The overall objective of this study was to test the hypothesis that insects can vector F.
tumidum conidia to infect gorse plants with the aim of developing an alternative approach to mycoherbicide delivery to control weeds. Four potential insect species (Apion ulicis, Cydia ulicetana, Epiphyas postvittana and Sericothrips staphylinus) were assessed for their ability to vector F. tumidum conidia. To achieve this, the external microflora (bacteria and fungi) and the size and location of fungal spores on the cuticle of these insect species were determined. In addition, the ability of the insects to pick up and deposit F. tumidum conidia on agar was studied. Based on the results from these
experiments, E. postvittana was selected for more detailed experiments to determine transmission of F. tumidum to infect potted gorse plants. The factors promoting pathogenicity of F. tumidum against gorse and the pathogen loading required to infect and kill the weed were also determined.
The external microflora of the four insect species were recovered by washing and plating
techniques and identified by morphology and polymerase chain reaction restriction
fragment length polymorphism (PCR-RFLP) and sequencing of internally transcribed
spacer (ITS) and 16S rDNA. A culture-independent technique (direct PCR) was also used
to assess fungal diversity by direct amplification of ITS sequences from the washings of the insects. All insect species carried Alternaria, Cladosporium, Nectria, Penicillium,
Phoma, Pseudozyma spp. and entomopathogens. Ninety four per cent of the 178 cloned
amplicons had ITS sequences similarity to Nectria mauritiicola. E. postvittana carried the
largest fungal spores (mean surface area of 125.9 ìm2) and the most fungal CFU/insect.
About 70% of the fungi isolated from the insects were also present on the host plant
(gorse) and the understorey grass. The mean size of fungal spores recovered from the
insect species correlated strongly with their body length (R² = 85%). Methylobacterium
aquaticum and Pseudomonas lutea were common on all four insect species.
Pseudomonas fluorescens was the most abundant bacterial species.
In the pathogenicity trials, the effectiveness of F. tumidum in reducing root and shoot
biomass of 16 and 8 wk old gorse plants was significantly increased with wounding of the
plants. Older plants (32 wk old) which were wounded and inoculated were significantly
shorter, more infected and developed more tip dieback (80%) than plants which were not
wounded (32%). This indicates that damage caused by phytophagous insect species
present on gorse through feeding and oviposition may enhance infection by F. tumidum.
Wounding may release nutrients (e.g. Mg and Zn) essential for conidia germination and
germ tube elongation and also provide easier access for germ tube penetration. Conidial
germination and germ tube length were increased by 50 and 877%, respectively when
incubated in 0.2% of gorse extract solution for 24 h compared with incubation in water.
Inoculum suspensions amended with 0.2% of gorse extract caused more infection and
significantly reduced biomass production of 24 wk old gorse plants than suspensions
without gorse extract. A minimum number of about 900 viable conidia/infection site of F.
tumidum were required to infect gorse leaves. However, incorporation of amendments
(which can injure the leaf cuticle) or provision of nutrients (i.e. gorse extract or glucose)
in the formulation might decrease the number of conidia required for lesion formation.
Scanning electron micrographs showed that germ tube penetration of gorse tissue was
limited to open stomata which partly explain the large number of conidia required for
infection. The flowers and leaves were more susceptible to F. tumidum infection than the
spines, stems and pods. An experiment to determine the number of infection sites
required to cause plant mortality showed that the entire plant needs to be inoculated in
order for the pathogen to kill 10 wk old plants as F. tumidum is a non systemic pathogen.
The number of infection sites correlated strongly with disease severity (R² = 99.3%). At
least 50% of the plant was required to be inoculated to cause a significant reduction in
shoot dry weight. F. tumidum, applied as soil inoculant using inoculated wheat grains in three separate experiments, significantly suppressed gorse seedling emergence and biomass production.
In experiments to determine the loading capacity of the insect species, E. postvittana, the
largest insect species studied, carried significantly more (68) and deposited significantly more (29) F. tumidum conidia than the other species. Each E. postvittana, loaded with 5,000 conidia of F. tumidum, transmitted approximately 310 conidia onto gorse plants but
this did not cause any infection or affect plant growth as determined by shoot fresh
weight and shoot height. E. postvittana on its own did not cause any significant damage
to gorse and did not enhance F. tumidum infection. It also failed to spread the pathogen
from infected plants to the healthy ones. There was no evidence of synergism between the
two agents and damage caused by the combination of both E. postvittana and F. tumidum
was equivalent to that caused by F. tumidum alone.
This study has shown that E. postvittana has the greatest capacity to vector F. tumidum
since it naturally carried the largest and the most fungal spores (429 CFU/insect).
Moreover, it naturally carried Fusarium spp. such as F. lateritium, F. tricinctum and
Gibberella pulicaris (anamorph Fusarium sambucinum) and was capable of carrying and
depositing most F. tumidum conidia on agar. Coupled with the availability of pheromone
for attracting the male insects, E. postvittana may be a suitable insect vector for
delivering F. tumidum conidia on gorse using this novel biocontrol strategy. Although it
is a polyphagous insect, and may visit non-target plants, F. tumidum is a very specific
pathogen of gorse, broom and a few closely related plant species. Hence, using this insect
species to vector F. tumidum in a biological control programme, should not pose a
significant threat to plants of economic importance. However, successful control of gorse
using this "lure-load-infect" concept would depend, to a large extent on the virulence of
the pathogen as insects, due to the large size of F. tumidum macroconidia, can carry only
a small number of it.
| Identifer | oai:union.ndltd.org:ADTP/183549 |
| Date | January 2007 |
| Creators | Yamoah, Emmanuel |
| Publisher | Lincoln University. Bio-Protection and Ecology Division |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
| Type | PhD doctorate |
| Rights | http://theses.lincoln.ac.nz/rights.html, Copyright Emmanuel Yamoah |
Page generated in 0.0026 seconds