154
AnnaSala
very low during cold months due to decreased evaporative demand, stomatal
conductance and photosynthesis rates were relatively high. Significant carbon input during fair winter conditions in Mediterranean sclerophylls was
also measured by Larcher and Tisi (1990). At the whole canopy level, simulated peak canopy carbon input during clear autumn days ranged from 11.3
to 15 flmol m- 2 s-I, values comparable with those reported by Valentini et al.
(1991).
Simulations of monthly transpiration and carbon input in holm oak canopies at the ridge top and valley bottom sites of the Avic catchment (Fig. 11.3)
were done by selecting representative days from detailed meteorological records available during 1989 at each site (Sala and Tenhunen 1996). The value
of gF for each month was determined from annual courses of \I'pd at each site
and the linear dependency between gF and 'Ppd' Because of reduced light intensity, the contribution of shade leaves deep in the canopy to canopy water
loss and carbon input is very small; thus canopy totals at the ridge top (LAI =
4.6 m 2 m- 2 ) and valley bottom (LAI = 5.3 m 2 m- 2 ) during periods of adequate
water availability were very similar. Stands at the valley bottom used more
water during the critical period in late spring and early summer, which led to
severe soil water depletion and summer water stress at the valley bottom site
(particularly during August; Sala and Tenhunen 1994). A decline in measured catchment streamflow following increases in precipitation suggests that
the basin was not fully recharged and that soil water may have been severely
depleted (e.g. August-September). Severe soil water depletion occurred during late spring and early summer (May-June) when simulated canopy transpiration was at a maximum and precipitation was low. The result was severe
water stress during August and September when trees exhibited very conservative gas exchange behaviour (Fig. 11.3; Sala and Tenhunen 1994). Periods
of high precipitation and low canopy transpiration allowed the soil to recharge and an increase in measured streamflow was measured.
Estimated total annual canopy transpiration was 464 mm year- 1 at the
valley bottom site and 453 mm year- 1 at the ridge top. This represents 85 and
87% of the annual precipitation during 1989 (534 mm) at the ridge top and
valley bottom sites, respectively. These values differ less than 10% from the
ones reported by Pinol (1990) by using the input/output hydrological balance
method. While this is not a validation, these results suggest, once more, that
model predictions of water use by holm oak canopies are realistic and consistent with measured hydrological parameters.
The results summarized here (see also Sala and Tenhunen 1996) demonstrate that the modelling approach used at the Avic catchment (which is
based on detailed field studies) provides a realistic description of diurnal
and seasonal patterns of leaf and canopy response to water availability. By
using the same approach we can now address specific questions regarding
the relative effect of physiological and structural adjustments to drought on
the water loss and carbon input of holm oak canopies.
AnnaSala
very low during cold months due to decreased evaporative demand, stomatal
conductance and photosynthesis rates were relatively high. Significant carbon input during fair winter conditions in Mediterranean sclerophylls was
also measured by Larcher and Tisi (1990). At the whole canopy level, simulated peak canopy carbon input during clear autumn days ranged from 11.3
to 15 flmol m- 2 s-I, values comparable with those reported by Valentini et al.
(1991).
Simulations of monthly transpiration and carbon input in holm oak canopies at the ridge top and valley bottom sites of the Avic catchment (Fig. 11.3)
were done by selecting representative days from detailed meteorological records available during 1989 at each site (Sala and Tenhunen 1996). The value
of gF for each month was determined from annual courses of \I'pd at each site
and the linear dependency between gF and 'Ppd' Because of reduced light intensity, the contribution of shade leaves deep in the canopy to canopy water
loss and carbon input is very small; thus canopy totals at the ridge top (LAI =
4.6 m 2 m- 2 ) and valley bottom (LAI = 5.3 m 2 m- 2 ) during periods of adequate
water availability were very similar. Stands at the valley bottom used more
water during the critical period in late spring and early summer, which led to
severe soil water depletion and summer water stress at the valley bottom site
(particularly during August; Sala and Tenhunen 1994). A decline in measured catchment streamflow following increases in precipitation suggests that
the basin was not fully recharged and that soil water may have been severely
depleted (e.g. August-September). Severe soil water depletion occurred during late spring and early summer (May-June) when simulated canopy transpiration was at a maximum and precipitation was low. The result was severe
water stress during August and September when trees exhibited very conservative gas exchange behaviour (Fig. 11.3; Sala and Tenhunen 1994). Periods
of high precipitation and low canopy transpiration allowed the soil to recharge and an increase in measured streamflow was measured.
Estimated total annual canopy transpiration was 464 mm year- 1 at the
valley bottom site and 453 mm year- 1 at the ridge top. This represents 85 and
87% of the annual precipitation during 1989 (534 mm) at the ridge top and
valley bottom sites, respectively. These values differ less than 10% from the
ones reported by Pinol (1990) by using the input/output hydrological balance
method. While this is not a validation, these results suggest, once more, that
model predictions of water use by holm oak canopies are realistic and consistent with measured hydrological parameters.
The results summarized here (see also Sala and Tenhunen 1996) demonstrate that the modelling approach used at the Avic catchment (which is
based on detailed field studies) provides a realistic description of diurnal
and seasonal patterns of leaf and canopy response to water availability. By
using the same approach we can now address specific questions regarding
the relative effect of physiological and structural adjustments to drought on
the water loss and carbon input of holm oak canopies.