|
II. Scientific Evidence Regarding Rainforest Ecology and Protection
8. DISEASE
Phytophthora cinnamomi is a soil and soil water borne
fungus that is widespread in Australasia (Weste and Marks 1987)
and is lethal to plants in many different taxa, including many
rainforest species (Marks and Smith 1991). For example, Podger
and Brown (1989) isolated P. cinnamomi from 39 species
of indigenous cool temperate rainforest flora in Tasmania. They
rated 30% of the rainforest flora as highly susceptible to the
disease and 5% as highly resistant. It is well established that
P. cinnamomi responds directly to elevated soil moisture
and soil temperature (Zentmyer 1980). CFL (1987) considered
P. cinnamomi not to be a threat to Victorian rainforests
because 'most warm temperate rainforests ... occur on well drained
sites which are not generally susceptible to infection by this
pathogen. Nor is it a threat to cool temperate rainforest because
of climatic conditions and soil temperatures' (p. 10). Nevertheless,
the Code of Forest Practice (CFL 1989) stipulates that care should
be taken to limit the spread of the disease. Despite the statements
by CFL (1987), the disease is prevalent in Tasmanian cool temperate
rainforests where soil temperatures and climatic conditions are
less favourable to the disease than they are in Victoria.
The distribution of the disease in Tasmania is symptomatic of
recent invasion although it may have been present for a very long
period (Burbidge 1960). In a State-wide survey, Podger et
al. (1990) noted that the distribution of P. cinnamomi
in remote areas of Tasmania was associated with mining exploration,
human settlement or walking tracks. Samples taken more than 2m
uphill of sharp boundaries between healthy and diseased vegetation
tested negative, suggesting a point infection and subsequent spread
in soil water. Podger and Brown (1989) found the disease to be
patchily distributed along exposed road and track edges, but they
did not recover it from healthy roadside vegetation or from undisturbed
rainforest. They suggested that infection is dependent on disturbance
that elevates soil temperatures above 15oC. Podger et al.'s(1990) analysis of its distribution in relation to climatic variables
suggested that ecological limits of the disease appear to be where
annual rainfall is less than 600mm and mean annual temperature
is less than 7.5oC. They concluded that it seemed unlikely that
practical measures for control could be devised for areas where
the disease is widely scattered.
Until recently, the known distribution of the disease in Victoria
did not suggest it would constitute a threat to cool temperate
rainforest. It is widespread in lowland forest in and around
catchments containing warm temperate rainforest. For example,
in East Gippsland, it is largely restricted to the south, distant
from most rainforest stands (Marks and Smith 1991). The evidence
from Tasmania suggests that there is at least the possibility
that P. cinnamomi is a relatively recent arrival in
the forests of the Central Victorian Highlands and in Victoria
generally, and that it is widespread in local populations (Marks
and Smith 1991). N. cunninghamii forests in Victoria have
a mean annual temperature of 10.6oC (s=1.6) and mean annual precipitation
of 1469mm (s=137; Busby and Brown, in press). Conditions in these
forests are warm enough and wet enough to support the disease,
contrary to the assumptions made by CFL (1987). However, it may
be that site susceptibility depends on soil fertility because
it is relatively virulent in Tasmania on oligotrophic soils (Podger
and Brown 1989, Podger et al. 1990). Weste and Vithanage
(1978) suggested that high organic matter content in soils may
stimulate competitive and antagonistic micro-organisms. Weste
and Ashton (1994) documented the recovery of most (but not all)
of the floristic elements in an open, dry sclerophyll forest in
Victoria 23-30 years after infection by the disease. The disease
appears to exhibit a characteristic decline following initial
spread, after which recovery commences. At their study site,
recovery is indicated but is far from complete.
The disease has been isolated from a recently dead individual
of Astelia australiana (CNR, Flora Section, personal communication).
This observation may have implications for the conservation of
one of Victoria's most endangered rainforest dependant plants.
Prediction of impact of the disease within a stand is likely to
be difficult because it depends on local variation in topography
and structural and floristic variation in local vegetation (Podger
and Brown 1989). The successful prediction of the occurrence
(or not) of an epidemic that has serious repercussions for rainforest
vegetation will depend on the accuracy of prognoses concerning
site and species susceptibilities, the rate of movement of the
disease among stands leading to new point infections, and the
probability and time frame for recovery after infection (see,
for example, Mollison 1977, Bailey 1980, Mollison and Kuulasmaa
1985). Monitoring for the disease in rainforest stands in Victoria
should be initiated on a regular basis. The monitoring program
should be designed to detect change in the status of the disease,
so that further action may be implemented, if warranted.
Myrtle wilt is the name given to the disease of Nothofagus
cunninghamii caused by invasion by the fungus Chalara australis.
It was first thought that the vector of the disease was the beetle,
Platypus subgranosus (Elliott et al. 1987, Roberts
1988, see Kile and Hall 1988). It was later discovered that the
pathogen enters the trees through exposed wounds, rendering the
tree susceptible to subsequent beetle attack (Kile et al.
1989). The disease causes the death of the plant, usually within
three years. The spores of the pathogen are air and water borne,
and infection may also occur by root contact below ground in dense
stands of N. cunninghamii, suggested by a relationship
between stand density and disease frequency and by the spatial
clumping of the disease (Elliott et al. 1987, Kile and
Walker 1987, Kile et al. 1989). Most new sites of infection
are believed to occur through stem and branch wounds (Packham
et al. 1992). The disease is much more prevalent in older
trees (Roberts 1988) because the chance of wounding increases
with age (Cameron and Turner 1994). Trees less than about 12
cm dbhob are rarely attacked unless they are fire scorched or
damaged by logging (Elliott et al 1987). Rainforest stand
isolation may mitigate the spread of the disease (Packham and
Kile 1992). The presence of other species in association with
N. cunninghamii such as blackwood in the forests of
the Otways and sassafras in the Central Highlands could inhibit
the local spread of the disease (Cameron and Turner 1994).
Many infectious diseases display fluctuations within host populations.
Susceptibility of individual trees to myrtle wilt will depend
on the number of fresh wounds, the amount of inoculum present
in the environment, disease resistance (if any), and the infectivity
of the wounds, usually related to their age (Packham and Kile
1992). The frequency of the disease in a population of N. cunninghamii
will depend on the amount of inoculum present, the average
wounding rate in the population, and the density of the stand.
The frequency of the disease among stands will depend on the
amount of inoculum present, the distances between stands, the
wounding rates and stem densities in stands, the rate of movement
of the disease among stands leading to new point infections, and
the probability and time frame for recovery after infection (see,
for example, Bailey 1980, Mollison and Kuulasmaa 1985). Because
this disease is a wound pathogen, and because forest harvesting
and roading activities result in many otherwise insubstantial
wounds to trees, it is reasonable to suggest that there should
exist a causal relationship between (particularly) mechanical
disturbance activities and disease frequency, either within or
among rainforest stands. In Victoria, road construction, harvesting
in mixed forest stands, and windthrow in buffers have the potential
to elevate the background wounding rate.
The response of the population to elevated disturbance will depend
on the relative importance of the processes that cause the disease.
Wounding and infection are natural processes, particularly in
mixed forests and ecotonal communities with eucalypt overstoreys
where branch fall must cause wounds. Cameron and Turner (1994)
suggest that background inoculation rates are sensitive to human
disturbance because fungal spores are wind dispersed. Thus, if
the wounding rate throughout the landscape increases, the amount
of inoculum in the environment will increase throughout the landscape
from a series of local point infections, resulting in an accelerated
infection rate. This scenario is not unreasonable, and deserves
the attention of a specific research project. The relationship
between the host, pathogen and beetle is likely to be the product
of a long process of co-evolution (Cameron and Turner 1994).
It is very likely that there have been cyclical or stochastic
fluctuations in the disease in the past. The most likely scenario
is that we are observing a natural phenomenon that has been exacerbated
in some areas by past human activity. Such activities may include
road construction, harvesting in mixed forest communities, windthrow
and land clearance for agriculture. There is also the potential
for time lagged responses. Past activities in rainforest may
have consequences that are only now observable in the form of
elevated mortality.
Disturbances may provide a disease focus through wounding and
subsequent infection of edge trees in ecotonal vegetation (Kile
et al. 1989). There is empirical evidence to suggest that
a relationship exists between disturbance and disease incidence
in the N. cunninghamii forests of Tasmania. Kile
et al. (1989) found that the proportion of diseased trees
in transects declined with distance from roads in extensive stands.
The relationship was strong (r2=0.53) and differences with distance
were appreciable (70% infection at 10m to 25% infection at 200m).
Neyland and Brown (1994) found the incidence of the disease in
isolated stands of rainforest was strongly positively associated
with the degree of disturbance. However, there is considerable
variability in the response of the disease to disturbance levels
(Packham 1990, Kile 1992, Cameron and Turner 1994).
The disease has implications for the ecology of Victorian N. cunninghamii
dominated forests in the medium term. It opens gaps within
rainforest canopies in areas that are usually more or less immune
from edge effects (Roberts 1988). Either rainforest species or
sclerophyll species, not necessarily eucalypts, may recruit into
these gaps, depending on local conditions and seed sources. Cameron
and Turner (1994) speculated that the current recruitment rate
of new plants in wilt affected areas may not be replacing the
adults eliminated from the population, particularly in the Otways
region, although the data are far from complete, and that the
creation of gaps may increase the risk of canopy fires in these
communities. Inference of medium term impacts is made difficult
by the fact that, despite the recent report by Cameron and Turner
(1994), there are very few data on the distribution or prevalence
of the disease in Victoria except in broad, qualitative terms.
There is little doubt that logging and roading operations and
such factors as windthrow have the potential to increase infection
and death rates among N. cunninghamii trees in Victorian
rainforest stands. It is likely that the Otways region of Victoria
is suffering an unusually high rate of disease, probably higher
than could be sustained in the medium term. Recommendations of
protection measures against the disease in Tasmania have taken
the form of buffers. Kile et al. (1989) and Packham (1990)
suggested that buffers of at least 50m and up to 350m are required
to protect against disturbance providing a focal point for infection
of extensive rainforest stands. In Victoria, because primary
rainforest is distributed in small, isolated stands which are
excluded from harvesting, the dynamics of the disease may be qualitatively
different. More extensive stands of mixed forest are also susceptible.
Management of the disease should focus on prescriptions for road
construction through Nothofagus forest, logging in mixed
forest, and minimisation of the impacts of windthrow. Suggestions
in the literature include treatment of wounds on N. cunninghamii
that result from silvicultural operations or windthrow, timing
of harvesting to coincide with periods when the inoculum is scarce,
and removal and burning of any dead trees, as far as is possible
without causing further damage. In particular, roading activities
are at least partly a function of silvicultural systems and coupe
planning (Burgess 1993), and thus the impact of roads on the dynamics
of the disease are partly under the control of forest planners.
|
