shortwave radiation reflection coefficient or albedo, which a recent review showed varied by a factor of 2 in the boreal zone (Baldocchi et al., 2000). Albedo depends also on ground-surface wetness (i.e., darkness), which is particularly relevant to the boreal zone because of the commonly sparse nature of the vegetation, and on the solar zenith angle because of vegetation architecture effects. For the boreal zone, it is thus relevant to note how fire can alter vegetation structure (Wirth et al., 2000). Second, the surface temperature, also used in Eq. (1) to determine e and A, largely determines the outgoing longwave radiation. Surface temperature in turn depends on Rt and its partitioning into AE and sensible heat (H), heat that you can sense because it warms the air, via the surface energy balance (R„ = A£ + H). Aerodynamic roughness of vegetation or of ground affects the surface temperature as well. In addition, for outgoing longwave radiation, a dry mineral or plant-litter surface will have a significantly lower emissivity than vegetation or wet ground (soil or litter) (Monteith and Unsworth, 1990). Vegetation density determines the amount of net all-wave irradiance that is absorbed. Ground surface wetness (thermal admittance) also affects the fraction of the remaining energy dissipated by conduction into the ground. The rate of conduction can be variably, and overwhelmingly, affected by underlying ice during summer in the boreal zone (e.g., Fitzjarrald and Moore, 1994).

For plant leaves, the surface may be considered explicitly by a model of E or transpiration rate (£,);

where D0 is air saturation deficit, expressed as a dimensionless specific-humidity deficit, at the leaf surface and gst is the stomatal conductance for water vapor transfer. It is gsl, rather than £,, that has been the focus of most research, although some careful studies including D0 have also been conducted (Mott and Parkhurst, 1991; Alphalo and Jarvis, 1991; 1993; Monteith, 1995). Using Eq. (2) and rearranging Eq. (1), we may write an equation for the terrestrial surface conductance for water vapor transfer (gs) as gs=[e/Me + l)]RJD0. (3)

Written similarly, gs is also called the climatological conductance (Monteith and Unsworth, 1990). In any case, like gsl, gs represents the surface control of E by balancing radiant energy supply and atmospheric demand (Du). Consequently, as shown below, evalu ating terrestrial E in relation to £eq includes an assessment of the underlying surface as well as the meteorology and water balance.

The value of £ in relation to £eq, like the value of gs, indicates the evaporative nature of the surface so £ < £eq or £/£eq < 1 reflects surface dryness or stomatal closure as well as the balance of energy exchange between the atmosphere and the underlying surface. By definition, £ > £eq can be caused only by advection. As implied above with respect to the partitioning of temperature, this may also result from the entrainment of dry air from above the convective boundary layer that develops daily over the earth surface. To further illustrate the relation between £ and £eq in terms of surface characteristics, it is helpful to write the Penman-Monteith equation (Monteith and Unsworth, 1990),

where p is the density of air, D is the air saturation deficit above the evaporating surface expressed as a dimensionless specific-humidity deficit, and gA is the total aerodynamic conductance assuming similarity between heat and water vapor transfer processes in the atmosphere. Some limits of this equation suggest that there are at least three surface-related conditions leading to £ = £eq (Raupach, 2001), namely:

(i) gs * x by definition for a completely wet surface with the corollary D —» 0 (Slatyer and Mclllroy, 1961), although this requires no entrainment of dry air from above the near-surface boundary layer (i.e., the closed box model of McNaughton and Jarvis, 1983),

(ii) gA * 0 by definition for a surface beneath a completely calm atmosphere (Thom, 1975), and

(iii) £ is not sensitive to gA (Thom, 1975) for a completely smooth surface or a surface completely isolated from the influence of D on £, a derivative of (ii).

Condition (i) depends mostly on precipitation frequency although precipitation interception, hydraulic conductance characteristics of soil, ground surface (e.g., litter), and plants, and ground water storage capacity also contribute. Conditions (ii) and (iii) depend on the surface roughness, mostly reflecting vegetation height or a lack of it. For all conditions, larger-scale meteorology is relevant too, as are the physical and physiological feedback processes critically analyzed recently by Raupach (1998).

Returning to plant leaves, through stomata, there is an intrinsic connection between £t and the net rate of carbon assimilation (A) that may be written as

where Ca and C0 are carbon dioxide concentrations in the atmosphere and substomatal cavity, respectively. Division of gst by 1.6 accounts for the difference in diffusion coefficients for water vapor and carbon dioxide (Massmann, 1998). Further, the maximum A or carbon assimilation capacity of the leaf has been found to be proportional to the leaf nitrogen content (Field and Mooney,

1986; Evans, 1989). According to Eq. (5) and when a wide range of plants are compared, maximum git is also proportional to leaf nitrogen content [Schulze et al, 1994; gst correlates with carbon assimilation capacity (Wong et al., 1979)]. In this way, stomata link water, carbon, and nutrient cycles. It is no wonder then that a voluminous literature reflects a virtually universal interest in stom-atal behavior among plant physiologists and biometeorologists (Körner, 1994).

Was this article helpful?

0 0

Post a comment