Optimizing biocontrol of purple nutsedge (Cyperus rotundus).

Abstract

Cyperus rotundus L. CYPRO (purple nutsedge) and Cyperus esculentus L. CYPES (yellow

nutsedge) are problematic weeds on every continent. At present there is no comprehensive

means of controling these weeds.. The primary means of control is herbicides, although the

weeds are becoming more resistant. Bioherbicide control of purple and yellow nutsedge is an

important avenue of research, with much of the focus being to increase the virulence of current

fungal pathogens of C. rotundus and C. esculentus.

The primary aim of this study was to increase the virulence of a fungal pathogen of C. rotundus

and C. esculentus, with the objective of creating a viable bioherbicide.

A possible means of increasing the virulence of a pathogen would be to increase the amount of

amino acid produced by the fungus. This was proposed as a means of increasing the virulence of

Dactylaria higginsii (Luttrell) M. B. Ellis. Overproduction of amino acids such as valine and

leucine result in the feedback-inhibition of acetolactate synthase (ALS), an enzyme which is a

target for many herbicides currently on the market. By applying various amino acids to tubers of

purple nutsedge and comparing the results with a reputable herbicide, glyphosate, it was possible

to determine the success of the amino acid applications. Only glutamine treatment at 600 mg.r1

resulted in significantly less (P<O.OOI) germination compared with the water control, while the

glyphosate application resulted in no germination. Four treatments were significantly different

(P<O.OOI) from the water control in terms of shoot length, but no pattern or conclusion could be

drawn from the results. Injecting amino acids and glyphosate into the leaves of the plants gave

similar results to those obtained with the tubers, with no visible damage on those plants injected

with the amino acids and complete plant death of those injected with glyphosate. Amino acids

had little effect on the growth of the C. rotundus plant or tuber. It was later determined by a

colleague (Mchunu1

, unpublished) working on the same project, that D. higginsii does not infect

the local ecotypes of C. rotundus in Pietermaritzburg, South Africa.

A second fungus, Cercospora caricis Oud., was isolated from C. rotundus growing in the region,

and confirmed as a Cercospora species by conidial identification. Like many Cercospora

species, C. caricis produces a phytotoxin, cercosporin. An increase in production of cercosporin

would theoretically lead to an increase in virulence of C. caricis. Mutation of hyphae by

i

J Makhosi Mchunu: Address: National department ofAgriculture; Private Bag 3917; Port Elizabeth; 6056

Email: Makhosimc@NDA.agric.za

ultraviolet-C light was perfected on C. penzigii Sacc., where 5 min exposure to DV-C light

resulted in approximately 99% cell death. Surviving colonies were analysed by spectrophoresis,

and the surviving mutant gave an absorbance value of approximately 5% more than the median.

Samples were analysed by high-performance liquid chromatography (HPLC) to determine the

presence of cercosporin. No definitive result was obtained. Exposure of C. caricis to DV-C for

5 min. resulted in approximately 65% hyphal cell death, with 20 min. resulting in approximately

95% death. A spontaneous mutant was observed in a colony that had been exposed to DV-C.

This mutant showed sectored growth with red and grey growth patterns. The red section of the

mutant was subcultured and analysed by spectrophoresis and HPLC. The red C. caricis gave an

absorbance reading of approximately 140 on HPLC compared with about 22 from the grey

colony. HPLC analysis of the wild-type C. caricis did not produce a peak corresponding to that

of the cercosporin standard, although no conclusion could be obtained on the presence or

absence ofthe toxin.

The virulence of the mutant C. caricis could not be determined as inoculation experiments were

unsuccessful, and had to be discontinued due to time constraints.