Soil Nitrogen Dynamics
229
were followed by high rates of N immobilization in September and October.
At Montseny, net N mineralization in the upper mineral soil was weakly but
positively correlated with soil temperature (r = 0.47, P < 0.05, n = 23). Net
nitrification showed more marked seasonality than net N mineralization,
probably because of the high environmental sensitivity of nitrifier populations (Schaefer 1973; Johnson 1992). Nitrification showed a positive relationship with soil moisture, both at Montseny (r = 0.52, P < 0.05, n = 23) and at
Prades (r = 0.82, P < 0.001, n = 12) in the surface horizons.
Considering the mineral soil of 0-20 cm depth, the Prades forest had on a
surface area basis a much lower mean net N mineralization than Montseny:
2.7 versus 6.7 kg N ha- l month-I, respectively. On an annual basis, about half
of the mineralized N at Montseny was produced in each of the two mineral
soil layers, whereas at Prades 71 % of the net N mineralization in the mineral
soil took place in the 5-20 cm layer (Fig. 16.3). These results imply that the
lower mineral soil cannot be ignored in ecosystem N dynamics.
Net nitrification was low in both holm oak forests and occurred mostly in
the forest floor and in the top mineral soil. At Prades, net nitrification in the
H horizon averaged 0.40 kg N ha- I month-I, and represented 52% of the
nitrified N in the soil profile (H and 0-20 cm of mineral soil). In the mineral
soil (0-20 cm), mean net nitrification rates were lower at Prades than at
Montseny (0.37 and 0.63 kg N ha- l month-I, respectively).
16.5 Nitrogen Leaching
At Montseny, the movement of mineral N in solution through the soil profile
was studied in a 3-m-Iong trench in the permanent plot at La Castanya
(Chap. 2) using zero-tension throughflow collectors placed below the H horizon and at 160-cm depth at the bedrock contact (Avila et al. 1995). At Prades,
and in the context of the burning and dear-cutting experiment referred to
above, percolating water was collected in zero-tension lysimeters under the
forest floor, and at 10- and 30-cm depths in the mineral soil (the latter being
the local bedrock contact). Leached N was also collected using anionic and
cationic exchange resins (Binkley and Matson 1983) in the same Prades site.
At Prades the maximum nitrate concentrations in the water leached from
all depths occurred in summer, whilst other peaks occurred in spring and
autumn. Concentrations were lowest in winter when the soil nitrate content
was also very low. The amount of nitrate collected in resins was strongly related to the volume of rainfall during the incubation period (r = 0.75,
P < 0.000l), so leaching of nitrate occurred during the major rainfall periods
(spring and autumn). At Montseny, nitrate concentrations in throughflow
water were highly variable, with little seasonal pattern (Avila et al. 1995).
Figure 16.4 shows the mean nitrate concentrations as water percolates
from rainfall to streamwater through each holm oak ecosystem. Infiltrated
229
were followed by high rates of N immobilization in September and October.
At Montseny, net N mineralization in the upper mineral soil was weakly but
positively correlated with soil temperature (r = 0.47, P < 0.05, n = 23). Net
nitrification showed more marked seasonality than net N mineralization,
probably because of the high environmental sensitivity of nitrifier populations (Schaefer 1973; Johnson 1992). Nitrification showed a positive relationship with soil moisture, both at Montseny (r = 0.52, P < 0.05, n = 23) and at
Prades (r = 0.82, P < 0.001, n = 12) in the surface horizons.
Considering the mineral soil of 0-20 cm depth, the Prades forest had on a
surface area basis a much lower mean net N mineralization than Montseny:
2.7 versus 6.7 kg N ha- l month-I, respectively. On an annual basis, about half
of the mineralized N at Montseny was produced in each of the two mineral
soil layers, whereas at Prades 71 % of the net N mineralization in the mineral
soil took place in the 5-20 cm layer (Fig. 16.3). These results imply that the
lower mineral soil cannot be ignored in ecosystem N dynamics.
Net nitrification was low in both holm oak forests and occurred mostly in
the forest floor and in the top mineral soil. At Prades, net nitrification in the
H horizon averaged 0.40 kg N ha- I month-I, and represented 52% of the
nitrified N in the soil profile (H and 0-20 cm of mineral soil). In the mineral
soil (0-20 cm), mean net nitrification rates were lower at Prades than at
Montseny (0.37 and 0.63 kg N ha- l month-I, respectively).
16.5 Nitrogen Leaching
At Montseny, the movement of mineral N in solution through the soil profile
was studied in a 3-m-Iong trench in the permanent plot at La Castanya
(Chap. 2) using zero-tension throughflow collectors placed below the H horizon and at 160-cm depth at the bedrock contact (Avila et al. 1995). At Prades,
and in the context of the burning and dear-cutting experiment referred to
above, percolating water was collected in zero-tension lysimeters under the
forest floor, and at 10- and 30-cm depths in the mineral soil (the latter being
the local bedrock contact). Leached N was also collected using anionic and
cationic exchange resins (Binkley and Matson 1983) in the same Prades site.
At Prades the maximum nitrate concentrations in the water leached from
all depths occurred in summer, whilst other peaks occurred in spring and
autumn. Concentrations were lowest in winter when the soil nitrate content
was also very low. The amount of nitrate collected in resins was strongly related to the volume of rainfall during the incubation period (r = 0.75,
P < 0.000l), so leaching of nitrate occurred during the major rainfall periods
(spring and autumn). At Montseny, nitrate concentrations in throughflow
water were highly variable, with little seasonal pattern (Avila et al. 1995).
Figure 16.4 shows the mean nitrate concentrations as water percolates
from rainfall to streamwater through each holm oak ecosystem. Infiltrated