By
J.I. House & D.O. Hall
Division of Life Sciences, King's College
London
|
|
|
|
| Leaf
Area Indices Albedo and radiation Roughness, conductance and resistance
|
Carbon
dioxide fluxes Production efficiency ACKNOWLEDGEMENTS
|
|
 |
BIOPHYSICAL PROPERTIES, FLUXES AND EFFICIENCIES | |
|
Biophysical properties, fluxes and
efficiencies vary enormously and are radically affected by the high
seasonality of savannas and fire. The few sites where such measurements have
been made in any detail are highlighted in Table 5.
|
||
|
Leaf Area Indices | |
|
LAI is highly seasonal e.g. in Lamto the
LAI increases from 0 after fire in January to a peak of 4.0 in November (Le
Roux et al, 1997). Values for Venezuela and the UNEP sites also represent
the complete seasonal ranges, while the Niger study captured part of the wet
and dry season. The Nylsvley study incorporated trees with measured LAI
under canopies (tree + grass, 2.25) and between canopies (grass only, 0.6)
giving an overall LAI of 1.2 (Scholes & Walker, 1993). The Brazil study
also measured tree LAI but figures seem relatively low (Miranda et al,
1997). Thus, according to this limited data set, the range in arid savannas
is 0 to 1.7 (or 2.3 if trees are included), and the range in humid savannas
is 0 to 4.8 (and could be higher with the inclusion of trees).
|
||
|
Albedo and radiation | |
|
Albedo is a function of underlying soil
as well as vegetation, and fires can darken the soil surface and affect its
spectral qualities. Albedo in Lamto was 0.07 over bare black soil, and
increased over open areas with low shrub from 0.08 to a peak of 0.23, with
an overall mean annual albedo of 0.194 (albedo in more shrubby areas was
similar) (Le Roux et al, 1994). In contrast, bare soil albedo at Niger was
0.4, greater than the average vegetation albedo of 0.2 (Braud et al, 1997).
Similarly in Nylsvley, with a light soil, albedo was at its lowest value of
0.11 over peak biomass, and rose to 0.15 at the end of the dry season, with
an annual average of 0.12. San Jose (1992) found that dry-season albedo was
lower in recently burned areas 0.08 compared to unburned areas 0.12, because
the smother burned surfaces trapped radiation more efficiently while
vegetation scattered it. In spite of this net radiation was also lower due
to higher emission of long-wave radiation as a consequence of higher daytime
mean temperature in the surface of the bare soil. Thus, in the examples
found, peak vegetation albedo varies from around 0.12 in Nylsvley/Venezuela
to 0.2 in Lamto/Niger, while bare soil albedo varies from 0.07 in Lamto to
0.4 in Niger.
|
||
|
Roughness, conductance and resistance | |
|
Spatial heterogeneity between different
patches of savanna vegetation causes a discontinuity in aerodynamic,
radiative, thermal and moisture features. This is exacerbated by fire as air
flows from the smooth and hot recently burned savanna (bare soil) to a
rougher and cooler unburned savanna (vegetative surface). San Jose (1992)
showed that the roughness length and aerodynamic conductance were far lower
in burned patches (Table 5), and that mixing due to thermal turbulence in
the burned savanna was 3 times greater than that due to roughness generated
turbulence in the unburned savanna. In Brazil and Venezuela (site 2) surface
conductance was higher with more leaves in the canopy, and the degree of
coupling (W ) was higher indicating less coupling between canopy and
atmosphere (San Jose et al, 1998; Miranda et al, 1997). [W approaches 1 in
well-watered and aerodynamically smooth canopies, where the transpiration
rate is driven by radiation, and approaches 0 over aerodynamically rough
canopies, where transpiration is driven by canopy to air saturation
deficit]. While roughness values in Niger were similar to those for the
unburned Venezuelan grassland site (Braud et al, 1997), the sparse tree
crowns of the Brazilian cerrado vegetation form a more aerodynamically rough
surface.
|
||
|
Carbon dioxide fluxes | |
|
Several studies have been carried out on leaf photosynthetic rate (or CO2 flux) of savanna grass species showing a range of 15 to 33 m molCO2/m2/s across a range of dry and humid sites with results showing a high variability (Le Roux & Mordelet, 1995 and Table 5). In Nylsvley alone, measurements for one grass species varied by 13-fold, dependent partly on techniques used and local conditions, although for further calculations, Scholes & Walker (1993) used an overall average of 25 m molCO2/m2/s. Some measurements were also made of tree leaf photosynthesis at Nylsvley and, on average, the C3 woody plants had 18% lower rates than the C4 grasses which are better adapted to high irradiation and heat and water stress. Primary productivity is much more closely linked to canopy CO2 flux than leaf flux, yet there are even fewer savanna studies on this. At Lamto, Le Roux & Mordelet (1995) measured CO2 canopy fluxes at the beginning of the wet season after a January fire. Despite low N at this site, the peak net canopy CO2 flux/assimilation F (measured above the canopy) was high at 24 m molCO2/m2/s for a LAI of only 1.9 not long after fire. They compare this to net canopy fluxes of 27 m molCO2/m2/s for LAIs of 7 and 8 in Amazonian and Malaysian rainforests (Fan et al, 1990; Aoki et al, 1975) sustaining the emerging opinion that the primary productivity of tropical savannas could be close to that of tropical forests (Atjay et al, 1979; Gifford, 1980; Long et al, 1989). Net canopy fluxes measured in Niger (LAI 1-1.25, Hannan et al, 1998) were similar to those found in the wet season in Brazil (LAI 0.5, Miranda et al, 1997) at 15 m mol/m2/s into the canopy during the day and –5 m mol/m2/s out at night. During the wet season in Brazil, the fluxes were higher and while vegetation was found to be a CO2 sink in the wet season of up to 0.15 mol/m2/d, it was a source of CO2 for a brief period at the height of the dry season (Miranda et al, 1997). San Jose et al (1991) measured CO2 fluxes in the wet season in Venezuela and found that the net community assimilation ranged from 6.6 - 7.9 g (DM) /m2/d which was similar to the mean community growth rate they calculated from biomass measurements (2.8 - 6.9 g (DM) /m2/d).
|
||
|
Production efficiency | |
|
Several different methods are used to
calculate efficiency, and some of these are represented in Table
5.
Radiation Use Efficiency (RUE) is calculated as total NPP/IPAR (incident
photosynthetically active radiation) and we have tried to convert all values
to this. IPAR can be assumed to be a fixed proportion of incident radiation
(Rs), usually between 45 and 50%. The range of RUE is 0.4
to 1.8 taking the annual results only (i.e. not Niger) and ignoring the
outlier Brazil. RUE is partly related to rainfall with the average across
the three semi-arid sites being 0.43, and the two humid sites 1.65. Many
arid species have a lower conversion efficiency as lower stomatal
conductance is an adaptation to water stress. The Lamto results show how
efficiency declines as the stand matures. In Thailand, one year of fire
reduced RUE from 1.47 to 1.35 and a second year of fire caused it to drop
further to only 0.17 (assuming IPAR = 0.45 Rs).
|
||
|
|
ACKNOWLEDGEMENTS | |
|
Many thanks to Bob Scholes (CSIR, South Africa), Xavier Le Roux (INRA, Clermont-Ferrand, France), Jonathan Scurlock (ORNL, USA) and Joe Scanlan (Department of Natural Resources, Queensland, Australia) for providing information and making corrections to the manuscript. Dale Kaiser & Sonja Jones (ORNL, USA) for calculating tropical % of the Olson et al (1983) "grasslands" category. Sadly, David Hall passed away in August 1999 before this chapter was published. His knowledge and love of savannas was only surpassed by his eagerness to learn and teach.
|
||
|
|
REFERENCES |
|
Table 1: Previous estimates of area, biomass and NPP of savannas and grasslands |
|
Table 2: Broad plant functional types found in African savannas (from Scholes et.al., 1997) |
|
Table 3: Biomass reported for tropical grasslands and savannas | |
|
Table 4: Primary production reported for tropical grasslands and savannas |
|
Table 5: Biophysical properties, fluxes and efficiencies |
|
Figure 2: The relationship between total NPP and aboveground NPP |
To read the complete "Chapter" by J. I. House & D. O. Hall, click on "Previous" or "Next"
 |
|
|
|
If you have entered this site from
UNEP.NET or SCOPE, then feel free to use the link below to view the full
site. 