CPBR logo
Centre for Plant Biodiversity Research

Home > CANBR > Summer Scholarships

Ecological Risk Assessment of Transgenic Virus-Resistant White Clover

Characterising Isolates of Clover Yellow Vein Virus

By Rosemary Golding

Supervisor, Bob Godfree

February 2003


White clover, or Trifolium repens, is an exotic species in Australia. Its introduction was probably incidental, but populations were well established by the mid 1800s. White clover is both an important component of agricultural pastures and an invasive weed, especially in subalpine ecosystems. It is found in both dry and wet Poa grasslands, as well as woodland environments. A recent study at Long Plain in the Kosciuszko National Park found that white clover had an average of 1-2% ground cover, but at some sites covered almost 25% of the surveyed area (Godfree et al., in review).

Several viruses are associated with T. repens populations around the world. One of these is Clover Yellow Vein Virus (CYVV), which has been detected in pasture clover in North America, Asia and Europe. More recently, CYVV has been found in Australian and New Zealand, with infection rates as high as 77% of plants in one pasture (Forster and Musgrave, 1985). Few studies have examined infection rates in weed populations of white clover, but CYVV has been recorded in wild populations.

Glasshouse studies indicate that CYVV infection reduces some aspects of growth (such as biomass) in white clover plants (Campbell and Mayer, 1983). Little is known about the variability that exists within viral populations, or the effects of CYVV infection on plants in the field.

A diet of grass supplemented by white clover is known to increase milk or wool yields in grazing animals, and for this reason clover is important to many industries including dairy. A genetically modified virus-resistant white clover has been developed for commercial release, to reduce the impact of viral infection on the growth of agricultural clover. Trifolium repens is an obligate out-crossing species, and the virus-resistance genes could potentially cross into wild populations of white clover. Virus resistance may confer some advantage to clover as a weed, with unpredictable consequences. This possibility necessitates ecological risk assessment prior to the release of the transgenic clover.


Numerous CYVV-infected clover plants were collected from both grassland and woodland sites in the Namadgi National Park (south of Canberra) by Robert Godfree. From these, 24 isolates of CYVV were extracted using single lesion isolation in Chenopodium amaranticolor. The aim of my project was to characterise twelve of these isolates in order to estimate the variability that exists in CYVV extracted from wild clover. The experiment consisted of both a white clover growth trial and an indicator host trial.

Growth Trial

Seven seedlings of the white clover cultivar "Sustain" were inoculated with material from infected Chenopodium amaranticolor. Plants were grown for six weeks in an airconditioned glasshouse with regular watering, and several growth parameters were measured weekly. Growth was assessed using leaf number, stolon number, maximum leaf width, maximum stolon length, average petiole length, average internode length and inflorescence number. After seven weeks, all above ground material was harvested, dried and weighed as an estimate of biomass. A control group was inoculated using uninfected material, and a blank group underwent no inoculation procedures.

Indicator Host Trial

Growth Trial

Indicator Host Trial

Five known susceptible species (Chenopodium amaranticolor, broad bean Vicia faba, French bean Phaseolus vulgaris, tobacco Nicotiana tabacum and garden pea Pisum sativum) and two known insusceptible species (cowpea Vigna unguiculata and cucumber Cucumis sativus) were inoculated, with several plants of all species for each of the twelve isolates. Plants were mechanically inoculated, using ‘Celite’ to gently abrade the surface of the leaf and applying a pulverised mixture of infected leaf material from C. amaranticolor with approximately 3mL of a pH 8.5, 0.057M K2HPO4 buffer solution. Symptoms were recorded for each isolate and compared to expected symptoms as described in Barnett et al. (1987).

Results and Discussion

Growth Trial

The mean biomass of the control group plants was significantly higher than six of the eleven other isolates in pair-wise comparisons (p<0.05). Other growth parameters were interpreted as genotypic diversity. However, the bioassays into C. amaranticolor detected no CYVV infection in any of the plants used in the growth trial. This suggests that none of the plants were infected during inoculation. This low inoculation efficiency may have been due to low virulence of the inoculum or insufficient concentrations of virus within the plants to be detected in a bioassay.

Indicator Host Trial

All twelve isolates displayed symptoms in C. amaranticolor. Some isolates produced systemic symptoms in new growth, while other isolates were restricted to inoculated leaves. Symptoms included necrotic local lesions, systemic interveinal chlorosis and systemic necrosis. In V. faba, only three isolates displayed detectable symptoms, such as stem necrosis. One isolate displayed unusual symptoms in N. tabacum, with a systemic mosaic appearing after four weeks. Initial attempts to inoculate P. vulgaris with inoculum from C. amaranticolor produced no signs of infection, but repeated attempts using infected material from white clover host plants yielded chlorotic lesions from all three of the isolates tested. No symptoms were observed in P. sativum, and both the known insusceptible indicator species were symptomless.

Changing Nature of Viral Isolates

Retrospective observation of lesion counts in C. amaranticolor revealed a rough trend for isolates to decline in severity of symptom (or number of lesions per plant) with successive rounds of inoculation. There were also substantial differences between the numbers of lesions for each isolate. The apparent decline in virulence could possibly be attributed to host passage effects (Yarwood, 1979). The passage of some viruses through a series of hosts, whether of the same or different species, can cause a variety of changes including increasing or decreasing virulence, changes in host range and different symptoms (Yarwood, 1979). It is possible that CYVV is subject to host passage effects through successive rounds of inoculation between white clover and C. amaranticolor.


Barnett, O. W., Randles, J. W. and Burrows, P. M. (1987) Relationships amongst Australian and North American Isolates of the Bean Yellow Mosaic Virus Subgroup. Phytopathology 77(6):791-799.

Campbell, C. L. and Mayer, J. M. (1983) Effects of clover yellow vein virus and Codinaea fertilis on growth of white clover. Plant Disease 67: 70-73.

Godfree, R. C., Lepschi, B. J. and Mallinson, D. J. Invasion of subalpine communities in southeastern Australia: ecological filtering of alien plants and community invasibilty (in review).

Forster, R. L. S. and Musgrave, D. R. (1985) Clover yellow vein virus in white clover (Trifolium repens) and sweet pea (Lathyrus odoratus) in the North Island of New Zealand. New Zealand Journal of Agricultural Research. 28: 575-578.

Yarwood, C. E. (1979) Host passage effects with plant viruses. Advances in Virus Research 25: 169-190.    

Updated 11 March, 2003 by Murray (cpbr-info@anbg.gov.au)