Biology Reference
In-Depth Information
high potential for loss of nitrogen through leaching or trace gas losses (Mack,
unpublished), a result corroborated by lower recovery of an N15 tracer added to
plots within each site (Mack et al. 2001).
In addition to these ecosystem changes, the elimination of native species due to
fire changes the substrates upon which nitrogen fixation, and thus ecosystem N
accretion, occurs. Native plant species with symbiotic nitrogen fixation are rare in
these unburned woodlands. Instead nitrogen fixation occurs via asymbiotic fixation
in association with the litter of native species (Ley and D'Antonio 1998; Mack et al.
2001). The primary substrates for fixation are leaf litter of the dominant unburned
tree species
M. polymorpha
and the organic (O) layer of the soil in unburned sites.
Because
M. polymorpha
is eliminated by fire and the O layer is greatly altered and
no longer fixes N, the capacity of these systems to fix N decreases by two orders of
magnitude. This could significantly alter ecosystem development since these sites
are on young volcanic soils low in nitrogen (Mack and D'Antonio 2003a).
The almost complete replacement of native woody species with invasive grasses
in the submontane seasonal zone should have a profound influence on wildlife
composition. However, no studies have been done to document such impacts.
Furthermore, native birds are rare in intact native forests below 1,200 m because of
the prevalence of introduced avian malaria. Nonetheless, we have observed two
species of native birds, the Amakihi (
Hemignathus virens
) and the Apapane
(
Himatione sanguinea
) in our unburned sites. By contrast we have not observed any
native bird species in the burned sites.
7.2.3
Difficulty of Control and Management
By the early 1970s, feral goats had extensively browsed the woody vegetation in
the coastal lowlands and submontane woodlands of this region. After their removal
in the 1970s, grasses invaded virtually all of the coastal lowland and submontane
shrublands and forest with at least some soil above the lava bedrock (T. Tunison,
personal observation). By the 1990s, roughly 80% of the submontane seasonal zone
and lowland shrub/grasslands had experienced at least one grass-fueled fire (Tunison
et al. 2001). The primary management strategy for the ecosystem was fire suppres-
sion achieved through restricting human access to sites during the dry season and
immediate fire suppression when ignitions occurred. Chemical and mechanical
control of grasses was only being done for
P. setaceum
, which is restricted in its
distribution within the park.
M. minutiflora, S. condensatum
, and
A. virginicus
were
considered too widespread for any sort of chemical or mechanical control. Use of
grazing animals, none of which would be native to this environment, was not con-
sidered a viable management strategy.
Restoration of the prefire native community was not considered to be a viable
management strategy because most of the prefire dominants are not resilient to fire.
Surveys of sites that had burned once but not burned again for 20 years demon-
strated that only one of the prefire native shrubs,
Dodonaea viscosa
, regenerates
