Gas Exchange and Water Relations
143
than 30 jllll were embolised and the resultant losses in hydraulic conductivity
were about 30%. At this stage, recovery was easy with a moderate water supply. With increasing water stress, losses in hydraulic conductivity reached
85% of the initial value, and the percentage of recovery fell dramatically (Lo
Gullo and Salleo 1993). In the permanent plot at La Castanya, the ratio (leaf
water potential)/(turgor loss point water potential) was usually close to 0.9,
indicating a low probability of embolism due to drought stress. At higher altitudes (above 1000 m a.s.l.) this ratio increased to 1.9-2.5, which can explain
the deleterious effects of drought stress on oak growth and survival (Tyree
and Dixon 1986).
Xylem embolism can occur in winter as well. Holm oak plants exposed to
an air temperature of -2.5 °C showed a 50% loss of hydraulic conductivity, a
loss that increased to 94% at -11 °C (Sakai and Larcher 1987). The recovery
rate was negatively correlated with the intensity and duration of exposure to
cold. The above temperatures are not uncommon throughout the geographical range of holm oak (Terradas and Save 1992) and they can limit the
growth and survival of holm oak at high altitudes.
10.9 Canopy Transpiration and Ecosystem Water Budgets
Annual water consumption by entire forests is difficult to determine directly.
From water budgets in our small experimental catchments at Montseny and
Prades, total evapotranspiration in dense holm oak forests is estimated to average 500-600 mm year- 1 (Chap. 19). This estimate not only refers to canopy
transpiration but also includes interception by the canopy (Chap. 15), interception by the forest floor, and direct evaporation from the soil. The model
described in Chapter 11 predicts a canopy transpiration of 450-460 mm for
1989 (a dry year) in two plots of contrasted topographic position within the
Avic catchment at Prades. On a site with high soil water availability, such as
the mesic permanent plot at La Castanya, integration of leaf transpiration
rates yielded a canopy transpiration of 628 mm in 1980 (Save 1986), again a
dry year. Despite being based on different methods, sites and years, these estimated water requirements agree with the geographical distribution of holm
oak forests, which are dominant in humid and sub-humid Mediterranean
climates. Nevertheless, in Mediterranean areas there is a great interannual
variability of precipitation, and even in sites whose average rainfall lies above
the water requirements of holm oak forests, years with much lower water
availability are frequent.
143
than 30 jllll were embolised and the resultant losses in hydraulic conductivity
were about 30%. At this stage, recovery was easy with a moderate water supply. With increasing water stress, losses in hydraulic conductivity reached
85% of the initial value, and the percentage of recovery fell dramatically (Lo
Gullo and Salleo 1993). In the permanent plot at La Castanya, the ratio (leaf
water potential)/(turgor loss point water potential) was usually close to 0.9,
indicating a low probability of embolism due to drought stress. At higher altitudes (above 1000 m a.s.l.) this ratio increased to 1.9-2.5, which can explain
the deleterious effects of drought stress on oak growth and survival (Tyree
and Dixon 1986).
Xylem embolism can occur in winter as well. Holm oak plants exposed to
an air temperature of -2.5 °C showed a 50% loss of hydraulic conductivity, a
loss that increased to 94% at -11 °C (Sakai and Larcher 1987). The recovery
rate was negatively correlated with the intensity and duration of exposure to
cold. The above temperatures are not uncommon throughout the geographical range of holm oak (Terradas and Save 1992) and they can limit the
growth and survival of holm oak at high altitudes.
10.9 Canopy Transpiration and Ecosystem Water Budgets
Annual water consumption by entire forests is difficult to determine directly.
From water budgets in our small experimental catchments at Montseny and
Prades, total evapotranspiration in dense holm oak forests is estimated to average 500-600 mm year- 1 (Chap. 19). This estimate not only refers to canopy
transpiration but also includes interception by the canopy (Chap. 15), interception by the forest floor, and direct evaporation from the soil. The model
described in Chapter 11 predicts a canopy transpiration of 450-460 mm for
1989 (a dry year) in two plots of contrasted topographic position within the
Avic catchment at Prades. On a site with high soil water availability, such as
the mesic permanent plot at La Castanya, integration of leaf transpiration
rates yielded a canopy transpiration of 628 mm in 1980 (Save 1986), again a
dry year. Despite being based on different methods, sites and years, these estimated water requirements agree with the geographical distribution of holm
oak forests, which are dominant in humid and sub-humid Mediterranean
climates. Nevertheless, in Mediterranean areas there is a great interannual
variability of precipitation, and even in sites whose average rainfall lies above
the water requirements of holm oak forests, years with much lower water
availability are frequent.