|
A Technical Guide for Wine Producers
|
|
RECENT ARTICLES | WYNBOER HOME
The Occurrence And Control Of Botrytis Cinerea (Grey Mould) In Grapes: A South African Perspective
The findings underlined the complex behaviour of the fungus, and its disease biology. The research approach followed created a basis for making meaningful recommendations to table and wine grape producers regarding the control of grey mould. The findings of the team also led to another important aspect; the complex disease biology underlined the fact that plant pathologists, viticulturists and nurserymen would have to act as a team to further refine control measures. For the first time guidelines were clearly stipulated for the strategies to be followed by nurserymen in order to increase resistance in grape cultivars against the dreaded fungal disease. This article covers highlights of the research, which is considered to be of great significance and enjoys international recognition.
THE DISEASE CYCLE
Botrytis cinerea is widespread, usually grows on dead plant material (on floral parts in particular) and attacks a variety of plants. The fungus is not considered a strong pathogen, and causes disease in young tissue, or damaged plant material. The fungus has a distinctive life cycle. It forms tough survival structures, known as sclerotia, which may develop on any infected plant material where they can survive the winter. The structures germinate in late winter or early spring, and form conidia (little spores with a diameter of approximately 0,01 mm). The conidia are released, land on plant parts, germinate, penetrate suitable tissue and then initiate the disease phase. If a lesion is formed, the hyphae grow outward and form a new generation conidia. The conidia then perpetuate the life cycle.
The infection cycle
Overseas literature indicates that the fungus may penetrate the berries in four ways. Firstly, conidia germinate on the stigma during flowering, penetrate the style and grow into the young berry, where they are then arrested by host resistance. The effectiveness of the resistance weakens during berry development, the fungus gradually resumes its growth and later in the season rotting sets in. Secondly, conidia germinate on the berry surface, germ tubes penetrate the fine cracks around the stomata, or penetrate the berry by growing into the stomata. Shortly afterwards rotting sets in. Thirdly, conidia land next to or in wounds. When they germinate close to the wound, they grow towards the wound and penetrate the berry. If they land in the wound, they develop directly inside the berry. In both cases rotting sets in shortly afterwards. These three infection pathways are generally considered to be the most prominent. The fourth pathway is seldom referred to. In this instance the fungus penetrates stamens during flowering, grows down into the filaments and penetrates the receptacle, where it is localised. With ripening the fungus then grows from the pedicel into the pedicel-end of the berry, and rotting occurs.
In the case of vines, grey mould is traditionally associated with the berries and rainy conditions. It is generally accepted that the fungus occurs in large numbers on the berries, that the berries are the primary part of the plant where infection occurs, and that infection requires free water. Researchers therefore direct their studies of the fungus on berries, and use ripe berries for studies of infection and other investigations. In order to obtain clear disease symptoms on berries, researchers usually apply masses of conidia to establish infection, and berries are usually kept wet. Data obtained from these studies form the basis for further conclusions about the fungus and the disease.
OCCURRENCES IN THE VINEYARD
In the vineyard the picture looks somewhat different. The fungus moves in vineyards mainly as conidia, carried by air currents, raindrops and insects. Investigations in local vineyards with spore traps have indicated that conidia move individually in air currents, not in groups. Air currents deposit conidia as single cells on plant surfaces. The conidia are hydrophobic, and are not easily wetted. This characteristic of conidia causes them to occur on the surface of raindrops, and therefore rub off individually on the plant surface when the drops roll over the surface. Each part of the vine, not only the berries, is therefore a potential target for the fungus.
The behaviour of airborne conidia on the berry surface
In nature, bunches of grapes are seldom completely and thoroughly wetted. Some parts will be exposed to rain or dewdrops, while other parts will only be humid. Conidia have been found to germinate within 1-3 hours of settling on humid (93% relative humidity) or wet berries. Conidia do not require extra nutrition for germination, germ tube growth and penetration. In view of this knowledge, nature was imitated as closely as possible in all the laboratory investigations. For infection studies airborne conidia were distributed over bunch parts in an inoculation tower, and bunches were either kept humid or wet. Fluoressence microscopy was used to investigate the growth pattern of solitary conidia, their vitality, penetration ability, penetration sites, die-back and host resistance (harmful compounds in berry secretions, skin components, suberin and stilbene formation). By killing off the fungal germlings on the berry surface after set periods, or leaving them intact, and applying a differential set of isolation, freezing and paraquat treatments, the extent of surface colonisation, infection and symptom development caused by the fungal germlings could be determined.
Microscopic investigations showed that airborne conidia never landed on stomata, nor grew into them. Germ tubes did not penetrate the micro cracks around stomata either. Penetration always occured directly through the skin. These findings are in conflict with overseas literature, where researchers use masses of conidia for infection studies. It is significant that conidia on humid berries constantly formed an extremely short germ tube (equal to or shorter than conidia). On wet berries the germ tubes grew more extensively.
Solitary conidia did not cause disease symptoms on humid or wet berries from pea size to harvest. However, a small minority of these conidia were able to penetrate the skin. Natural resistance of the berry skin was very efficient at the pea size and bunch closure stages, and the few germlings that penetrated, were arrested and killed off by natural host resistance. The germlings formed symptomless infections in the berry skin from véraison to harvest, but at very low frequencies. Natural host resistance was still active during these phases, and no symptom expression occured.
Histological studies of germling dieback and host resistance clearly showed that harmful compounds in berry secretions, skin components, and suberin and stilbene formation each contributed to the resistance of the berry skin. It was clear that both harmful compounds in berry secretions and skin components were prominent resistance factors.
Since it was apparent that airborne conidia did not readily infect unblemished berries, the interactions between airborne conidia, germlings and wounding were investigated. Green and ripe berries were used for this purpose. Symptom development occurred only where fresh conidia were applied to wet berries with fresh wounds.
Airborne conidia only cause symptoms when they land on wet berries with fresh wounds. Established germlings on the berry cheek do not cause wound infection.
The behaviour of airborne conidia on the berry surface of cold stored grapes
It has been found that cold storage (-0.5řC for 2 weeks) moderately lowered the natural resistance of the berry skin. In contrast with fresh grapes, conidia caused lesions on cold stored berries. Furthermore, both conidia and established germlings could infect wounds on cold stored grapes.
Occurrence of natural infections in bunch parts
The above findings put a question mark behind the relative importance of the various pathways available to the fungus to penetrate berries, to establish and eventually cause symptom expression. The occurrence of natural infections, and the initial site of symptom development within the grape bunch, were therefore investigated in grapes from various vineyards (table and wine grape cultivars and different production areas) over a seven season period. In the studies symptom development was researched in superficial conidia and established germlings (symptomless infections). As a rule symptom expression occurred only after manipulations in the laboratory. It was found that the fungus, in the case of both infection types, followed a constant pattern. Symptoms consistently developed first on the pedicel, lateral and to a lesser extent the rachis. In the pedicel the fungus developed mostly out of the receptacle. The first signs of symptom development seldom occurred in the berry cheek, and hardly ever in the style end. When symptoms did appear on the berry, it started from the pedicel end. Histological studies indicated that the fungus penetrated the berry systemically from the receptacle through the vascular tissue, followed by the onset of rotting.
These findings unequivocally showed that in nature, B. cinerea infected bunch parts using an infection route seldom referred to in overseas literature.
Growth stage when natural infections are established on vines
The frequency at which the fungus occurred in the various bunch parts differed from year to year and from vineyard to vineyard. This means that certain years and certain vineyards were more conducive to fungal development. Apart from these differences, there was a corresponding pattern: in the pea size and bunch closure stages the infection frequency was high in the rachis, lateral and the pedicel in particular, followed by a drastic reduction. The fungus seldom occurred in these parts at harvest. As a rule cheek infections did not occur from pea size stage to véraison, but was found at low frequencies at harvest.
Resistance of bunch parts against the establishment of natural infections and symptom expression
The finding that natural symptomless infections in the rachis, lateral and pedicel decreased drastically as the season progressed, suggested that the level of natural resistance in these parts increased gradually, and that infections in these parts were eventually eliminated. Inoculation studies with airborne conidia confirmed that the structural bunch parts were susceptible in the young stage and that their resistance increased moderately as the season progressed. The only exception here was the pedicel, which was susceptible throughout the season, and in the case of certain cultivars even increased in susceptibility.
Uniform behaviour in bunch parts of different grape cultivars
The fungus followed a consistent behavioural pattern in bunches of various table and wine grape cultivars. However, grape cultivars differed with the resistance of the various bunch parts to the establishment of infections and with regard to symptom expression. This phenomenon was more noticeable in the case of the pedicel.
In the course of the investigation researchers in Australia, Switzerland, France, California and New York were requested to conduct similar studies, to determine whether the fungus followed a constant pattern of infection world-wide. Their findings confirmed that this was indeed the case. Berry rotting developing from the style end, as mentioned in the literature, could not be proved.
Resistance of leaves to the establishment of natural infections and symptom expression
In the case of leaves the fungus was found to occur naturally at high levels in symptomless leaves. As a rule leaves were resistant to symptom expression in the adult stage, but inoculation studies indicated that young leaves readily developed symptoms. The disease reaction of the petiole corresponded noticeably with that of the pedicel. Symptom development on leaves therefore followed a corresponding pattern. Beginning at the petiole, it spread into the leaf blade.
Inoculum levels in the vineyard
Investigations into the occurrence of B. cinerea in nectarine, plum and pear orchards, and vineyards, indicated that the fungus occurred at high levels from early September to late November, and that the levels then dropped. This phenomenon can be ascribed to the fact that the fungus grows on dead material of a wide variety of plants, and is generally associated with flower infection of weeds. A variety of plants flower during this period and serve as a potential inoculum source. Climatic conditions in late spring, and occasionally in early summer, promote the formation and release of conidia.
DISEASE EPIDEMICS IN THE VINEYARD
In view of the findings regarding the infection route, the natural occurrence of the fungus in the various bunch parts, and the resistance thereof to symptom expression, there is uncertainty about the specific set of factors giving rise to symptom expression in vineyards. The fungus clearly requires a "helping hand" to enable the infection cycle to run its full course, and generate a symptom. This helping hand is to be found in viticultural practices, cultivar characteristics and environmental factors. There are strong indications that factors which may cause "weak spots" (compact bunch structure, wind damage, wounding, berry burst) at the pedicel end of the berry, play a determining role.
The role of insects is currently being investigated, and there is good reason to believe that insects play a prominent role in symptom expression in the vineyard. Insects seem to carry with them masses of inoculum, depositing them individually or in groups on plant parts. Insect secretions sometimes also contain groups of conidia. Moreover insects fulfil another important role: they wound the tissue and deposit conidia or mycelia in the wound cavity, which is an important prerequisite for symptom development. In the case of the fruit fly, it was found that the insect fed mainly around wounds on the berry and placed conidia and mycelial fragments in wounds. Symptoms usually developed from these wounds. However, there was also another important observation. On undamaged berries, fruit flies drastically increased the occurrence of pedicel end infections on ripe berries, and symptoms developed in the absence of humid conditions or free water. Microscopic studies showed the flies to deposit the two types of inoculum (conidia and hyphae) on adult berries mostly at the pedicel end. This is ascribed to seeping of berry juice at the pedicel/berry attachment, thus attracting the fruit flies.
The phenomenon that young bunches can carry high levels of the fungus symptomlessly, and older bunches low levels, has important implications for disease control. In practice disease symptoms are prominent in vineyards only after the bunch closure stage. Symptom development in wine grapes is noticeable at an earlier stage than in table grapes. This indicates that in practice the "helping hand" impacts on symptom development only after bunch closure.
The phenomenon that young bunches can carry high levels of the fungus symptomlessly, and older bunches low levels, has a further important implication for disease control, and the value and reliability of disease prediction models. In the Western Cape wet climatic conditions from September to November, and drier conditions from January to February, play a determining role in the availability of inoculum, the number of infection cycles that occur, and the development of a disease epidemic. Symptomatic plant tissue forms conidia under humid and wet conditions. Inoculum levels will therefore increase in the vineyard only after a protracted wet period, and infection will occur only afterwards during a subsequent wet period. Periods conducive to the creation of "new" inoculum, and an infection cycle, occur more generally early in the season, and only sporadically later in the season. The level at which infections occur at a given developmental stage in the bunch parts, is therefore determined by the inoculum level in the vineyard, and the occurrence of subsequent infection cycles. Monitoring trials with symptomless naturally and artificially infected bunches indicated that, following laboratory manipulations, a time lapse occurred before symptoms developed and new conidia were noticeable on lesions. Symptoms usually appeared 4-7 days after the stress factor (freezing or paraquat treatment) had been applied, and the first conidia usually developed 2-4 days later. Under conditions that were extremely favourable for disease development, a germinating conidium took approximately 9-13 days to produce a new generation of conidia on berries. The development of a disease epidemic after a single rainshower late in the season now becomes questionable. According to the pattern described here, an outbreak of the disease following a short-lived wet period late in the season cannot be ascribed to new infections, but to the development of symptomless infections established earlier in the season.
PROTECTION OF BUNCH PARTS AGAINST INFECTION
In view of this knowledge regarding inoculum levels, the infection route, the establishment of infections, and symptom expression, the protective effect of various biocontrol agents and fungicides on the different bunch parts was investigated.
Biological control
In the case of biological control, the premise was that a combination of different organisms, each able to perform a specific task, had to be used. Ulocladium atrum, Gliocladium roseum, Trichoderma harzianum and Trichosporon pullulans were selected. Ulocladium atrum is a highly competitive saprophyte known for its ability to rapidly colonise dead tissue. This fungus was selected for its ability to quickly occupy dead tissue (flower parts, aborting grains, lesions) in the bunch and prevent B. cinerea from becoming established in, and forming conidia on it. Gliocladium roseum grows systemically in woody parts of plants and is a well-known antagonist of B. cinerea. This fungus was selected to colonise the rachis, lateral and pedicel, and to protect the bunch parts against infection by B. cinerea. Trichoderma harzianum is a well-known antagonist of B. cinerea. The fungus is able to grow on any of the bunch parts, and is also known for its ability to activate the inherent resistance mechanisms of the vine, thereby increasing the natural level of protection. Trichosporon pullulans operates only on the berry surface of ripe grapes, where it acts antagonistically towards B. cinerea. Two cultivars (Chardonnay and Dauphine) that differ regarding bunch structure and susceptibility to the disease, were used.
It was found that under natural conditions the organisms did indeed occupy the niches for which they had been selected, but that climatic conditions played a determining role in their establishment. They erratically colonised the bunch parts. The extent of biological control thus obtained was therefore extremely erratic. Trichoderma harzianum was the most successful antagonist.
Control by means of fungicides
The protective effect of various fungicides (iprodione, pyrimethanil, fenhexamid, cyprodinil/fludioxinil) on the respective bunch parts and on leaves was investigated. Two cultivars (Merlot and Dauphine) that differ regarding bunch structure and susceptibility to the disease, were used. Shoots with leaves and bunches were sprayed under controlled conditions in a spraying tent so as to ensure good penetration of the inner bunch parts, and subsequently exposed to airborne inoculum. This study provided significant data regarding disease control.
The fungicides were highly effective and protected the rachis, lateral and pedicel of both cultivars efficiently during the various growth stages. The only exception was the receptacle/pedicel end infection, which was not significantly controlled by the pre-harvest application. In the case of berry cheek infections, the occurrence thereof in both cultivars was too low during pea size, bunch closure and véraison stages to reach meaningful conclusions about the protective effect of the fungicides. Infection frequencies in the berry skin were marginally higher at harvest, and spraying at this stage reduced berry skin infections. The protective effect was more noticeable in Merlot (more susceptible berry skin) than Dauphine (more resistant berry skin). The fungicides protected leaves effectively against infection.
Resistance to fungicides in fungus populations
Various studies have been conducted on this topic. The findings provide a good basis for implementing an effective fungicide resistance control strategy.
Dicarboximide fungicides (iprodione, vinclozolin and procymidone). Resistance to the dicarboximide fungicides was generally encountered in B. cinerea populations in vineyards. The occurrence of resistant subpopulations was more common in the Paarl region, less in the Hex River valley, and low in the Orange River area. The majority of the isolates had a low level of resistance, and as a rule, in the absence of these fungicides they were less aggressive on unblemished berries than the sensitive population. However, the resistant isolates could infect fresh wounds just as readily as the sensitive isolates. The increase of the resistant subpopulation followed a set pattern annually. It increased drastically after flowering and reached a peak at the bunch closure stage. Thereafter it declined drastically.
Benzimidasole fungicides. Resistance to these fungicides was generally found. Large numbers of isolates were resistant to both the dicarboximide and the benzimidasole fungicides.
Folpet. In some of the vineyards where dicarboximide resistance occurred, the frequency of the resistant subpopulation sometimes increased early in the season in the absence of dicarboximide applications. This pattern did not comply with the general rules regulating resistance development. Once all possible factors had been considered, it was found that these vineyards received a few successive folpet sprays early in the season to control Phomopsis viticola (Phomopsis cane and leaf blight). Cross-resistance between dicarboximide fungicides and folpet is unknown. The sensitivity of isolates to folpet from vineyards with a known history of dicarboximide resistance was consequently investigated. It was found that generally dicarboximide resistant subpopulations were less sensitive to folpet, and that incomplete cross resistance to these two fungicides occurred in B. cinerea. Early, successive folpet sprays in a vineyard with a dicarboximide resistant subpopulation, exerted severe selection pressure on the dicarboximide resistant subpopulation which then increased rapidly in the absence of dicarboximide applications. This phenomenon is especially noticeable since the fungus population is largest early in the season.
Pyrimethanil. No cross-resistance between this fungicide and dicarboximide, or benzimidasole fungicides, was observed in fungus populations with a known history of fungicide resistance. Monitoring done in vineyards where the fungicide was sprayed for one, two or three successive seasons, showed however that a resistant subpopulation developed at low frequencies in the third season. Resistance to pyrimethanil in B. cinerea can therefore increase in vineyards with increased selection pressure, resulting from successive applications of the fungicide.
Fenhexamid. No cross-resistance between this fungicide and dicarboximide, or benzimidasole fungicides, was observed in populations with a known history of fungicide resistance. Since the fungicide is still new on the market, selection pressure has not been exerted on the fungus. According to overseas sources, resistance to fenhexamid may well build up in B. cinerea populations.
DISEASE CONTROL STRATEGY
From the above it is clear that between flowering and harvest, control strategies should focus on four "window periods". These stages are late flowering/set, bunch closure stage, véraison and shortly before harvest. Fungicide application during the first two stages is of cardinal importance. This is the period when early primary infection is established under a wide range of environmental conditions in the rachis, lateral and pedicel. These infections must be prevented or reduced. Spraying in the first window should comprise a dicarboxymide fungicide, and in the second window either pyremethanil, fenhexamide or cyprodinil/fludioxinil. The third period is véraison, when infection may still be established in the rachis, lateral and pedicel, as well as the berry skin. This will only happen provided inoculum is available, in other words successive wet periods that give rise to the buid-up of high inoculum levels, and successive infection cycles. In the third window either a dicarboxymide, or fenhexamid or pyremethanil may be used. Select the fungicide according to the anti-resistance strategy applicable to the vineyard in question. Spraying in the fourth window is not recommended. The reasons for this are that spraying under controlled conditions in the laboratory showed berry skin infections to be only marginally reduced, and that the fungicide does not penetrate the bunch sufficiently to limit infections developing from the receptacle.
In the case of table grapes, there is a fifth important window period, namely harvest and handling thereafter. It is essential that new infections be prevented during this window period. It has been found that fungicides are unable to fulfil this task, and that SO2-treatment is required. The correct use of SO2-sheets is therefore of absolute importance to clean bunch parts of superficial conidia. Since wound infection during this window period largely contributes to rotting, any practice that damages the berry skin should be prevented.
ACKNOWLEDGEMENTS
The following organisations are thanked for financial support of the research: Deciduous Fruit Producers Trust; National Research Foundation; Stellenbosch University; and THRIP. The following producers are thanked for making vineyard blocks available for field trial purposes, or for donating vine material for experimental purposes: J.D. Kirsten (Irene); P. Benadé (Lievland); A. Hoekstra (Nancy); D.A. Clift (De Hoop); M. Viljoen (Nuwehoop); D. Coetzer (Arbeids Adel); F. Rossouw (Meiringshoop); A. Bredell (Dellrust); P. Reynolds (R. Müller, Buitenverwachting); C.J. Bröcker (Groot Constantia); J.H. Smit (Jakkalsfontein); R. Jeffrey (Timberley); G.H. de Kock (Somerslus); A.F. Viljoen (Werda); F.J. Rossouw (Mon Repos); S.W.C. Rabie (Buffelskraal); A.F. Hill (Clovelly); M. Lotter (Cramix); J.J. Haasbroek (La Bri); A.J. van Velden (Overgaauw): J.P. Bredell (Heldersig); and C.B. Kerr (Montana). A special word of thanks to J.D. Kirsten for support of the research in general.
|
|
Wynboer is incorporated in WineLand, magazine of the SA wine producers.
Visit our sister sites:
|
Gustav Holz
| UP |
COPYRIGHT (C) 2000 WineLand |