J. Romanya et al., Short-term effects of fire on soil nitrogen availability in Mediterranean grasslands and shrublands growing in old fields, FOREST ECOL, 147(1), 2001, pp. 39-53
In three different plant communities growing in Mediterranean old fields we
studied the short-term changes in soil nitrogen availability that occur af
ter the fire. Two of these communities were grasslands with great capacity
of resprouting and contrasted N availability, one dominated by Brachypodium
retusum, and the second one dominated by B. retusum and the N fixing shrub
Genista scorpius. The third community was an obligate seeder community (sh
rubland) with low N availability and was dominated by Rosmarinus officinali
s. We selected six plots for each type of vegetation and therefore performe
d 18 experimental fires. During fires we measured temperatures at the soil
surface. Maximum temperature recorded during fire and time-temperature inte
gral were used as indexes of fire severity. During the 6 months following f
ires we measured Net N mineralization and plant uptake by field incubations
using the resin-core technique in paired burnt and control plots.
Fire severity increased with plant biomass. In grasslands heating of the so
il surface increased with plant biomass up to a Limit of 1 kg m(-2) of abov
e-ground biomass. For high biomass a large proportion of heat released duri
ng fire was probably transmitted to the atmosphere or to the deeper soil ho
rizons. The increase of soil mineral N was larger in fires of greater sever
ity. Most mineral nitrogen released to the soil during fire was ammonia. In
creases of ammonia post-fire depends on the temperatures measured on the so
il surface while increases of the less volatile N form (nitrate) were relat
ed to the amount of burnt biomass and were highly dependent on the type of
vegetation.
The amount of nitrogen released to soil during fire represented a small pro
portion of the N mineralized during the 6 months following fire and thus th
e amount of nitrogen mineralized per unit of N released during fire was ver
y different across the different types of vegetation. In grasslands fire in
duced changes in N mineralization decreased as fire severity increased. In
contrast, in shrublands we observed the opposite trend. Differences in pote
ntially mineralizable and in net mineralization N between unburnt grassland
s and shrublands could account for this fact. Despite the depression in nit
rification that we observed in grasslands between 40 and 80 days after the
fire, high nitrate concentration in the soil during that period increased N
leaching in burnt plots. No plant uptake was detected at that time. In gra
sslands the onset of plant uptake in burnt plots was delayed as compared to
control. Cumulative changes in N did not depend on the burnt biomass in gr
assland communities, but it did in the seeder community. On the contrary, s
oil temperatures measured during fires related to changes in N observed in
grasslands but not in the seeder community. It appears therefore, that post
-fire N mineralization and leaching in grasslands may have been driven by t
he changes induced by heating the soil surface while in shrublands it may h
ave been driven by the quantity of ash deposited on the soil surface. (C) 2
001 Elsevier Science B.V. All rights reserved.