Vegetation and Disturbance - Environmental Modelling: Finding Simplicity in Complexity - page 227

composition of eucalypt forests and showed that fire and
grazing accounted for more variation than the combina-
tion of all other environmental and spatial variables.
Although they are perhaps the two most important
sources of disturbance, they differ very much in char-
acter; the former causes an abrupt discontinuity in the
system while the latter is characterized by a gradual and
continuous process of selective biomass removal.

of these obligate seeders to fire in terms of projected
plant cover is much slower than that of the resprouters.

In general, tree survival can vary according to the inten-
sity of the fire, the height of the crown and bark thickness
(Ryan and Reinhardt, 1988). Other considerations, such
as fire intensity, fire season and water availability at
the time of the fire and during the immediate post-fire
period can be of great importance in determining rate of
recovery for seeders and, to a lesser extent, for sprouters
(Moreno and Cruz, 2000). The shade tolerance of species
and light conditions after fire can also be fundamental in
determining the post-fire recovery of vegetation.
In most cases, however, plants are 'top-killed' by fire;
plant succession in fire-adapted systems, such as for
example in most Mediterranean communities, consists
largely of the recovery of the species present before fire
through their respective life-cycle and growth processes.
This is autosuccession (Hanes, 1971) in which the burned
stand, although initially appearing to be different from
the pre-burned one, retains its floristic identity in time.

14.1.1 Fire

The immediate effects of fire are the removal of the
majority of above-ground biomass and the generation of
a pulse of heat through the upper soil horizons. Some
plants are killed, but those that survive and can resprout
are changed physiologically because of the imbalance
in the root:shoot ratio. The exposure of the modified
soil horizons results in a new post-fire microclimate
and changes in hydrology. However, the resilience of
vegetation to fire has been reported for several ecosys-
tems. For example, Quercus coccifera garrigue and Ulex
parviflorus and Rosmarinus officinalis shrublands regained
their total cover in less than five years after burning (Fer-
ran and Vallejo, 1998). Similarly, few changes following
fire were reported for oak scrub in Florida uplands where
species composition and structure largely recovered in less
than two years (Abrahamson and Abrahamson, 1996). In
general, after fire, the most obvious change is that of
the composition and relative abundance of herbaceous
species. As succession proceeds and the canopy closes, the
early flush of herbaceous vegetation is either restricted to
small canopy openings or remains dormant in the soil
in the form of seeds waiting for the next fire to come
(Arianoutsou, 1998). Similar results were found in Cali-
fornia (Hanes, 1971), in Australia (Specht et al ., 1958), in
Italy (Mazzoleni and Pizzolongo, 1990), and in Portugal
(Espırito-Santo et al ., 1993).
Two broad groups of communities can be distinguished
related to the reproductive strategy of the dominant
plant species and their comparative rates of plant cover
regeneration after fire:

14.1.2 Grazing

Grazing effects on vegetation have been studied in many
ecosystems. It is clear that herbivores cause changes in
species richness (Wimbush and Costin, 1979; Basset,
1980; Bakker et al ., 1983; Persson, 1984).
An important parameter of grazing disturbance is the
specificity of grazers to species to age and size classes
and to community structure. However, there are large
differences in selectivity between animals. Large animals
are generally less selective in their diet than are smaller
animals. In general, animals select the smaller or younger
plants within a species, they select only some species
within a plant community and select only some plant
community types within a landscape (Dumont, 1997;
Legg et al ., 1998). Grazing animals exert controls on the
rates of several important processes in ecosystems. The
significance of this fact can be easily understood from
the conclusion by Wolfe and Berg (1988) that selective
browsing of young deciduous trees by ungulates has effec-
tively eliminated many deciduous hardwood species from
European forests, whereas conifers usually escape dam-
age. Many similar studies can be found in the literature for
other herbivores. The spatial distribution of wild rabbit
( Oryctolagus cuniculus ) in dry conditions, for example,
was found to be correlated with water and forage avail-
ability, but also with the existence of a protective cover
of shrubs or oak trees. At the same time consumption of

1. Plant communities dominated by re-sprouters with
deep root systems (e.g. species of Quercus, Pistacia,
Arbutus, Rhamnus, Phyllirea, Erica, Daphne and some
Ulex ) that recover very rapidly after fire from organs
which have been preserved below ground;
2. Plant communities dominated by species regenerating
after fire from seeds (for example, species of Cistus, Ros-
marinus officinalis and Ulex parviflorus ). The response