A review of Grapevine Leafroll associated Virus type 3 (GLRaV-3) for the New Zealand wine industry

Charles JG, Cohen D, Walker JTS, Forgie SA, Bell VA, Breen KC. February 2006

Report to New Zealand Winegrowers HortResearch Client Report No 18447 HortResearch Contract No. 20498

Charles JG, Cohen D, Forgie SA. Mt Albert Research Centre 120 Mt Albert Road, Private Bag 92 169 Mt Albert, AUCKLAND, NEW ZEALAND Tel: +64-9-815 4200 Fax: +64-9-815 4201

Walker JTS, Bell VA, Breen KC. HortResearch Hawkes Bay Cnr Crosses and St Georges Roads Private Bag 1401, HAVELOCK NORTH, NEW ZEALAND Tel: +64-6-877 8196 Fax: +64-6-877 4761

DISCLAIMER
Unless agreed otherwise, HortResearch does not give any prediction, warranty or assurance in relation to the accuracy of or fitness for any particular use or application of, any information or scientific or other result contained in this report. Neither HortResearch nor any of its employees shall be liable for any cost (including legal costs), claim, liability, loss, damage, injury or the like, which may be suffered or incurred as a direct or indirect result of the reliance by any person on any information contained in this report.

This report has been prepared by The Horticulture and Food Research Institute of New Zealand Ltd (HortResearch), which has its Head Office at 120 Mt Albert Rd, Mt Albert, AUCKLAND. This report has been approved by:

__________________________________ Research Scientist Date: 24 August 2006

________________________________ Group Leader, Bioprotection Date: 24 August 2006

CONTENTS
Page EXECUTIVE SUMMARY........................................................................................................ 1 PREFACE AND TERMS OF REFERENCE .......................................................................... 15 Preface.................................................................................................................................. 15 METHODS AND STRUCTURE OF THE REVIEW ............................................................. 17 SECTION 1 THE GLRAV-3 VIRUS ...................................................................................... 18 Early reports of grapevine leafroll disease....................................................................... 18 Vine pathology ................................................................................................................. 19 Viruses associated with Grapevine leafroll disease ......................................................... 19 What types of grapevine leafroll associated viruses are present in New Zealand?.......... 21 Virus strains of GLRaV-3 ................................................................................................ 23 Other viruses infecting grapevines................................................................................... 23 Serological detection ........................................................................................................ 25 Molecular detection methods ........................................................................................... 25 Seasonal movement of GLRaV-3 within the vine ........................................................... 26 Virus movement and detection in newly infected grapevines ......................................... 27 Thermotherapy ................................................................................................................. 27 Grapevine resistance to GLRaV-3 ................................................................................... 28 The International Council for the Study of Virus and Virus-like Diseases of the Grapevine (ICVG)............................................................................................................ 29 SECTION 2 THE IMPACT OF GLRAV-3 ON VINE GROWTH, GRAPE YIELD AND JUICE QUALITY .................................................................................................................... 31 Introduction .......................................................................................................................... 31 Leafroll virus and vine performance in New Zealand ..................................................... 31 Effects of GLRaV-3 on vine growth ................................................................................ 32 Grape yield ....................................................................................................................... 37 Grape quality.................................................................................................................... 37 SECTION 3: GLRAV-3 VECTORS/TRANSMISSION ECOLOGY..................................... 47 Introduction .......................................................................................................................... 47 A short history of mealybugs on grapes in New Zealand ................................................ 47 Pseudococcidae the mealybugs ..................................................................................... 48 Pest status of mealybugs in New Zealand........................................................................ 48 Pest status of mealybugs on grapes in New Zealand ....................................................... 52 Developmental biology and life history characteristics of mealybugs ............................ 53 Endosymbionts of mealybugs .......................................................................................... 54 Mealybug vectors of plant viruses ................................................................................... 55 Mealybug vectored viruses of plants other than grapes ................................................... 56 Mealybug vectored grape viruses other than GRLaV-3 .................................................. 61 Transmission of GLRaV-3 by mealybugs........................................................................ 62 Transmission of GLRaV-3 by other insects..................................................................... 64 Transmission ecology of GLRaV-3 ................................................................................. 64

SECTION 4: MANAGEMENT OPTIONS WITHIN THE VINEYARD TO LIMIT GLRAV3 INFECTION AND SPREAD................................................................................................ 67 Introduction .......................................................................................................................... 67 Insects as vectors.............................................................................................................. 67 Managing the spread of GLRaV-3 in vineyards .............................................................. 68 Managing the spread of GLRaV-3 originating within vineyards..................................... 68 Managing spread of GLRaV-3 originating in other vineyards ........................................ 69 Managing infested planting material................................................................................ 70 Cultivar susceptibility to mealybug ................................................................................. 71 Mealybug control in vineyards......................................................................................... 72 Sampling for mealybugs and thresholds for action.......................................................... 73 Insecticidal control of mealybugs .................................................................................... 75 Degree-day models for timing mealybug insecticides ..................................................... 77 Biological control of mealybug........................................................................................ 78 Management of ant populations ....................................................................................... 78 Weed control within vineyards ........................................................................................ 79 SECTION 5: SUMMARY AND RECOMMENDATIONS.................................................... 81 Summary and integration of key findings ............................................................................ 81 REFERENCES......................................................................................................................... 87

EXECUTIVE SUMMARY
A review of Grapevine Leafroll associated Virus-3 (GLRaV-3) for the New Zealand wine industry
A report to New Zealand Winegrowers Charles JG, Cohen D, Walker JTS, Forgie SA, Bell VA, Breen KC. February 2006

A key goal of the New Zealand Wine Industry is to control the spread and impact of virus diseases, particularly Grapevine leafroll associated Virus-3 (GLRaV-3), in New Zealand vineyards. Yet a recent survey of winegrowers commissioned by New Zealand Winegrowers (NZWG) indicated that there is a wide divergence of views on the importance of GLRaV-3 to the industry. The survey clearly indicated that more could be done to inform the industry about GLRaV-3 and its various effects, plus how it might best be managed. To that end, NZWG contracted HortResearch to Review the relevant research that has been carried out, both in New Zealand and internationally, in order to provide a platform for the NZWG programme to control the spread and impact of GLRaV-3 Identify priorities for further research based on knowledge gaps. Literature review databases (e.g. CAB Abstracts, VITIS VEA, Agricola and Web of Science), key technical reports to New Zealand MAF, the wine industry and New Zealand Winegrowers, and internet-based reviews were searched for key words associated with grapevine virus diseases. Although focused on GLRaV-3 the review included studies of other grapevine viruses, or other plant viruses, when it was considered that they provided a useful insight into GLRaV-3 or its vectors. When considering mealybug vectors, more than 2,000 published references in the scientific literature on mealybug biology and control, dating back over 100 years, are available. An appropriate selection of those papers and reviews, judged to make a significant contribution to knowledge of mealybugs in relation to plant viruses and grapes, was reviewed. The literature is presented in four sections, reflecting the different components of grapevine leafroll disease - GLRaV-3, grapevines and mealybug vectors Section 1: The GLRaV-3 virus Section 2: Impacts of GLRaV-3 infection in the vineyard Section 3: GLRaV-3 vectors and the ecology of transmission Section 4: Management options in vineyards to limit GLRaV-3 infection and spread. Key elements of these sections are summarised in Section 5. It became quite clear from the review that although the symptoms of grapevine leafroll disease have been recognized in New Zealand (and elsewhere) for more than 100 years, an understanding of the causes of the disease came slowly. Recognition that the disease was caused by a virus provided some early focus on the type of problem, even though initial emphasis on roguing (with no knowledge of a vector) did little to achieve control. The small size of the New Zealand wine industry in the 1960s to early 1980s, centred on American hybrid cultivars, provided little impetus to develop further understanding. The first implication of mealybugs as vectors of grapevine viruses overseas was followed by further laboratory experiments to identify the viruses they could transmit and to measure transmission efficiency. These studies were hampered substantially by a lack of tools to (a)

2 accurately identify the viruses themselves, and (b) measure low virus titres in both vines and mealybugs. Rapid advances in molecular biology and improved ELISA techniques then provided some rapid advances, but it is really only in the past 2-5 years that a reasonably clear understanding of the identity of the causal agents and vectors of grapevine leafroll disease have been elucidated. Even so, our understanding remains piecemeal. Molecular biologists have worked on virus taxonomy; virus epidemiologists have measured the spread of the disease in vineyards; entomologists have confirmed transmission in the laboratory; and grapevine physiologists have measured the effects of the disease on grapevines. Having said that, research programmes on all aspects of grapevine leafroll disease continue in different parts of the world, and there is increasing evidence that cross-disciplinary approaches are, and will continue, to increase our ability to understand and control the disease. It is now clear that grapevine leafroll disease is predominantly caused by the ampelovirus GLRaV-3, which, in New Zealand, is vectored between plants by three species of mealybugs. Other closteroviruses and insects may also be involved, at least sometimes. However, despite this basic knowledge of the mechanics of GLRaV-3 transmission and hence spread of grapevine leafroll disease, many questions with a direct bearing on the practical issue of how to manage the disease in New Zealands vineyards remain to be answered. Such fundamental questions include: What are the effects of GLRaV-3 on vine yield and wine quality in different cultivars and regions? How does the virus titre ebb and flow during the season? Is an apparent (observed) plant resistance due to a particular strain of virus, a different species/cultivar/clone of vine, or the ability of mealybugs to transmit the virus from one plant to another? How can we best measure the transmission ecology of different mealybugs in different regions on different cultivars/clones? What is the best or most appropriate strategy for vector control? How do ants, natural enemies and ground cover influence mealybug populations and hence the spread of GLRaV-3?

The search of the global scientific literature shows that there are no great advances in science elsewhere in the world that New Zealand can immediately appropriate to answer all of these questions. Valuable research in South Africa has given grape growers there quite extensive management guidelines through technical data sheets, but the key mealybug vector (Planococccus ficus) is not present in New Zealand, and, in any case, has a rather different biology to our species. In addition, some of the management options (e.g. use of broad spectrum insecticides) are not appropriate under SWNZ. The literature on mealybug vectors of other plant viruses does, however, illustrate how the interactions between the plant, environment, virus and vector may be unexpected, complex and interdependent. Some interactions that can occur, and perhaps might be investigated in grapes include the role of mealybug endosymbionts in virus transmission by mealybugs; the effects of a virus complex within the plant on mealybug transmission and disease symptoms; how plant resistance at the species and cultivar level, combined with plant nutrition and natural enemies, can affect disease symptoms; the effect of temperature on transmission efficiency. In reality, the combination of mealybug species, climate and vineyard environment that comprises the New Zealand wine industry is unique. New Zealand appears to be one of the very few countries that is warm enough for mealybugs to regularly reach high population densities, yet not warm enough to provide sufficient additional ripening to compensate for the effects of GLRaV-3 on grape quality and harvest.

3 It seems to us that the only way to provide successful solutions to grapevine leafroll disease in New Zealand is to establish an holistic, multi-disciplinary research programme with a specific goal to achieve those solutions. This is particularly so if the goal is to provide longterm management over the course of decades, rather than simply from year to year. The virus vine vector relationship (V3) should be viewed as a single entity, each component of which is inter-dependent on the other two. Research to understand V3 should be tackled by a multi-disciplinary team of plant virologists, entomologists, vine physiologists, pest controllers, vineyard managers, grapevine breeders/improvers and winemakers. Even though there would be many (some quite narrowly focused) research projects, this approach would provide the framework within which the many participants would be able to view all aspects of the V3 relationships. This would inevitably lead to better priority setting and design of research projects, better communication and understanding of the contributions from individual projects, better and more innovative projects incorporating appropriate ideas from overseas, and ultimately, better and more robust long-term solutions to the grapevine leafroll disease problem in New Zealand. This approach has been adopted in South Africa, where a coordinated, collaborative, multidisciplinary programme to study and control grapevine leafroll disease has been established. Researchers meet annually at a workshop to discuss progress, problems, and future studies, and research results are regularly provided to growers using a structured extension programme. If the concept of V3 takes off, then the design of a V3 research programme should take into account some of the key findings from this review: Section 1: GLRaV-3 the virus GLRaV-3 has probably been in New Zealand for more than 100 years, although it was first positively documented in 1964. Leafroll symptoms were seen in European Vitis vinifera, but not in US rootstock species. Virus-free vines resulting from thermotherapy were released by DSIR during the early 1970s. Vine pathology was first measured in the 1970s. Recognition of the similarity of disease symptoms to potassium deficiency provided an indication of the mode of pathology of the virus. Leafroll disease was first associated with closterovirus-like particles in the mid-1980s. Monoclonal antibodies and improved techniques allowed sensitive detection of GLRaV-3 from 2000. By 2003, 55 viruses known to infect grapevines had been identified globally. Reverse transcriptase polymerase chain reaction (RT-PCR) is the most sensitive, but difficult, technique for virus identification. Recent advances in virus taxonomy (based on molecular structure) have clarified the relationships between different grapevine viruses: GLRaV-3 is placed in the Ampelovirus group. Different groups of viruses are characterised, in part, by their transmission by different groups of insects. Ampelovirus is transmitted by mealybugs and scale insects. The existence of strains of GLRaV-3 with different pathogenicity has been postulated, but not yet demonstrated.

Section 1 Possible research topics Determine why different cultivars/species of grapevine infected with GLRaV-3 exhibit a range of symptoms (from mild to very severe). Investigate whether this is due to: o genetic tolerance within the plant to the presence of GLRaV-3 o Presence of more than one type of virus, or strain of GLRaV-3 o Presence of other viruses or infectious agents o Combination of a number of strains of GLRaV-3 differing in severity Determine whether different grapevine viruses in New Zealand are additive with respect to symptoms.

Section 2: impact of GLRaV-3 on vine growth and productivity. Early work in New Zealand supported overseas data that leafroll disease adversely affected vine growth, yield, fruit colour and sugar content. Degeneration of phloem cells in leaves, stems, and fruit petioles has been reported in GLRaV-3 infected vines. This is usually accompanied by an accumulation of starch in leaves, which may be the feedback mechanism whereby photosynthetic activities are shut down in infected leaves. GLRaV-3 can depress photosynthesis by 25-65% (depending on cultivar and environment), which is likely to directly affect all aspects of growth and cropping. The appearance of disease symptoms depends on many factors, but GLRaV-3 depresses photosynthesis even in vines that do not reveal visual symptoms. Foliar nutrients may alleviate the virus effects on photosynthesis. The effect of GLRaV-3 on root growth of infected vines is unclear. It is unclear whether GLRav-3 plays a role in graft incompatibility, although other viruses have been implicated. Viruses (including GLRaV-3) reduce weight and girth of canes, but large variation caused by environmental conditions makes the impact difficult to quantify. Increased vigour of virus-free vines may necessitate changes in vine management. A GLRaV-3 plus fanleaf virus infection reduced number, area and mineral content of leaves. GLRaV-3 infection may affect seasonal phenology through delayed budbreak, delayed flowering and delayed berry maturity, although these effects are not well quantified especially in New Zealand. GLRaV-3 infection may increase vine longevity due to reduced lifetime productivity. Evidence from around the world suggests that where present, GLRaV-3, either on its own or in combination with other virus and virus-like diseases, severely reduces grape yield among a wide range of white and red varieties. Similarly, among white and red varieties, evidence from a majority of studies suggests that in leafroll infected vines fruit sugar levels at harvest are lower, titratable acidity is higher and berry anthocyanin levels are reduced, compared with the juice from healthy vines. In almost half the studies where data were presented, fruit pH differed significantly from that found in healthy vines. Wine quality from GLRaV-3 infected vines is reduced. The effects of GLRaV-3 on quantitative and qualitative parameters of vine performance can impose severe limitations on wine production and quality.

Section 2 Possible research topics Identify and develop management tools for GLRaV-3 infected vines to achieve high must quality and yield and to manage timing of harvest (e.g. foliar fertilisation, crop loading) Identify and develop management tools for GLRaV-3 free (increased vigour) vines in the New Zealand environment (nutrition, summer pruning, botrytis control, etc.) to support virus free plantings. Determine the susceptibility of rootstocks to GLRaV-3 and the effect of the virus on graft compatibility and vine longevity. Identify GLRaV-3 tolerant/resistant Vitis species, identify method of tolerance/resistance (can it be translocated from rootstock to scion? identify gene level?), breed GLRav-3 tolerant/resistant rootstocks/vines. Quantify the impacts of GLRaV-3 infection on timing of budbreak, flowering and berry maturity. Quantify the long-term impacts of GLRaV-3 on vine yield in New Zealand cultivars and regions. Quantify the effects in New Zealand of GLRaV-3 on berry quality (especially sugars, titratable acidity and anthocyanins) and resulting wine (particularly sensory attributes). Section 3: GLRaV-3 vectors and ecology of transmission Three (3) species of exotic mealybugs feed on leaves, shoots, fruit, and occasionally roots of grapevines in New Zealand. None of the c.120 native species of mealybugs are known to feed on grapes. Twenty three (23) additional species worldwide are also known to feed on grapes. Seventeen (ex 26) species are considered to be pests of winegrapes. At least 10 of these are known plant virus vectors. Mealybugs are small, hemimetabolous, phloem feeding insects. They excrete large quantities of honeydew (on which grow sooty mould fungi). New Zealand species develop through 2-3 generations a year. Numbers increase greatly between spring and autumn. Mealybugs contain endosymbiotic bacteria, which may synthesise amino-acids essential for mealybug nutrition, and may also facilitate virus transmission. Transmission experiments indicate that mealybug transmission of plant viruses is usually, if not always, semi-persistent. Both mealybug and virus are required to induce mealybug wilt of pineapple (MWP) symptoms. MWP control strategies are either through the plant (meristem propagation of virus-free plants or development of transgenic plants through gene-silencing mechanisms) or the vector (mealybug control is especially through increased biological control). Presence of Argentine ant may be positively correlated with an increased rate of MWP virus spread. Unauthorised movement of cassava germplasm within and between continents has led to spread of pests. Control of mealybug vectors of cassava virus has been achieved by combinations of host plant resistance, biological control and cultural practices. Mealybug parasitoids were more effective if plants were selected for strong antixenosis but low antibiotic characteristics.

6 Environmental conditions (e.g. wet v. dry), mono- v. polycultures and Integrated Pest Management (IPM) methods all can have significant impacts on mealybugs and virus control in different crops. Plant viruses are typically able to be transmitted by several species of mealybugs, with differing efficiency. Virus spread can occur through airborne dispersal of young crawlers. Virus infections in some cocoa plants can remain latent for >2 years, and some cultivars/hybrids are more resistant to disease than others. Mild-strain cross protection is also developed in cacao. Grapevine fanleaf virus (GFLV) is transmitted only by a soil nematode. Development of resistant rootstocks using genes resistant to nepoviruses is under investigation. GVA and GVB (trichoviruses) are transmitted by the same species of mealybugs that transmits GLRaV-3. Pseudococcus longispinus transmits GLRaV-5 as well as GLRaV-3, but not -1 or -2. Mealybug vectors may acquire GLRaV-3 in much shorter times than those typical for mealybug acquisition of other plant viruses which have acquisition periods varying from 0.25-12h with retention of virus for 12h - 5 days. Virus may persist in a mealybug after post-acquisition moulting; but mealybugs may quickly lose the ability to transmit virus after transfer. Transmission efficiency of mealybugs is variable (typically 15-25%), but may be very high for Planococcus ficus and GLRaV-3 in South Africa. Efficiency may be greater after post-acquisition feeding than fasting. Temperature may be an important variable in transmission efficiency.

Section 3 Possible research topics Define the ecological relationships between mealybugs (especially Pseudococcus. longispinus and P. calceolariae) in different regions and their grapevine cultivar preferences and efficiency of GLRaV-3 transmission (i.e. transmission ecology and V3 interactions). Determine the role of mealybug endosymbionts in virus transmission, and seek ways to disrupt the endosymbiosis and hence improve GLRaV-3 management. Section 4: Managing GLRaV-3 infection and spread within the vineyard The international distribution of graft transmissible grapevine viruses is largely due to early indiscriminate exchange of graftwood and rootstocks. There has been little international interest in measuring or limiting the extent of the virus in national vineyards at least until recently. There is virtually no literature on management strategies to reduce or eliminate the spread of GLRaV-3, other than by reliance on certificated planting material. In some countries (e.g. USA, Spain) the role of mealybugs as vectors of GLRaV-3 has only recently been realised, and control strategies are not yet widely in place. By far the most advanced vineyard management systems in place for managing the spread of GLRaV-3 are in South Africa. Even here the plans are in technical guides, rather than scientific literature. Although the South African management strategies are useful in principle to New Zealand, fundamental differences mean that they are not entirely transferable. Key differences are: different mealybug vectors; heavy use of broad-spectrum insecticides; influence of several species of disruptive ants.

7 In South Africa, the most common pattern of GLRaV-3 spread occurs initially within rows between immediately adjacent infected vines before moving across rows and vineyards through various combinations of: mealybug dispersal and vineyard machinery and personnel. Removal of infected vines (roguing) may reduce the rate of spread of GLRaV-3 where the incidence of primary and secondary infected plants is low and where latent infection is not excessive, but it is not expected to be effective under all conditions. Latent infection in asymptomatic vines is managed by spraying targeted insecticides. Edge-effects can indicate the spread of disease from other vineyards. Management strategies include planting large, isolated vineyards with shelterbelts, upwind from existing virus sources; and working daily in youngest, least infected blocks first. Good hygiene, especially when replanting an old infected vineyard, is important to minimise point sources of infection. Some grape cultivars are more susceptible to mealybugs than others. Beware of transmitting GLRaV-3 through top-grafting vines to new cultivars. Mealybug control is the key to GLRaV-3 control, especially when vineyards are susceptible to sustained mealybug immigration. Mealybugs are monitored by a combination of physical examination of vines and pheromone traps (for Planococcus ficus only). Pheromones are new tools that help to reduce the labour intensive monitoring and determine the need for control actions. Pesticide application are recommended to targeted vines during dormancy, to minimise negative effects on natural enemies. Common insecticides in S. Africa are either not registered in New Zealand, or are not compatible with SWNZ. There are widespread control failures due to insecticide resistance. Biological control of P. ficus can provide effective control. Ant control (including Argentine ant) is crucial for mealybug control in S. Africa. Appropriate cover crop and weed control strategies are an important part of managing ants, mealybugs, natural enemies and hence the spread of GLRaV-3. Section 4 Possible research topics Design and implement GLRaV-3 management strategies based on New Zealand fauna, industry practices (especially SWNZ) and environments. Develop new monitoring strategies using sex pheromone traps as key resources to rapidly locate small patches of mealybugs in large or small vineyards. Develop new biological control strategies through the introduction of new natural enemies or manipulation of existing species (including options of augmentative control). Determine the impact of Argentine ant (and other ant species in New Zealand vineyards) on mealybug control and spread of GLRaV-3 in New Zealand. Develop new mealybug control options using safe chemistry or non-chemical means. Quantify dispersal behaviour and distances moved by mealybug males and crawlers, and the benefits from shelterbelts or other physical barriers. Design the best (most appropriate) vineyard plant biodiversity for maximum disruption to mealybugs and hence spread of GLRaV-3 in New Zealand. The suggested research topics are by no means the only ones that could or even should be undertaken. At least one project has already started (a NZWG funded project to measure the effects of grapevine viruses on Sauvignon Blanc in Marlborough). However, they hopefully

8 provide a platform for next phase of the programme to establish industry and research priorities for managing GLRaV-3 in New Zealand. It is a daunting task, with an almost endless set of possible questions to answer from an inevitably limited research budget. Most vineyards will remain under sustained annual pressure from mealybug vectors, yet the goal is to manage the spread of grapevine leafroll disease over a possibly 50+year life span of a vineyard. The table ES1 (with preliminary estimates of technical difficulty, outcomes and time-frames etc.) provides a framework within which possible projects or strategies can be viewed and prioritised.

For further information, please contact:

John Charles (mealybug ecology) Dan Cohen (grapevine virus) HortResearch Mt Albert 120 Mt Albert Road, Private Bag 92169 Mt Albert, AUCKLAND, NEW ZEALAND Tel: +64-9-815 4200 Fax: +64-9-815 4201 Email: jcharles@hortresearch.co.nz dcohen@hortresearch.co.nz Or Jim Walker (mealybug control) HortResearch Hawkes Bay Cnr Crosses and St Georges Roads Private Bag 1401, HAVELOCK NORTH, NEW ZEALAND Tel: +64-6-877 8196 Fax: +64-6-877 4761 Email: jwalker@hortresearch.co.nz

Suggested research and implementation activities for reducing GLRaV-3 disease impact and virus transmission and enhancing control of mealybugs in vineyards; their limitations, ease of implementation and indicative costs and benefits. S1, S2, S3, S4 = see Sections 1, 2, 3 and 4 respectively L = low; M = medium, H = high e = easy, m = moderate, d = difficult St = short timeframe (<4y); It = intermediate timeframe (4-7y); Lt = long timeframe (>7y)

Table ES1

Proposed Action

Key assumptions

Use elsewhere

Key outcome(s) expected Research difficulty (e,m,d) Risk of failure (L,M,H)

Likely impact on target (L,M,H) Time to develop (St, It, Lt)

Ease to implement (e,m,d)

Relative cost to grower (L,M,H)

MEDIUM-LONG TERM STRATEGIC RESEARCH OPTIONS Vine Resistance To Virus And Vector Species, cultivars, tolerance / S1. Identify GLRaV-3 scions, rootstocks, resistance tolerant / resistant Vitis show differential noted in some species resistance to species and GLRaV-3 hybrids M-H m M Yes e.g. mildew, black spot on apples M -H Increased resilience to the presence of GLRaV-3 in NZ vineyards Ability to manipulate resistance in vines d GLRaV-3 resistant vines M-H d M

It

Species / cv or clones show differential resistance to GLRaV-3 Yes e.g. mildew, black spot on apples Yes e.g., cassava, avocado

Lt

Lt

M L in long term

S1. Identify mechanisms of tolerance/resistance (biochemistry, gene identification, graft transmissibility) S1. Develop or select GLRaV-3 tolerant/resistant rootstocks/varieties (gene insertion, microarray) S3. Detect, measure and select for, host plant resistance to mealybugs Lower populations of mealybugs on vines. Slower rate of spread of GLRaV-3 M-H

Tolerance / resistance to GLRaV-3 is gene based 1. Resistant species or cultivars of Vitis exist 1a. Nutritional factors in Vitis may deter or limit mealybug presence 1b Genes in Vitis spp

L- M

St-It

e-m

M (L in long term)

10

Proposed Action

Key assumptions

Use elsewhere

Key outcome(s) expected Research difficulty (e,m,d) Risk of failure (L,M,H) Time to develop (St, It, Lt)

Likely impact on target (L,M,H) Ease to implement (e,m,d)

Relative cost to grower (L,M,H)

for resistance to mealybugs can be identified L-M m M It - Lt m L in longterm

Mealybug (Vector) Control S4. Improved control of mealybugs in vineyard landscape Widespread many successes on other mealybug species M-H d H Lower populations of mealybugs on vines. Improved biocontrol of mealybugs and hence lower mealybug populations It - Lt

S4. Introduce new exotic parasitoids for mealybug control (Classical biological control)

M (L in long term)

New technologies will soon be available for NZ mealybug species 1 Able to locate and import suitable parasitoids 2 Climatic/ecological compatibility 3 No non-target effects on indigenous insects. 4 Approval from ERMA/MAF Commercial supplies of natural enemies are available Yes, mostly in glasshouses M-H Improved biological control d H It- Lt

S4. Develop augmentative biological control parasitoids or predators Manipulate Ecology Of Transmission S3. V3: Understand the vine/virus/vector interactions Hardly developed. Mealybug control or vine removal is usual. Development of the ideal NZ vineyard to minimize the effects and spread of GLRaV-3 M-H

There is a 3-way, interlinked relationship between vine, virus and vector Managing GLRaV-3 requires deep knowledge of the whole relationship No. Unproven concept 1.Fewer mealybugs on

St-Lt

e-m

L-H (L in longterm

S3. Disrupt endosymbiont relationship with mealybug

1.Mealybugs require endosymbionts for

1. M-H

Lt

unknown

unknown

11

Proposed Action

Key assumptions

Use elsewhere

Key outcome(s) expected Research difficulty (e,m,d) Risk of failure (L,M,H) Time to develop (St, It, Lt) vines 2. Reduced disease incidence; slower spread of virus 2. M-H

Likely impact on target (L,M,H) Ease to implement (e,m,d)

Relative cost to grower (L,M,H)

survival 2. GLRaV-3 requires endosymbionts for transmission

SHORT-MEDIUM TERM STRATEGIC RESEARCH OPTIONS Yes under development in NZ, RSA, USA. L m L Australia Accurate assessment of mealybug population size and dispersal St-It e L

S4. Develop monitoring tools for mealybugs in vineyards especially trapping to measure dispersal (natural and via human activity)

S4. Control ants

Synthetic pheromones are effective lures for winged male mealybugs within the vineyard Sticky traps measure long-distance immigration Ants, especially Argentine ant, limit mealybug control by natural enemies Ant control improves mealybug biocontrol by parasitoids H Improved biocontrol L-M m L

St-It

e-m

S4. Enhance/conserve natural enemies in vineyard

Yes- this species disrupts mealybug control in many countries Yes but limited impact so far

It

m-e

S2. Identify and develop management tools for infected vines Yes crop management specific to crop

Yes other diseases on various crops

L-M

St

S2. Identify and develop management tools for virus free vines

Parasitoids populations enhanced in vines, inter-row, headlands and local environment Yield and must quality can be increased by correct vine management Virus elimination and subsequent vine growth in NZ will Increased yield and must quality in existing vines Increased yield and must quality

St-It

12

Proposed Action

Key assumptions

Use elsewhere

Key outcome(s) expected Research difficulty (e,m,d) Risk of failure (L,M,H) Time to develop (St, It, Lt)

Likely impact on target (L,M,H) Ease to implement (e,m,d)

Relative cost to grower (L,M,H)

induce high vegetative growth M e-m L St-It e

S2. Determine GLRaV-3 effect on graft compatability and vine longevity

GLRaV-3 infection alters graft incompatibility and/or vine longevity

and environment characteristics Yes graft compatibility known to affect yield and plant longevity Yessuccessful in S. Africa H e-m L Increased ability to determine impacts of GLRaV-3 on NZ wine industry List of interrow and vineyard plant species for NZ conditions It-Lt Direct chemical control beneficial to predators if naturalenemyfriendly chemicals applied M-H e-m L It e

S4. Develop understorey biodiversity for mealybug control

vineyard plant biodiversity can be designed to enhance mealybug control

SHORT-TERM STRATEGIC RESEARCH OPTIONS 1. All actively On citrus S4. Chemical Controlfeeding mealybugs Direct injection are exposed to Applies to all insecticide. sap-feeding 2. Residues insect pests on acceptable at harvest orchard plants

S2. Quantify impacts of GLRaV-3 on vine phenology

Infection changes timing of bud break, flowering and veraison

Increased ability to determine impacts of GLRaV-3 on NZ wine industry

St

Results not designed for vineyard

13

Proposed Action

Key assumptions

Use elsewhere

Key outcome(s) expected Research difficulty (e,m,d) Risk of failure (L,M,H) Time to develop (St, It, Lt) St Results not designed for vineyard L e L

Likely impact on target (L,M,H) Ease to implement (e,m,d) L

Relative cost to grower (L,M,H)

S2. Quantify long term effects of GLRaV-3 infection on vine yield in NZ

Infection reduces vine yield

Increased ability to determine impacts of GLRaV-3 on NZ wine industry Increased ability to determine impacts of GLRaV-3 on NZ wine industry M L St Yes H e Effective strategy for control of mealybug incursion(s) Better spray timing, cheaper, more effective monitoring M-H m L St m-d

Results not designed for vineyard

Infection changes S2. Quantify effects of berry and wine GLRaV-3 infection on characteristics berry quality and resulting wine quality (trials on Sauvignon Blanc in Marlborough underway, February 2006) Implementing or improving existing technologies Planning for S3. Biosecurity and potential mealybug contingency planning pest incursions on Vitis will lead to effective management Pesticide timing and/ S4. Improved forecasting or scouting for (using development rate mealybugs can be models) linked to decisionimproved by support systems prediction of mealybug phenology Yes. Widespread in New Zealand M-H

It-Lt

14

15

PREFACE AND TERMS OF REFERENCE

PREFACE
A key goal of the New Zealand Wine Industry is to control the spread and impact of virus diseases, particularly Grapevine Leafroll associated Virus type 3 (GLRaV-3), in New Zealand vineyards. Yet a recent survey of winegrowers commissioned by New Zealand Winegrowers (NZWG) indicated that there is a wide divergence of views on the importance of GLRaV-3 to the industry. Some vineyard owners observed such significant deterioration of both quantity and quality of grapes from blocks infected with GLRaV-3 that all the vines were removed, and the land treated and fallowed before replanting. Conversely, other vineyard owners reported that some varieties, known to be infected with the virus, continued to produce grapes that yielded premium wines. Yet other growers do not appear to be aware of the existence of viruses in grapevines. There were similar differences in understanding of the virus life cycle resulting in somewhat ad hoc application of recognised best practices for GLRaV-3 management in grapevines. It may be as a consequence of such ambivalence within the industry that the presence of GLRaV-3 in New Zealand vineyards is not declining significantly and may, in fact, be increasing (although any purported increase may be attributed in part to increased awareness). The survey clearly indicated that more could be done to inform the industry about GLRaV-3 and its various effects, plus how it might best be managed (see the Chain of Custody concept implicit in the new Grapevine Health Standards). To that end, NZWG contracted HortResearch to: Review the relevant research that has been carried out, both in New Zealand and internationally, in order to provide a platform for the NZWG programme to control the spread and impact of GLRaV-3 Identify priorities for further research based on knowledge gaps. HortResearch agreed to: Undertake a review of the New Zealand and international scientific literature reporting the research on all aspects of the incidence, identification, impact (on both quantity and quality of grapes), spread (including vectors), management, and control of GLRaV-3 Provide a summary of key findings Identify areas where further research is warranted to provide base information that will improve GLRaV-3 management. Both parties understood that the review was intended to provide the base documentation from which appropriate research, education and extension programmes could be developed. It was accepted that the review would NOT make comment on the development and maintenance of a high health planting scheme as NZWG had recently completed this exercise. However, the use of high health material as being good horticultural practice would be touched on for completeness. Review Structure HortResearch consulted with NZWG (particularly Rod Bonfiglioli, Diane Stewart and Philip Manson) to confirm the structure of the review report. This would contain sections on the virus, its impact on grapevines, its transmission by insect vectors, and management options. A

16 summary of the key findings would be prepared, gaps in knowledge identified, and draft recommendations for potential R & D written for further discussion. Personnel It was agreed that HortResearch would draw on the available expertise within the Institute, covering virology - including virus taxonomy and detection (Dr. Dan Cohen); entomology including insect vector ecology and insect/virus interactions (Dr Jim Walker, Mr. John Charles, Mr. Vaughn Bell and Dr. Shaun Forgie) and plant physiology (Mr. Ken Breen) to produce the initial draft of the review document. Draft and Final versions of the review HortResearch presented a final draft of the review to NZ Winegrowers in February 2006. NZWG submitted the draft for international review, which was completed by late June 2006. Comments made by the reviewers were discussed at a subsequent meeting in July between Winegrowers and HortResearch, and the extent of consequent changes/additions to the draft review were accepted. The final review was then sent to NZ Winegrowers in August 2006.

17

METHODS AND STRUCTURE OF THE REVIEW

Key words and names of authors to identify studies with links to grapevine virus diseases, vine performance, and virus vectors were entered into a number of literature review databases (e.g. CAB Abstracts, VITIS VEA, Agricola and Web of Science). To ensure that important material was not overlooked during this search, the bibliographies of many of the identified papers were also checked. Technical reports to New Zealand MAF, the wine industry and New Zealand Winegrowers since the association of GLRaV-3 with mealybugs was first mooted were examined. Several internet based reviews were also accessed and assessed. Literature in foreign languages was reviewed where possible, and analysis included when English language abstracts and captions for figures and tables provided a useful insight into the study methodology and results. Others, where no English translation was available, were excluded where relevance and/or usefulness were difficult to determine. Although the review focuses primarily on GLRaV-3, studies of other grapevine viruses, or other plant viruses were included when it was considered that they provided a useful insight into this primary goal. Many studies were published before the current taxonomic classification of the leafroll virus complex was determined, and there are several studies that were published after the virus was described but where the disease was simply defined as leafroll virus. It is therefore unclear if the results outlined in all these studies could be directly attributed to GLRaV-3 or were instead the result of another leafroll virus or perhaps a range of virus and virus-like diseases. Where the name GLRaV-3 was explicitly used in a study, it is included in this review, otherwise the descriptive term leafroll virus, or GLR unassigned is often used. There are more than 2,000 published references in the scientific literature on mealybug biology and control, dating back over 100 years. An appropriate selection of those papers and reviews, judged to make a significant contribution to knowledge of mealybugs in relation to plant viruses and grapes, was reviewed. The literature review is reported in four sections, reflecting the different elements of GLRaV3, grapevines and mealybug vectors: Section 1: The GLRaV-3 virus Section 2: Vineyard impacts of GLRaV-3 infection Section 3: GLRaV-3 vectors and the ecology of transmission Section 4: Management options within the vineyard to limit GLRaV-3 infection and Spread. In a fifth section, some of the key features of the previous four sections have been extracted in bullet format (key findings). From these bullets, the authors have posed a number of relevant questions, in the form of potential research projects. The projects are intended to highlight questions that have not been satisfactorily addressed in either New Zealand or internationally, and that would help to improve management of grapevine leafroll disease in New Zealand. The executive summary attempts to encapsulate Section 5 into an holistic overview of the potential management of GLRaV-3 in New Zealand, and provides a table of suggestions from the authors of various research projects that could contribute to the goal of reducing GLRaV-3 disease impact and virus transmission, and enhancing control of mealybugs in vineyards. It includes very preliminary estimates of their limitations, ease of implementation and relative costs and benefits.

18

SECTION 1 THE GLRaV-3 VIRUS

Early reports of grapevine leafroll disease Scheu described a virus disease of grapes in Germany that involved the downward rolling of the margins of leaves (Rollkrankheit) and he estimated that over 80% of vines planted in Germany were infected (Scheu 1936, 1950). Some varieties were completely infected and the varietal descriptions were based on leafroll-infected plants. A number of grape diseases with similar symptoms have been described in France such as brunissure, rougeau and flavescence and Goheen et al. (1958) considered these to be the same as leafroll. Scheu recognised that leafroll-like symptoms could be caused by a number of other factors and listed mechanical bending, Botrytis infection, plough wounds and nutritional disorders. A key factor that distinguished leafroll virus from other factors was the ability to transmit symptoms from symptomatic vines to healthy vines. Of particular interest was the recognition that potassium deficiency caused very similar symptoms to leafroll. Herschler (1936) found scattered vines in German vineyards with symptoms of potassium deficiency in autumn and these vines had a very low level of potassium in their leaves although soil levels of potassium were high. Other symptoms associated with the disease include poor fruit quality and low fruit-sugar content (Goheen et al. 1958). Leafroll symptoms were seen on vinifera grapes but have little or no effect on American rootstocks. In California, a disease of Emperor grapes referred to as white Emperor disease was found to be identical to leafroll (Goheen et al. 1958). These authors reported that although there may be some natural spread in German vineyards, indications are that there is very little if any natural spread in California. The first published account of virus diseases in New Zealand was a paper by McKissock (1964) who recorded the presence of both grapevine fanleaf and grapevine leafroll (leaf-roll). With hindsight, we recognise that Bragato (1902) in a report on grapegrowing in New Zealand, described symptoms suggesting that the disease was present in New Zealand over 100 years ago. Bragato noted that many vines of the variety Cabernet Sauvignon failed to fruit and these vines displayed early reddening of leaves. McKissock was able to distinguish leafroll from potassium deficiency by graft transmission from symptomatic vines to virus-free indicator plants imported from the University of California. McKissock referred to work by Goheen & Cook (1959) that showed potassium deficiency symptoms first appear on leaves in the middle of canes in early summer. In grapevines with leaf-roll virus there is an increased carbohydrate content and a decreased amount of potassium in the leaf blade, and a blockage and accumulation of potassium in the leaf petiole. In 1960 a survey of grapevine virus diseases was started in Auckland and Hawkes Bay vineyards. A preliminary report on this work was published in the Wine Review (Chamberlain 1967) and listed many grapevine cultivars showing clear-cut leafroll symptoms in the field. He also listed the major effects of leafroll infection as reduced yield, sugar content and colour in red-fruited varieties and delayed ripening of fruit. He also noted that hybrid grape varieties were grown in preference to European varieties because they were not as seriously affected as hybrid varieties. In a later paper Chamberlain et al. (1970) discussed the early history of grapevine introductions to New Zealand, including the introduction of phylloxera-resistant rootstocks and propagation methods used. These authors noted that since it is only in comparatively recent years that grapevine viruses have been generally recognised and their significance appreciated, it is probable that many of the introduced vines were virus infected. The

19 practice of regrafting vines to new varieties was noted in by Bragato (1903) and by 1970 Chamberlain et al. reported that they were unable to find any Pinot Meunier or Pinot Chardonnay (Chardonnay) vines free from leafroll symptoms. Vine pathology Over de Linden & Chamberlain (1970a) reported the effect of leafroll virus on white (Baco 22A) and red (Mission) varieties over 6 years. Buds from an infected Pinot Meunier vine were used to inoculate virus-free cuttings of the 2 selected cultivars. Twelve healthy and 12 infected vines of each variety were planted in December 1963, and symptoms, fruit yields, sugar and tartaric acid levels were measured from 1966-69. In summary they found: 1. By harvest time each season many lower leaves of Mission vines were reddish-purple with downward rolling of leaf margins and the fruit was just beginning to colour, whereas the leaves on healthy vines were still green and not rolled, and fruit had developed full colour 2. For Baco 22A, no difference in leaf or fruit colour between infected and healthy vines was detected at harvest but leaves of infected vines showed yellow vein banding compared with uniformly yellow leaves on healthy vines 3. Infected vines of both varieties showed later bud burst on infected vines with flowering delayed by 10-14 days 4. Vine growth of infected Mission vines was reduced, but no differences in vine growth were seen between infected and healthy Baco22A 5. For both varieties, infected vines yielded significantly less fruit 6. In most seasons, sugar levels were significantly reduced in infected vines of both varieties 7. No significant differences were detected in tartaric acid levels 8. Pigments in juice and wine made from grapes harvested from infected Mission vines were significantly reduced compared with healthy Mission controls. Over de Linden & Chamberlain (1970a) attributed the reduced symptoms on infected Baco 22A compared with those on infected Mission vines to the hybrid origin of the Baco 22A variety. It was known that American Vitis species did not show symptoms of leafroll, but, although not noted by these authors, symptoms are much harder to detect on white grape varieties. Table 1-1 Gross Returns from healthy and virus infected Cabernet Sauvignon and Pinotage grapevines based on 1975 prices. From Thomas (1976). Variety Yield tons/ acre 4.50 11.20 Healthy Sugar Price/ content ton $ (Brix) 21.00 260 20.50 220 Total return $ 1170 2464 Yield tons/ acre 1.50 9.20 Infected Sugar Price/ content ton $ (Brix) 14.50 80 18.77 176 Total return $ 120 1619

Cabernet Sauvignon Pinotage

Work on grapevine virus diseases at the Mt Albert Research Centre of the then Department of Scientific and Industrial Research (DSIR), continued through the 1970s and some of this later work was reviewed in a short paper in Wine Review (Thomas 1976). Data presented in this paper clearly demonstrated delayed accumulation of sugar and higher tartaric acid levels in leafroll-infected Cabernet Sauvignon. Economics of growing grapes free of known viruses was also discussed and Table 1-1 is taken from this paper. Viruses associated with Grapevine leafroll disease The identification of the viruses associated with leafroll infected vines has been a long task, made difficult by the presence of graft incompatibility-associated viruses as well as leafroll in

20 many infected vines. There have been reports of an association with a potyvirus (Tanne et al. 1977, 1989) and closteroviruses (Namba et al. 1979, Faoro et al. 1981, Castellano et al. 1983, Mossop et al. 1985). However some viruses previously described as closteroviruses are now classified in newly created genera and some of the viruses studied were probably associated with corky bark and stem pitting diseases. The nomenclature of grapevine leafroll-associated viruses was discussed at the 10th Meeting of the International Council for the Study of Viruses and Virus diseases of the Grapevine in 1990. This meeting proposed using the name Grapevine leafroll-associated virus followed by Roman numerals I to V (e.g. GLRaV-I). However the International Committee on Taxonomy of Viruses (ICTV has determined that virus acronyms that have numbers are to be written in Arabic numerals separated by a hyphen from the letters. By 1995 there were 6 types recognised and Boscia et al. (1995) referred to these viruses as GLRaV-1 to GLRaV-6. The classification of filamentous viruses associated with a variety of virus diseases in plants has undergone a series of revisions over the past 10 years. Much of the work has involved detailed sequencing of viral genomes and identification of the order of genes in these genomes. Until the molecular studies and been carried out, long thin flexuous particles with a length of above 900 nm as seen in the electron microscope were called clostero-like (Clostero = thread) and were tentative members of the Clostervirus genus. Most of these viruses had a diameter of 10-13 nm. Sometimes shorter length particles were also referred to as closterolike, even though their particle lengths were 650-850 nm. These viruses are now classified into 2 distinct virus families Closteroviridae and Flexiviridae. The Closteroviridae has 3 genera, Closterovirus, Ampelovirus and Crinivirus, based on genome organisation, conservation of key gene sequences and mode of transmission (Martelli et al. 2002).. The Flexiviridae has 9 genera, of which 3 contain viruses that infect grapevines, viz Vitivirus, Foveavirus and Trichovirus, although there has been one report of a Potexvirus (Potato virus X) in Tunisian grapevines (Chabbouh et al. 1993). All members of the Closterovirus genus, with the exception of Grapevine leafroll-associated virus 2 (GLRaV-2) and Grapevine stem lesion-associated virus (GSLaV) (which is now recognised as a molecular variant of GLRaV-2 (Martelli 2003) are transmitted by aphids. The vector for GLRaV-2 is unknown, but is unlikely to be a mealybug or scale insect. Some species of Closterovirus can be sap transmissible, although with difficulty, and this is the case with GLRaV-2 between some tobacco species. The type member of the genus Ampelovirus is Grapevine leafroll-associated virus 3 (GLRaV3). This genus also includes GLRaV-1, 4, 5, 6, 8 and 9 (and possibly, as discussed by the ICVG) other GLRaV-types e.g. -10, -11, -12. Numerous studies have demonstrated transmission of GLRaV-3 in vineyards by several species of mealybugs. GLRaV-1 can be transmitted by the soft scale Parthenolecanium corni and the mealybugs Heliococcus bohemicus and Phenacoccus aceris (Sforza & Greif 2000, Sforza et al. 2003). The majority of other members of this genus can be transmitted by either mealybugs or soft scales under experimental conditions. We are unaware of any report showing mechanical transmission of ampeloviruses to other plants. The type member of the Crinivirus is Lettuce infectious yellows disease. This virus and some other members of the genus Crinivirus are transmitted by whiteflies. None of the members of Crinivirus are known to infect grapevines. Grapevine leafroll-associated virus 7 (GLRaV-7) is a member of the family Closteroviridae but has not yet been assigned to a genus. In addition, there are several new Grapevine leafrollassociated viruses that have yet to be characterised, some of which appear to be related to GLRaV-4, -5, and -6. In the Flexiviridae, the genus Vitivirus has 3 species that infect grapevines - Grapevine virus A, B and D (GVA, GVB and GVD). GVA and GVB can be transmitted by mealybugs and can

21 also be transmitted by sap inoculation to several tobacco species, but not to grapevines. Grapevine rupestris stem pitting-associated virus (GRSPaV) is a member of the genus Foveavirus. Although this virus has no known insect vector, it is probably the most widespread virus infecting grapevines worldwide. Grapevine berry inner necrosis virus is a member of the genus Trichovirus (Yoshikawa et al. 1997). A list of species in families Closteroviridae and Flexiviridae from the NCBI website (http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi) is presented in Table 1-2. Grapevine virus C, which was previously listed as a tenetative species of Vitivirus is now considered to be serologically related to GLRaV-2 (Masri at al. 2006) What types of grapevine leafroll associated viruses are present in New Zealand? Groups in Europe and the USA have produced polyclonal and monoclonal antibodies to purified preparations of virus extracted from leafroll-infected grapevines (Gugerli et al. 1984, Teliz et al. 1987, Zimmerman et al. 1990a, b). Antibodies to a range of Grapevine leafrollassociated viruses are now available from commercial companies in Switzerland, France and Italy. These antibodies provide a tool for scientists to determine which types of leafroll were present in vines displaying symptoms. In New Zealand, a survey of vineyards using ELISA for GLRaV-1 and GLRaV-3 showed that GLRaV-3 was detected most frequently and only GLRaV-3 was associated with secondary spread within the vineyard (Petersen & Jordan 1992). Since that time extensive testing for these viruses by both HortResearch and Linnaeus Laboratories confirm the widespread occurrence of GLRaV-3. Other than in the variety Chardonnay Mendoza which is 100% infected with GLRaV-1, this virus is detected at a much lower frequency and there has been only one report of spread of GLRaV-1 in New Zealand (Bonfiglioli et al. 2001). Over the past 10 years a large number of New Zealand grapevine samples have been tested by RT-PCR at Waite Laboratories in Adelaide and Linnaeus Laboratories in Gisborne for a range of grapevine leafroll-associated virus species. In addition, as new improved monoclonal antibodies have been produced overseas, ELISA tests for GLRaV-2, GLRaV-5 and GLRaV-6 have also been carried out in New Zealand. These tests have detected GLRaV species 2, 4, 5 and 9 in one or more varieties. GLRaV-2 is the only leafroll-associated virus in the genus Closterovirus and this virus is also associated with graft incompatibility problems (Boscia et al. 1995). It has been detected in a range of varieties imported from Bordeaux in the 1980s. GLRaV-5 has been detected only in Tempranillo imported from Australia in 1990 and GLRaV-9 is present in Cabernet Sauvignon SA125, imported from Australia in 1974. Recently Linnaeus Laboratories have detected GLRaV-4 in a clone of Sauvignon Blanc (Bonfiglioli & Edwards pers. comm.).

22 Table 1.2. Viruses in the Closteroviridae and Flexiviridae. Species that infect grapevines are shown in bold.
Family Closteroviridae Genus Ampelovirus Family Flexiviridae Genus Foveavirus

Grapevine leafroll-associated virus 1 Grapevine leafroll-associated virus 3 Grapevine leafroll-associated virus 5 Little cherry virus 2 Pineapple mealybug wilt-associated virus 1 Pineapple mealybug wilt-associated virus 2
unclassified Ampelovirus

Grapevine leafroll-associated virus 4 Grapevine leafroll-associated virus 6 Grapevine leafroll-associated virus 8 Grapevine leafroll-associated virus 9 Plum bark necrosis and stem pitting-associated virus
Genus Closterovirus

Apple stem pitting virus Apricot latent virus Peach asteroid spot virus Peach sooty ringspot virus Cherry green ring mottle virus Cherry necrotic rusty mottle virus Grapevine rupestris stem associated virus
unclassified Foveavirus

pitting-

African oil palm ringspot virus Asian prunus virus 1 Asian prunus virus 2 Asian prunus virus 3 Prunus mume foveavirus
Genus Potexvirus

Beet yellow stunt virus Beet yellows virus Citrus tristeza virus Citrus tristeza virus strain T36 Grapevine leafroll-associated virus 2
unclassified Closterovirus

Fig leaf mottle-associated virus Mint virus 1 Olive leaf yellowing-associated virus
Genus Crinivirus

Abutilon yellows virus Beet pseudo-yellows virus Cucurbit yellow stunting disorder virus Lettuce infectious yellows virus Sweet potato chlorotic stunt virus Sweet potato sunken vein virus Tomato chlorosis virus Tomato infectious chlorosis virus unclassified Crinivirus Bean yellow disorder virus Blackberry yellow vein virus Cucumber yellows virus Potato yellow vein virus Strawberry pallidosis associated virus
unclassified Closteroviridae

Alternanthera mosaic virus Bamboo mosaic virus Cactus virus X Cassava common mosaic virus Clover yellow mosaic virus Cymbidium mosaic virus Foxtail mosaic virus Hosta virus X Hydrangea ringspot virus Lily virus X Narcissus mosaic virus Nerine virus X Papaya mosaic virus Pepino mosaic virus Plantago asiatica mosaic virus Nandina mosaic virus Potato aucuba mosaic virus Potato virus X Scallion virus X Strawberry mild yellow edge virus Tulip virus X White clover mosaic virus
Genus Trichovirus

Grapevine leafroll-associated virus 7 Little cherry virus 1 Mint vein banding virus

Apple chlorotic leaf spot virus Cherry mottle leaf virus Grapevine berry inner necrosis virus Peach mosaic virus
unclassified Trichovirus

Apricot pseudo-chlorotic leaf spot virus

Genus Vitivirus

Grapevine virus A Grapevine virus B Grapevine virus D Heracleum latent virus

unclassified Vitivirus

Mint virus 2

From these testing results, it is apparent that leafroll disease as first described in New Zealand (McKissock 1964, Chamberlain et al. 1970) was predominantly what is now known as GLRaV-3, but the presence of other viruses in these vines may have contributed to the severity of symptoms.

23 These results are in contrast to the situation in Australia. Habili et al. (1996) reported on the leafroll-associated types present in a range of leafroll infected varieties and clones obtained from CSIRO collections and private vineyards. GLRaV-1 was found in 8/18 Sultana clones described by Antcliffe et al. (1979) and many of these clones also contained GLRaV-4 and GLRaV-5. GLRaV-1 was also found in Zante Currant and a low yielding Cabernet Sauvignon clone. GLRaV-3 was detected in several Pinot Noir clones and in a number of other varieties, but was not the dominant type. There was evidence of considerable spread of GLRaV-3 in this Pinot Noir block (Habili et al. 1996, Habili & Nutter 1997) but no mealybugs or scales were observed on these vines. Habili & Symons (2000) later reported the results of viruses detected from almost 2500 samples using RT-PCR. GLRaV-1, 2 and 3 were detected in 3.7%, 2.4% and 4.2% of samples respectively. Over the past 5 years, the proportion of samples that Waite Diagnostics have found to be positive for GLRaV-3 has markedly declined. This is attributed to increased awareness of leafroll problems and the success of roguing when infected vines were detected (Habili pers. comm.). Virus strains of GLRaV-3 Different cultivars of grapevine infected with GLRaV-3 can exhibit a range of symptoms from very mild to severe. From the earliest descriptions of leafroll disease, it has been recognised that symptoms are most severe with black or red-fruited vinifera varieties. On some green-fruited V. vinifera varieties symptoms appear to be less severe and most rootstock and some hybrid varieties may be infected yet show no foliage symptoms. However, the growth of vines, the yield of fruit and sugar levels may be reduced. There are a number of possible explanations for these observations: 1. 2. 3. 4. Some varieties may be genetically tolerant to the presence of GLRaV-3 There may be more than one strain of leafroll present in the variety There may be other viruses or infectious agents present in the variety There may be a number of strains of GLRaV-3 differing in severity.

Evidence can be presented to support each of these explanations. Evidence for different strains of GLRaV-3 differing in severity comes from two analyses of sequence variation in different isolates of GLRaV-3 (Turturo et al. 2005, Jooste & Goszczynski 2005). Both studies showed that the genome of GLRaV-3 shows low molecular variability. For most of the regions selected for study the isolates fitted into two main groups. At this stage there is no indication that these groups are related to the biological properties of the strains studied. Both papers referred to studies carried out on strains of Citrus tristeza virus (CTV), an aphid transmitted member of the genus Closterovirus. In CTV, partial correlations between molecular divergence and differences in pathogenic properties of isolates have been found (Ayllon et al. 2002). Studies in New Zealand at Linnaeus Laboratories indicate that some strains of GLRaV-3 that show up as weakly positive in ELISA have sequences that diverge markedly from GLRaV-3 sequences so far published (Belton et al. 2006). The presence of divergent strains and the relationships between strains and pathogenicity is an area that requires further work. Two papers are scheduled for presentation to the International Council for the Study of Virus and Virus-like Diseases of the Grapevine (ICGV) conference in South Africa in April 2006 (q.v. page 29) which may shed more light on this topic. Other viruses infecting grapevines A large number of viruses have been detected in grapevines around the world. Some of these viruses have only been reported on one occasion whereas other are major pests in many grape growing areas. The nepoviruses which cause fanleaf disease are a major group in most countries. The presence of fanleaf was recorded by McKissock (1964) and was widely studied

24 by scientists working at DSIR in Mt Albert during the 1960s and 1970s (Chamberlain 1967, Chamberlain et al. 1970). Although grapevine fanleaf was once very common, the absence of the nematode vector in New Zealand, roguing of old cultivars, and the use of virus indexed rootstocks and scionwood has virtually eliminated this disease in commercial vineyards. Table 1-3. Taxonomy of currently known grapevine viruses (Martelli, 2003) Family
Comoviridae

Genus
Nepovirus Fabavirus

Number of viruses
16 1 1 1 2 2 1 2 7 1 2 2 1

Bromoviridae

Alfamovirus Cucumovirus Ilarvirus

Tombusviridae

Tombusvirus Carmovirus

Closteroviridae

Closterovirus Ampelovirus
Unassigned to a genus

Tymoviridae

Macluravirus Marafivirus

Bunyaviridae Unassigned Genera

Tospovirus

Sobemovirus Necrovirus Potexvirus Foveavirus Tobamovirus Vitivirus Trichovirus Idaeovirus

1 1 1 1 2 4 1 1

25 The number of viruses known to infect grapevines was recorded as 55 during the ICVG meeting in Italy (2003) (Table 1-3). Information about a number of new viruses (including a new leafroll associated virus) were presented at the 15th meeting of the ICVG in South Africa (q.v. p29). Serological detection Identification of the viruses that cause grapevine diseases has proven to be very difficult. The most common reasons are that most diseased grapevines are infected with more than one virus, or virus titre is often very low or most of the viruses can not be transmitted by sap inoculation to uninfected grapevines. Virus extracts prepared from infected plants have been used to produce antibodies in rabbits. When a virus is injected into an animal, the immune system of the animal produces a mixture of antibodies against the viral protein. This mixture is referred to as polyclonal since many different antibodies are produced that react with different chemical groups on the surface of the viral protein. If the original viral extract contained protein from more than one virus then the antisera will often contain antibodies reacting with different viruses. Scientists have sometimes been unaware of the presence of another unknown virus in the original extract and results obtained using the polyclonal antiserum have given misleading results. For example one of the early antisera prepared against GLRaV-2 contained antibodies to both GLRaV-2 and GLRaV-6 (Boscia et al. 1995). Goszcynski (ref) prepared GLRaV-1, GLRaV-2 and GLRaV-3 specific polyclonal antisera from mixed infections through separation of viruses by electrophoresis, blotting to nitrocellulose, isolation of bands and immunisation. This is a successful approach that can be used with mixed infections of other grapevine viruses. Over recent years several groups have prepared monoclonal antibodies in mice. Monoclonal antibodies react to a single chemical group (epitope) and the mouse cell line producing this monoclonal antibody can be preserved and later reused to produce further batches of antiserum with the same reactivity. Monoclonal cell lines can be selected to react with high specificity, or sensitivity or to detect many different strains of a virus and the monoclonal antibodies produced by these cell lines have been very useful in the identification of a number of distinct viruses which are associated with grapevine leafroll disease. Many of these monoclonal antibodies are available from commercial suppliers and are used in laboratories around the world. Since the serological detection of leafroll viruses was first applied on a large scale in the early 1990s, the quality and specificity of the antisera available commercially have improved. An improved extraction medium has also greatly improved the sensitivity of detection for GLRaV-1 (Cohen & van den Brink 2000). For large scale screening, detection methods using an enzyme linked immunosorbent assay (ELISA) can be used to detect very low levels of either GLRaV-1 or GLRaV-3 in composite samples prepared from leaves of bark collected from 5 vines per sample. Virus diagnostic laboratories in New Zealand provide very sensitive and cost effective procedures to screen both rootstock and scionwood propagation material for the presence of GLRaV-1 and GLRaV-3. Molecular detection methods The past 20 years has seen the application of a range of molecular methods to elucidate the biology of viruses and provide rapid, sensitive and specific diagnostic procedures. Techniques developed to determine the sequence of chromosomes in humans and plants have also been applied to sequence the very small genomes of plant viruses. Plant virologists around the world determine sequences to either small parts of a virus or to whole virus genomes and submit these to databases that are readily accessible on public websites. The accessibility of this information is revolutionising plant virology. For diagnostic purposes, sequence

26 information can be used to design primers for use in polymerase chain reaction (PCR) amplification of fragments from particular viruses. The genome of viruses infecting grapevines consists of a single or multiple strands of RNA. Before carrying out a PCR reaction, it is necessary to extract and purify the virus or viral RNA and produce a DNA copy of the viral RNA using reverse transcriptase. The copy DNA (cDNA) is then amplified in the PCR reaction. The use of reverse transcriptase and PCR is referred to as RT-PCR. Although RT-PCR is capable of detecting extremely low levels of virus, and in some circumstances is 10-1000 times more sensitive than ELISA, it is also more difficult to carry out and is more prone to failure than ELISA. Factors that affect reliability are the presence of inhibitors in RNA extracts, the design of the primers used to initiate the reactions and the quality of reagents used. RT-PCR is of greatest value in cases where high quality antibodies are not available. Seasonal movement of GLRaV-3 within the vine There have been several studies of the ability to detect GLRaV-3 in different tissues collected from known infected vines over the course of a season. Teliz et al. (1987) used ELISA to detect GLRaV-3 in Pinot Noir vines from winter dormancy, through inflorescence development and flowering to veraison and fruit harvest. Bark phloem of both dormant and growing canes was a good source of GLRaV-3. Leaf tissue was much less reliable, but these authors do not appear to have sampled sections of the lamina and concentrated on vein tissue. GLRaV-3 was first detected in basal leaves, but, soon after flowering, virus was detectable in the youngest expanded leaf although at a much reduced titre. Virus was also detected in young fruit and tendrils and a low titre was detected in root samples. Matthews (1996) followed the accumulation of GLRaV-3 in the leaves of 5 year old Pinot Noir vines. Samples were taken from the principal veins of the leaves. The results of this study were similar to those of Teliz et al. (1986). GLRaV-3 accumulated in leaves as the season progressed and accumulation was more rapid in basal than in apical leaves and it was observed that GLRaV-3 could be detected 2 months before the onset of symptoms. Ling et al. (2001) compared the sensitivity of ELISA with several RT-PCR protocols for the detection of GLRaV-3 in field samples. They found that bark scrapings were the most reliable sample whether using ELISA or RT-PCR. Since these authors followed the procedure of Teliz et al. (1986) for leaf laminar samples, they did not concentrate on vein tissues. They reported higher levels of virus in petiole samples. Although detection of GLRaV-3 was very good in symptomatic leaves harvested post veraison, asymptomatic leaves were less reliable. Monis & Bestwick (1996) investigated virus titre in both greenhouse grown and tissue cultures grapes infected with GLRaV-1, 2 and 3. They demonstrated that virus titre in the petiole and midrib was much higher than in the leaf blade for both GLRaV-1 and GLRaV-3. This result is expected since both viruses are localised in phloem tissue. Virus titre was also higher in symptomatic and/or basal leaves than in asymptomatic leaves and/or young leaves. High titres of GLRaV-3 were found in all tissue culture grown explants and the titre was higher than the titre in the originally tested internode samples. The picture that emerges from these studies is that a high titre of GLRaV-3 is present in phloem tissue of the canes and trunk. With the onset of growth in the spring, virus moves into the expanding shoot and can be detected in a range of tissues soon after budbreak. Virus titre appears to decline in some of these tissues as leaves and fruit expand, but accumulation of virus in leaves occurs through the season. We have found that leaf vein tissue collected from basal leaves from January to May is almost as reliable as bark scrapings, although virus titre may be lower in the leaf veins.

27 Virus movement and detection in newly infected grapevines In New Zealand it has been found that even when GLRaV-3 infected vines are identified and removed from a vineyard, further infected vines are detected in subsequent years, often adjacent to positions from which infected vines were removed. There are several possible explanations for this, but the most likely seems to be that virus that has been transmitted is difficult to detect in newly infected vines. If vines become infected late in the growing season, infection may not be detected in the following autumn or winter using ELISA, possibly because of uneven virus distribution in newly infected vines. In a project funded by Winegrowers of New Zealand, virus was transmitted to young vines growing in 16-litre pots trained to two canes. The technique used to transmit virus was by grafting small strips of bark from grapevines infected with both GLRaV-1 and GLRaV-3 near the base of one of the two canes in January of 2002 and 2003 (Cohen & van den Brink 2003, 2005). Samples were taken from different parts of the grafted vines during the following winter and spring to determine the presence and spread of GLRaV-1 and GLRaV-3 in the vines, using both ELISA and RT-PCR. Movement of virus in a plant is influenced by the direction of carbohydrate movement in the phloem (Hull 2002). In a grapevine, translocation of sugars will be predominantly to young expanding shoots. As the shoots mature, sugars move from the leaves to developing fruit and then into the canes, trunk and roots. In most of the grapevines tested by Cohen and van den Brink virus titre was high in the nodes below a graft and very low above the graft in autumn and winter. However, for the 2002 grafted vines one Gravesac vine was identified with high GLRaV-3 titre almost to the tip of the grafted cane. In the 2003 vines there were two Cabernet Sauvignon vines and one Riesling vine in which a high titre of GLRaV-3 was measured more than 10 nodes above the graft. Clearly, some new infections can spread acropetally soon after infection. In many deciduous plants, the current seasons phloem cells die in autumn and new phloem differentiates in spring. In grapevine canes, however, phloem from the previous season is reactivated before the cambium begins growth in the spring (Esau 1948). Grapevine leafroll viruses are found only in the phloem and as for most new infections these viruses are mainly spread with the remobilisation of sugars in the roots and trunk in the spring. This accounted for the appearance of virus in the ungrafted canes of most infected vines in early spring following infection. These results have implications for the selection and virus indexing of potential budwood blocks. Where mealybugs are spreading GLRaV-3, most new infections will not be detected during routine indexing in the following autumn or winter. This means that some infected buds could be collected for propagation. This risk will be highest if budwood is collected from blocks with a high level of GLRaV-3 infection and high mealybug numbers. Conversely, the risk of collecting infected budwood will be lower if collection is from blocks in which infected vines are routinely identified and then destroyed, and where mealybug numbers are low. Thermotherapy A programme to eliminate fanleaf and leafroll viruses using thermotherapy was initiated by the DSIR in 1964. By 1970, Over de Linden & Chamberlain reported that 167 vines have been established from 437 shoot tip cuttings taken from 291 heat treated plants of the 18 varieties treated. Pinot Meunier free of these two viruses was released in 1968, a further 6, 7 and 4 varieties were expected to be released in 1971, 1972 and 1973 respectively, with further varieties to follow. In this programme presence of virus was determined by symptoms when grafted to 3 indicators, St George, Mission and LN33 (Over de Linden & Chamberlain 1970b).

28 Work on thermotherapy was continued at DSIR, Mt Albert, and all the major varieties listed in a 1975 survey of New Zealand vineyards had been either heat treated or indexed free of known viruses (Thomas 1976). In the early 1980s a new project was initiated by Dr. Richard Smart, MAF, Ruakura, to determine the virus status of major grapevine varieties in New Zealand and, where necessary, to eliminate these viruses using heat therapy or tissue culture. Material that was later indexed as free of known viruses was to be established in a repository at a block south of Hamilton known as Blands. A total of 7 heat treatment runs were carried out during 1983 and 1984 involving both rootstocks and wine grape varieties. Virus indexing was mainly by grafting on to virus-free indicator varieties, LN33 and Cabernet Franc. Indexed vines from this programme formed the basis of the Blands collection which was a major source of grapevine propagation material through the 1990s. Plants from this collection have been relocated to a block at Rowley Crescent (Marlborough). A new project to eliminate a range of viruses from elite clones selected by the New Zealand Grapevine Improvement Group (NZGVIG) is currently being carried out by HortResearch. Indexing of plants coming from this programme will be indexed initially by ELISA and then using RT-PCR. Grapevine resistance to GLRaV-3 To the best of our knowledge, resistance to GLRaV-3 has not yet been identified in any grape rootstock or wine cultivar (e.g. Lahogue & Boulard 1996). American grape species and hybrids often do not show symptoms, but this is tolerance not resistance. GLRaV-3 cannot be detected in leaves of GLRaV-3 infected grapevine rootstocks using ELISA, although phloem scrapings from canes from the same vines index positive for this virus. GLRaV-3 can be detected in leaf samples by RT-PCR. Many white grape varieties show little if any symptoms in leaves and, although yield may be reduced, grape quality may not appear to be affected. Grape varieties that show few, if any, symptoms when infected with GLRaV-3 pose a special problem in devising control strategies. It is difficult to rogue infected vines in the vineyard on the basis of symptoms and there are no tests that are economical for mass screening unless scion wood is being collected for propagation. Asymptomatic vines are a reservoir of infection that can spread to other vines in the vineyard, and also can spread to neighbouring properties. However, if the aim of NZWG (or NZVIG) is to eradicate GLRaV-3 from New Zealand vineyards, then the presence of the virus in tolerant varieties would be a problem, rather than a solution. It might be possible to specifically breed for virus resistance, either using traditional methods or by genetic modification of existing varieties. Although often successful with other crops, neither of these approaches could give a solution to current vineyard problems in the short term. Breeding GLRaV-3 resistance into grapevines might not solve the problem over the longer term either. Most wine production around the world is based on varieties that have been selected over a long time period. New Zealand wines are aimed to the super-premium category in the international market place and it may be difficult to breed new resistant cultivars that would be acceptable to this market. Depending on the mechanism of the resistance, it would probably also be necessary to graft resistant wine grape varieties on to resistant rootstocks. Some forms of pathogen resistance are more durable than others, but there is always a risk that genetic variants of the pathogen will arise that overcome resistance. This is less of a problem with annual plants because production of new resistant cultivars is an ongoing task. However for perennial crops, particularly where clonal propagation of a cultivar is used resistance breeding is difficult. For example, it took more than 50 years to breed a resistance gene to apple black spot (Venturia inaequalis) into commercially acceptable apple cultivars.

29 By this time resistant strains of the fungus had been identified. Grapevines are expected to be productive for more than 25 years. Over the past 20 years there has been enormous progress in the development of virus resistance using genetic transformation methods. It has been shown that introduction of a small part vial genome into a plant can trigger a process called post-transcriptional gene silencing (PTGS). PTGS is a mechanism whereby a plant (or animal) detects small doublestranded RNA molecules and destroys them. Some plant viruses have developed strategies to inhibit PTGS. Development of successful virus resistance strategies is currently a very active area of research. One example was the successful development of Rainbow papaya, resistant to Papaya ringspot virus (Gonsalves 1998). In a similar project, tamarillos resistant to Tamarillo mosaic virus were produced by HortResearch (Cohen et al. 2000). These virusresistant tamarillos were succesfully tested over a 3-year period in a contained field trial, but the work was not continued because of sustained pressure from GE Free NZ and the enormous regulatory costs required to take this project further. In grapevines, successful genetic transformation methods have been developed and vines resistant to Grapevine fanleaf virus (GFLV) have been produced (Krastanova et al. 1995) and successfully trialled (Fuchs et al. 2000). Vines were infected either by grafting infected wood or by nematode transmission. Since the nematode is present in soil and infects grapevine roots, a transgenic resistant rootstock might provide effective protection to non-transgenic GFLV susceptible scion varieties. Results from a field trial in France showed that some transgenic rootstock gave good levels of protection (Fuchs et al. 2000), but this trial was later stopped because of anti-GE sentiments in France. Since mealybugs and soft scales can transmit GLRaV-3 into grapevine roots, trunk, cordons and leaves, providing a resistant rootstock would not provide sufficient protection to the vine. GLRaV-3 coat protein genes have been transformed into grapevine rootstocks (Xue et al. 1999), but there have not yet been any reports of induced resistance. Genetic transformation is being investigated to induce grapevine resistance to closteroviruses by engineering viral genes or expressing single chain recombinant antibodies to GLRaV-2 and GLRaV-3 (Nolke et al. 2003, Orecchia et al. 2006). Workers at the University of Stellenbosch in Sth Africa have developed techniques to transform embryogenic cell cultures of grape and intend to insert genes that might confer broad spectrum control of grapevine leafroll disease. Descriptions of these projects can be found at the website: http://www.winetech.co.za/project_t25.php3. It must be emphasised that these are long-term studies that are still at a proofofconcept stage. Similar work was conducted at Cornell University in the USA and reached a field trial stage, only for the commercial sponsor to withdraw funding and destroy all plant material. The International Council for the Study of Virus and Virus-like Diseases of the Grapevine (ICVG) Following the completion of this review, the 15th meeting of the ICVG was held in Stellenbosch in April 2006. During the meeting there were, inevitably, a number of important papers that related to grapevine leafroll disease (and other diseases of grapevines). Rather than attempt to interpolate the information into the existing review, the important body of information from this (and earlier meetings) of the ICVG is highlighted here. Extended Abstracts of the 13th - 15th meetings, and bibliographies from earlier meetings can be downloaded from http://www.icvg.ch. Many of the papers are relevant to the New Zealand wine industry, and will continue to be so in the future as additions are made to the website. As an illustration of the relevance of the ICVG conferences to this review, a selection of titles from the April 2006 conference is presented below. Comprehensive summaries of these research papers are freely available in .pdf format from the ICVG website.

30 Selected papers from 15th ICGV Meeting, Stellenbosch, South Africa Author Martelli Angelini Borgo Jelkmann Jooste Rowhani Rowhani Goszczynski Borgo Title Grapevine virology highlights 2004-2005 Characterization of grapevine leafroll-associated virus 2 strain BD Graft incompatibility and leafroll symptoms in grapevines affected bydifferent GLRaV-2 variants Genetic variability of grapevine leafroll-associated virus 1 Differentiation between two distinct molecular variants of GLRaV-3 Molecular characterization of grapevine leafroll-associated viruses 4 and 6 A putative new ampelovirus associated with grapevine leafroll disease Molecular variants of Grapevine virus A (GVA) associated with shiraz disease in south africa The relationship between grapevine rupestris stem pittingassociated virus and rupestris stem pitting and vein necrosis diseases Generic and specific detection of grapevine leafroll associated viruses using ramped annealing nested PCR Use of remote sensing to monitor the spread of grapevine leafroll disease in South Africa Rapid identification of three mealybug species by multiplex PCR Spatio-temporal distribution dynamics of grapevine leafroll disease in western cape vineyards Grapevine leafroll-associated virus 3 - vector interactions: transmission by the mealybugs Planococcus ficus and Pseudococcus longispinus (Hemiptera: Pseudococcidae) Agronomical and enological performances of a marzemino clone beforeand after virus (GLRaV-1 and GVA) elimination Modification in field behaviour and grape quality, with focus on terpenes, after GLRaV-3 eradication in a clone of white muscat (Vitis vinifera l.) Improvement in grapevine chemotherapy Vergelegen, South Africa: a case study of an integrated control strategy to prevent the spread of grapevine leafroll disease Molecular characterization of a divergent strain of grapevine leafroll-associated virus 3 Detection of grapevine leafroll viruses using direct immunoprinting (dip)-ELISA Grapevine leafroll-associated virus 3 transmission efficiency of Planococcus ficus and Pseudococcus longispinus (Hemiptera: Pseudococcidae) Further evidence of the involvement of grapevine leafrollassociated virus 2 in graft incompatibility Partial characterization of two divergent variants of grapevine leafroll-associated virus 4 Planococcus citri (risso) (Homoptera: Pseudococcidae) as GLRaV-3 vector Page 13 22 25 28 32 30 52 72 75

Maliogka Pietersen Saccaggi Pietersen Krger

115 122 124 126 130

Malossini Mannini

132 136

Panattoni Spreeth Angelini Cabaleiro Douglas

139 142 148 180 191

Pirolo Saldarelli Segura

242 246 252

31

SECTION 2 THE IMPACT OF GLRAV-3 ON VINE GROWTH, GRAPE YIELD AND JUICE QUALITY
INTRODUCTION
The economic consequences of GLRaV-3 are potentially severe. Vines infected with GLRaV3 show marked decline in quantitative and qualitative parameters of performance. Ultimately, wine quality is affected. This section focuses on a review of vine growth, grape yield and juice quality. The global literature search revealed the virtual absence of published data from South Africa. GLRaV-3 is known to be a major impediment to the future development of the South African wine industry, yet, despite extensive searching, few links between GLRaV-3 and vine performance were found. To confirm relevant literature was not overlooked Professor Gerhard Pietersen, a plant virologist from the University of Pretoria with research interests in grapevine leafroll disease, was contacted directly. Professor Pietersen confirmed there were no quantifiable data from South Africa on the effect of GLRaV-3 and grapevine yield components. Instead of generating yield loss data, the South African wine sector has opted to invest limited research funds into developing methods for controlling this disease (G. Pietersen pers. comm.). Leafroll virus and vine performance in New Zealand In New Zealand, the existence of grapevine leafroll virus and its effect on vine performance has been documented for many years. Among the first to record the occurrence of leafroll virus in grapevines was McKissock (1964) who noted the presence of visual symptoms in many New Zealand vineyards. By looking to overseas experiences McKissock (1964) highlighted that among infected vines, leafroll adversely affected vine growth, yield, fruit colour and sugar content at harvest. Soon after, Chamberlain (1967) presented preliminary results of research undertaken in Auckland and Hawkes Bay in which the effects of leafroll virus on vine performance were measured. These early observations supported the results of overseas studies and were developed more fully over the next few years (Chamberlain et al. 1970, Over de Linden and Chamberlain 1970a, b, Thomas 1976). Following the work of Over de Linden, Chamberlain and Thomas, there appears to have been no further applied research undertaken in New Zealand into leafroll and its quantitative and qualitative effects on vine performance. Some papers in local publications have discussed a range of issues including the spread of the virus in vineyards (Jordan 1993, Bonfiglioli et al. 2002), qualitative effects of leafroll on Australian vines (Krake 1993) and the results of a survey within the wine industry to elucidate understanding of virus and vector management (Bonfiglioli and Stewart 2005). Further, a report for New Zealand Winegrowers presented and discussed data that modelled the spread of the virus in vineyards as well as providing an economic analysis of GLRaV-3 to the industry (Walker et al. 2004).

32

Effects of GLRaV-3 on vine growth Photosynthesis Over the last decade, a number of authors have investigated the impact of GLRaV-3 on photosynthetic activities and pigments in grapevines. These include net photosynthesis (Pn), transpiration rate (E) and stomatal conductance (gs), and chlorophyll (Chl) and carotenoid (Car) content. In Cabernet Sauvignon plants grown in vitro, Hristov and Abrasheva (2001) found lower Chl and Car content in GLRaV-3 infected plants compared with GLRaV-3 free plants. However, in Albario shoot tips cultured in vitro, Gonzlez et al. (1997) found no difference in Chl or Car content between clean and GLRaV-3 infected plants. The latter result seems to complement results on Albario grown in pots, where no change in Pn, E or gs was found (Cabaleiro et al. 1997). Conversely, in Malvasia grown in pots, plants infected with both GFLV and GLRaV-1,2 and 3 had a reduced Pn of up to 45% depending on leaf age, and reductions in gs (44%), Chl (41%) and Car (29%) were also measured (Sampol, Bota et al. 2003). It is often difficult to transfer results from in vitro and pot trials to field situations as conditions are very different. Although no change in Pn was found in Albario plants grown in pots, both Cabaleiro et al. (1997) and Cabaleiro et al. (1999) measured reductions in Pn of between 53 and 65% (without consistent change in E or gs) when comparing GLRaV-3 infected and uninfected field grown vines. Similar reductions in Pn have been found in work on other varieties infected GLRaV-3 in association with other viruses, and this was often accompanied with reduced E, gs, Chl and Car (Table 2.1). It is difficult to separate out cultivar effects in this data due to the level of variation within cultivars, probably arising from the large differences in environmental conditions and virus associations between experiments. Nonetheless, it is apparent that in all situations, Pn was depressed by at least 25% and in some cases up to 65%. This level of suppression is likely to have direct consequences on all aspects of growth and cropping. Not all vines infected with GLRaV-3 show symptoms of leafroll. Vine age (Sampol, Bota et al. 2003), time of year (Mannini, Argamante et al. 1996, Guidoni, Ferrandino et al. 1997, Mannini et al. 1997), cultivar (Credi & Babini 1997, Kovacs et al. 2001) and virus strain (Woodham et al. 1983) among other factors, may play an important role in symptom development. In both Malvasia (Sampol et al. 2003) and Albario (Cabaleiro et al. 1997, Cabaleiro et al. 1999) Pn in vines showing symptoms of leafroll was reduced by 48 to 59% compared with vines that did not show symptoms but were nevertheless still infected with GLRaV-3. In other words, the appearance of symptoms in infected vines had a major negative effect on their photosynthetic activity. Of more significance however, are results that show that photosynthetic activity is reduced in vines having GLRaV-3 but NOT showing symptoms, as this implies that the effects of lower photosynthetic activity may impact growth before growers are aware of the problem. In Nebbiolo cl 415, Mannini et al. (1996), Guidoni et al. (1997) and Mannini et al. (1997) have shown that in GLRaV-3 infected plants, Pn is depressed early in the season, well before the appearance of leafroll symptoms.

33 Table 2.1. Literature (Source) citing country of research (Country), measured changes in net photosynthesis (Pn), transpiration rate (E) and stomatal conductance (gs); and chlorophyll (Chl) and carotenoid (Car) content in different cultivar rootstock combinations (CV/RS) of field grown grape vines, infected with GLRaV-3 (alone or in combination with other viruses) as compared with those not infected with GLRaV-3. CV/RS Albario Albario / SO4 & 197-17C Lagrein / Kober 5BB LN 33 Nebbiolo 415 / Kober 5BB Nebbiolo 415 / Kober 5BB Nebbiolo A&B / Kober 5BB Country Spain Spain Italy Korea Italy Italy Italy -56% -50% -30 to -60% -20 to -60% Pn -26 to -65% -40 to -50% -65% E +19 to 0% ns -48% gs ns* ns -59% -38% -18 to -40% -22% Chl Car Source (Cabaleiro, et al. 1997) (Cabaleiro, et al. 1999) (Bertamini, et al. 2004) (Kim et al. 2003) (Guidoni, et al. 2000) (Mannini, et al. 1997) (Mannini, et al. 1996)

* ns infers that no significant differences were found

In one field experiment, the use of a foliar nutrient supplement was found to reduce the effects of the leafroll on Pn, and gs (Sampol et al. 2003). If this result is found to apply more widely, it may prove to be an effective method of alleviating the influence of GLRaV-3 commercially. Anatomical Effects Grape leafroll viruses are considered to be phloem limited or phloematic viruses (Guidoni, Mannini et al. 2000, Pietersen 2004). Considerable degeneration of phloem cells in leaves (lamina veins and petiole), stems, and fruit petioles has been reported, and this is usually accompanied by an accumulation of starch in leaves. This phloem degeneration could impede carbohydrate export and result in starch accumulation (Hoefert & Gifford 1967), which could in turn be the feedback mechanism whereby photosynthetic activities are shut down in infected leaves. Roots do not always show signs of phloem degeneration and it is not apparent in shoot apices or mature canes (Hoefert & Gifford 1967). Rootstocks and the root system There is very little published research on the effects of GLRaV-3 on grapevine root systems. In Albario grown in vitro Gonzlez et al. (1995) found that GLRaV-3 infection had no effect on the number of explants that formed roots, or the numbers of roots formed per explant. GLRaV-3 infection did however reduce root length by 43%. In contrast, Hristov et al. (2001) found that GLRaV-3 infected Cabernet Savignon explants produced 17% fewer initial roots than uninfected explants and total length was reduced by 13%. Root mass was also reduced by 18%. In field grown Cabernet Franc vines, grown on their own roots, leafroll infected vines did not show reduced root growth (Clingeleffer & Krake 1992). This result is similar to the findings of Over de Linden and Chamberlain (1970a) in New Zealand with Baco 22A, although these authors found that leafroll infected Mission vines grown on their own roots did produce fewer roots than healthy vines.

34 In commercial vineyards, vines are planted as grafted plants, with particular rootstocks used for specific situations. Published data on root systems in grafted vines appears to be almost non-existent, although Goheen (1988) mentions that affected plants have slightly smaller root systems. Although much of the research on leafroll has been done in commercial vineyards, very little has been done on specific rootstock effects. Credi and Babini (1996) found that growth in three rootstocks (420 A, Kober 5BB and Teleki 5A) was reduced by 43 to 90% when they were infected with virus associations that included GLRaV-3. Differences in growth reduction were due to different virus associations rather than rootstocks. Borgo and Michielini (2000) also found that there was no difference between 420 A and Kober 5BB with respect to the spread of GLRaV-3. Cabaliero et al. (1999) reported that, among other variables, growth of Albario on GLRaV-3 infected SO4 rootstocks was significantly better than on GLRaV-3 infected 196.17C rootstocks, although they do not present supporting data. It is a strong possibility that this response was due to factors other than tolerance to virus infection. Most grape rootstocks, particularly American hybrids, do not show symptoms of leafroll even though they may well carry it (Goheen 1988, Pietersen 2004, Kovacs, Hanami et al. 2001). This can present significant problems, as symptoms of the resulting scion infection, are usually only seen well after vineyard planting is complete. As mentioned earlier, this lack of symptoms does not necessarily imply a genetic tolerance or resistance to the virus (Kovacs, Hanami et al. 2001). Differences in disease response have been measured among different Vitis species and viral isolates (Golino 1993), however resistance to GLRaV-3 has not yet been demonstrated in Vitis vinifera (Pietersen 2004). There is strong evidence that viruses can cause graft incompatibility in grapevines (Golino 1993), but there is very little information on this for GLRaV-3. Lider et al. (1975) found no difference in graft take of healthy or leafroll infected Burger/Dogridge grafts, but Golino et al. (2002) maintain that grape leafroll has been implicated in graft incompatibility. Stem and Cane Growth There are a large number of reports that include results on the effect of GLRaV-3 (both alone and in association with other viruses) on the growth of aerial parts of grape vines. Most of these include some measure of cane growth (usually pruning weight or cane length), but in some instances include measurement of weight of older wood or stem circumference. Most available literature cites a reduction in weight of canes of 0 to 20% as a result of infection due to GLRaV-3 and associated viruses, although at least a third of the literature cites weight reductions in excess of this, some as high as 78% (Table 2.2). Weight of oneyear-old wood may be reduced by 0 to 40% (Table 2.2) and stem circumference may be reduced by 0 to 30% (Clingeleffer & Krake 1992, Cabaleiro & Segura 1996, Cabaleiro, Segura et al. 1999). Only one available paper discussed the effects of this growth reduction on stem and cane morphology. In field grown Cabernet Franc vines that were spur pruned, node number per shoot did not differ with virus infection, although internode length was reduced by 15% (Clingeleffer & Krake 1992). The above figures are cited on a variety of cultivar, rootstock and environmental conditions. There seems to be little consistency in level of growth inhibition within cultivars, but this is probably due to the influence of other factors, most significant of which are probably environment and the specific virus association on which the study was done. Woodham et al. (1983) found that the extent of shoot growth reduction varied 0 to 50% depending on season, which suggests large variation related to the prevailing seasonal environment. There are very few studies that include data on stem and cane growth in the more important varieties

35 currently grown in New Zealand vineyards. In work conducted on Mission and Baco 22A in New Zealand, leafroll was found to reduce vine size in Mission, but not in Baco 22A (Over de Linden & Chamberlain 1970a). As mentioned in the previous section, the absence of symptoms in leafroll infected plants does not imply that the virus is not producing a negative impact on stem and cane growth compared with uninfected plants. However, stem and cane growth in Sultana and Cabernet Franc vines with mild leafroll symptoms may be less affected than in vines with more severe symptoms (Woodham, Krake et al. 1983, Clingeleffer & Krake 1992). Guidoni et al. (1997) and Mannini et al. (1998) suggest that increased vigour in virus free vines may cause increased management demands in cool climate vineyards because of increased requirement for summer pruning and Botrytis control. This may influence management strategies in New Zealand vineyards. Leaf Growth We could only find one paper that included measurement of leaf growth in the results. In Malvasia grown in pots, Sampol et al. (2003) found that infection with a virus association of GLRaV-3 and Fan Leaf Virus caused a 25% reduction in leaf number and 53% reduction in leaf area. These authors also found significant changes in leaf mineral content, notably increases in sodium and phosphorus and decreases in calcium and nitrogen. Phenology We could not find any papers dedicated specifically to phenological changes in grapevines due to GLRaV-3 infection, but some authors do include limited data in their results. In Nebbiolo, budbreak was found to be significantly earlier in GLRaV-3 infected vines (Mannini et al. 1999), although how many days earlier is difficult to establish from their data. Conversely, in Sultana, Mission and Baco 22A, budbreak was found to be delayed in infected vines (by about 1 day in Sultana) (Over de Linden and Chamberlain 1970a, Woodham et al. 1984b). Although GLRaV-3 infection has been linked to early budbreak in Nebbiolo vines, Guidoni et al. (1997) found that veraison in the same cultivar was delayed by 1 week in GLRaV-3 infected vines. In Mission and Baco 22A flowering was delayed by 10 to 14 days (Over de Linden and Chamberlain 1970a). In Burger, leafroll infection delayed fruit maturity by more than one month, although this delay could be overcome by reducing the crop load on infected vines by one half (Lider et al. 1975). Although this rather severe treatment may be impractical from an economic point of view, it indicates that there may well be scope to reduce the effects of GLRaV-3 through vine management. Goheen (1981, 1988) and Pietersen (2000) state that leafroll infection delays berry maturity, and this delay may be by as much as one month (Goheen 1981), but supporting data is not presented. The implications of this delay will vary depending on how the resulting harvest integrates with climate and other vineyard activities.

36 Table 2.2. Literature (Source) citing country of research (Country), measured changes in stem circumference, one-year-old wood and cane growth in different cultivar rootstock combinations (CV/RS) of field grown grape vines, infected with GLRaV-3 (alone or in combination with other viruses) as compared with those not infected with GLRaV-3. CV/RS Albario Albario / SO4 & 197-17C Cabernet Franc Albana cl AL 14T / Kober 5BB, SO4 Treb. Romag. Cl TR 7T/ Kober 5BB, SO5 Nebbiolo 415 / Kober 5BB St Vincent Vidal blanc (Vidal 256) Burger//Dogridge Nebbiolo A&B / Kober 5BB Zinfandel, White Riesling / St George Cabernet Franc cl A.C. 72.8186 Sultana cl H5 Sultana cl H5 Italy Italy Italy USA USA UAS Italy USA Australia Australia Australia Country Spain Spain Australia ns Stem 1-yr wood Circumference weight -29.5% ns Cane growth weight length -33% Source

(Cabaleiro & Segura 1996) (Cabaleiro, et ns al. 1999) (Clingeleffer 0 to 0 to -25% 0 to -9% & Krake 12% 1992) (Credi & 0 to -68% Babini 1997) (Credi & 0 to -78% Babini 1997) (Guidoni et al. -15% 1997) (Kovacs, et al. ns 2001) (Kovacs et al. ns 2001) (Lider et al. -23 to -37% 1975) (Mannini et -15% al. 1996) (Wolpert & ns Vilas 1992) (Woodham et 0 to -50% al. 1983) (Woodham -20 to -40% 1984a) (Woodham ns 1984b)

* ns infers that no significant differences were found

Vine longevity The reviewed literature displays a range of opinions on the effects of leafroll on vine longevity. Goheen (1988) states that leafroll has no effect on vine longevity, but according to Golino et al. (2002), leafroll has been implicated in young vine failure. In the only paper that gives supporting data, Borgo & Michielini (2000) show that GLRaV-3 infection increased the longevity of infected Merlot vines and they suggest that this was due to the lower stress levels on infected plants from lower production capacity. This again suggests that vine management techniques may be useful in overcoming some of the negative effects of the virus.

37 Grape yield During this literature review, 22 studies were identified that presented data on leafroll virus and its effects on yield. Over half (59%) reported a statistically significant decline in the yield of leafroll infected vines compared with healthy vines. Where significant differences were found, the extent of the reduced yield varied considerably between studies. For example, in leafroll infected Burger vines with a full crop, yield over two seasons declined by an average 14% compared with healthy vines with the same crop level (Lider et al. 1975). Similar percentage reductions in yield were observed elsewhere (Hale & Woodham 1979, Schoefling 1980, Woodham et al. 1983, 1984a, McCarthy et al. 1989, Borgo 1991). In contrast, over seven years the cumulative yield in GLRaV-3 infected Albana vines declined by 73% and in a second cultivar (Trebbiano Romagnolo) yield declined by 80% compared with the cumulative yield of healthy vines (Credi & Babini, 1997). Several other studies also found substantial declines in yield, which ranged from 35 72% (Goheen & Cook 1959, Over de Linden & Chamberlain 1970a, Woodham et al. 1984b, Kovacs et al. 2001). In another study, GLRaV-3 lowered fruit development and therefore harvest weight from infected vines compared with healthy vines, but the extent of this difference was unclear (Cabaleiro & Segura 1996). In discussing their results, Over de Linden & Chamberlain (1970a) referred to the research of a German scientist, Scheu, who studied the effects of leafroll on vines grown in Germany. His work, published in 1936, covered six years and included ~10,000 individual weight records from healthy and infected vines. Scheu estimated that in the absence of leafroll virus the German wine sector could conservatively increase production by up to 60%. Over de Linden & Chamberlain (1970a) argued similar productivity gains could also be expected in New Zealand. The results of other studies reported finding no significant yield difference between healthy and leafroll infected vines (Clingeleffer & Krake 1992, Wolpert & Vilas 1992, Guidoni et al. 1997, 2000, Mannini et al. 1994, 1996, 1998, 1999, Cabaleiro et al. 1999). There are potentially many factors likely to have contributed to such an outcome. For example, vine age was suggested as playing a potentially important role in this regard (Cabaleiro et al. 1999). Using young vines (<6 years), these authors suggested that the absence of yield differences in their study may be due to vines initially directing resources to root and shoot development rather than the berries. Similarly, lower net photosynthesis in leaves with symptoms of GLRaV-3 may affect vine yield in the long term. Further, in young vines phloem damage from virus may not be sufficient to significantly delay overall plant development (Cabaleiro et al. 1999). The deleterious effect of leafroll virus on phloem development and function was discussed elsewhere (Hoefert & Gifford, 1967). Grape quality Since 1994 the New Zealand wine sector has experienced rapid growth with the amount of land under cultivation and production having tripled to 18,300 ha and 162,000 tonnes, respectively (HortResearch 2004). Despite this expansion, the quantities of wine that the New Zealand sector can realistically produce are unlikely ever to be more than just a tiny proportion of total world production, so that competing on the basis of volumes alone is unlikely. If New Zealand wines are to succeed on the world stage, it is important that the sector have a point of difference, a marketing edge that distinguishes it from its overseas competitors. Currently, wine quality is that point of difference. Many New Zealand wines are now recognised for their outstanding quality and it is this aspect that has ensured access into and reward within many important overseas markets. Ensuring future success, however, relies on the continued production of a premium product and any adverse influence on wine quality and the enjoyment derived from it will inevitably tarnish brand image.

38 In New Zealand, GLRaV-3 is widely recognised as having the potential to severely disrupt the industrys ability to produce premium wines (Bonfiglioli et al. 2002, Bonfiglioli & Stewart 2005). Factors such as sugar accumulation, titratable acidity, berry anthocyanins and pH are all critical components in producing high quality wine. Largely because of the direction taken by overseas researchers, these are the constituent parts of wine quality that will be discussed in this section. Sugar accumulation ( Brix) Twenty-seven studies were identified from this review that specifically presented data related to leafroll virus and its effect on fruit sugar levels (Table 2.3). In 21 (78%) of these studies, data showed fruit from vines infected with leafroll virus had significantly lower sugar levels than levels recorded in healthy vines. Among those studies where significant differences were found, sugar levels ranged from 0.3 to 5.1 Brix higher in fruit from healthy vines compared with fruit from vines with leafroll virus. Schoefling (1980) and Ueno et al. (1985) found that sugar levels in leafroll infected vines were lower than those in healthy vines but the exact nature of the impact was unclear (Table 2.3). Despite a number of studies finding no significant difference in the level of fruit sugars between healthy and infected vines, the results generally favoured healthy vines, where fruit sugar levels were often marginally higher (Woodham et al. 1984b, McCarthy et al. 1989, Mannini et al. 1994, Credi & Babini 1997, Guidoni et al. 1997). Several authors suggested that a non significant result may be linked to a difference found in yields between healthy and leafroll infected vines. In California, it was found that when 50% of the crop load was physically thinned from leafroll infected vines, total soluble solids at harvest did not differ from that in healthy vines with a full crop (Lider et al. 1975, Kliewer & Lider 1976). In Italy, the cumulative yield of one cultivar (Albana) with several virus and virus-like infections, which included GLRaV-3, was 73% lower than that from healthy vines (Credi & Babini 1997). These authors argued that the significantly higher fruit sugar levels found in the leafroll infected vines than in the healthy vines was related to the very low crop yield in the infected vines. Based on the results of a number of these studies, it was argued that where present leafroll virus reduced the capacity for vines to accumulate sugars in the fruit, thereby delaying fruit maturity and the development of full fruit colour (Goheen & Cook, 1959, Over de Linden & Chamberlain 1970a, Lider et al. 1975, Kliewer & Lider 1976, Woodham et al. 1983, Krake 1993, Cabaleiro & Segura 1996, Cabaleiro et al. 1999). Of course in cool climate conditions like those found in New Zealand, any factor delaying fruit maturity may have severe financial consequences. In one European study, it was noted that market sensitivity to reduced sugar content in leafroll-infected Albario grapes had the potential to severely reduce growers incomes (Cabaleiro et al. 1999, Walker et al. 2004).

39 Table 2.3. Studies identified from around the world where data on fruit sugar levels (Brix) and Titratable Acidity (TA) were assessed between healthy and leafroll infected grapevines. Unless otherwise stated, all positive (+) Brix and negative (-) TA results favoured healthy v. infected vines. Results were statistically significant to at least 0.05 level of significance. Sugar levels found in healthy compared with infected vines ( Brix) Mean +1.1 Mean +1.1 Mean +1.1 Titratable acidity in healthy compared with infected vines (g/L) Mean -0.5 Mean -1.1 ns Lead author and publication date Borgo 1991

Country of study

Length of study (years) 3

Grape variety *

Italy

Cabernet Franc Cabernet Sauvignon Merlot

Italy Italy

2 2

Merlot Merlot Carmenre Cabernet Sauvignon

Mean +1.6 Mean +1.7 Mean +0.9 Mean +0.7 Mean +1.2 Mean +1.0 Mean +0.7 ns ns +1.5 ns Mean +0.5 No data

Mean -1.3 Mean -1.3 Mean -0.7 ns ns Mean -1.0 ns Mean +0.8 ns No data ns Mean -0.4 ns

Borgo et al. 2002 Borgo et al. 2003

Spain Spain Australia Italy

3 3 2 6

Albario Albario Cabernet Franc Albana Trebbiano Romagnolo

Cabaleiro et al. 1996 Cabaleiro et al. 1999 Clingeleffer et al. 1992 Credi & Babini 1997 Goheen & Cook 1959 Guidoni et al. 1997 Guidoni et al. 2000 Hale & Woodham 1979 Kim et al. 2003

USA Italy Italy Australia

1 3 4 1

Zinfandel Nebbiolo Nebbiolo Sultana

Korea

Kyoho

Mean +3.9

Mean -0.07%

40 Sugar levels found in healthy compared with infected vines ( Brix) +3.8 & +3.5 Titratable acidity in healthy compared with infected vines (g/L)

Country of study

Length of study (years) 1

Grape variety *

Lead author and publication date

USA

Burger

-0.1 & -0.08 Kliewer et al. g 1976 tartaric/100ml -0.6 -0.8 No data Kovacs et al. 2001 Krake 1993

USA Australia

1 1

Vidal blanc St. Vincent Emperor

+1.0 +0.6 + (linked to leafroll variant) Mean +3.1 & +3.4 ns ns Mean +0.3 Mean +0.5 Mean +0.4 Mean +5.1 Mean +1.6

USA

Burger

Mean -0.1 & 0.08 g tartaric/ 100ml -0.8 (in 1988) ns ns Mean -0.4 Mean -4.0 meq/ L ns ns

Lider et al. 1975 McCarthy et al. 1989 Mannini et al. 1994 Mannini et al. 1996 Mannini et al. 1998 Mannini et al. 1999 Over de Linden & Chamberlain 1970 Schoefling 1980 Ueno et al. 1985 Wolpert & Vilas 1992 Woodham & Krake 1983

Australia Italy Italy Italy Italy New Zealand

3 3 4 6 7 2 4

Muscadelle Nebbiolo Nebbiolo Nebbiolo Nebbiolo Mission Baco 22A

West Germany Japan USA

1 ? 3

Riesling Zenkoji & Koshu White Riesling Zinfandel Cabernet Franc

+ (?) + (?) Mean +1.2 +1.7 1986) Mean +0.6

-0.9 -1.0 -0.5 (in 1987) (in -0.6 (in 1986) Mean -0.2

Australia

41 Sugar levels found in healthy compared with infected vines ( Brix) ns Titratable acidity in healthy compared with infected vines (g/L) ns

Country of study

Length of study (years) 5

Grape variety *

Lead author and publication date Woodham et al. 1984a

Australia

Thompson seedless Sultana H5

Australia

Sultana clones ns

ns

Woodham et al. 1984b

*Column excludes variables such as rootstocks and/or clones Unit is g/L, unless otherwise stated Means weighted by yield levels ns; no significant difference found between healthy and leafroll-infected vines Difference between virus infected and healthy vines where crop was not thinned Difference between virus infected and healthy vines where crop was thinned

Titratable acidity In an attempt to quantify the impact of leafroll virus on levels of titratable acidity, data from twenty-six studies were reviewed (Table 2.3). In seventeen (65%) of these studies, grape juice from vines infected with leafroll virus was found to have significantly higher levels of titratable acidity (range +0.2 to 1.3 g/L) than was recorded from healthy vines. In contrast, results from a number of other studies revealed that levels of total acidity did not differ between leafroll infected and healthy vines. However, in several of these studies titratable acidity was still marginally higher among infected vines compared with levels recorded from healthy vines (Hale & Woodham 1979, Borgo et al. 1991, 2003, Clingeleffer & Krake 1992, Cabaleiro & Segura 1996, Guidoni et al. 1997). Based on the results from the majority of these studies, it was apparent that levels of total acidity were elevated when leafroll virus was present. The results of one study in particular are notable. In Italy, titratable acidity among healthy vines was significantly higher than either of two treatments containing a mix of virus and virus-like diseases, both of which included GLRaV-3 (Credi & Babini 1997) (Table 2.3). As discussed earlier, these data were thought to relate to the very low crop yield in the infected vines compared with the healthy vines (Credi & Babini, 1997). Generally, data on levels of malic acid and tartaric acid were included in the studies where titratable acidity was assessed; combined these acids represent over 90% of total titratable acidity (Mullins et al. 1992). Where total acidity was adversely affected in leafroll-infected vines, there were only two studies found where both malic and tartaric acids were similarly affected (Kliewer & Lider 1976, Hale & Woodham 1979). In other words, in most studies levels of one or other of malic acid and tartaric acid found in leafroll infected vines did not differ significantly from levels recorded in healthy vines. Anthocyanins During the ripening of the berries the accumulation of sugars and anthocyanins are closely associated (Mullins et al. 1992). Given that sugar accumulation has previously been shown to be adversely affected in leafroll infected vines, it is perhaps no surprise then that in those studies where berry anthocyanin levels were directly measured adverse effects were also

42 observed. Relevant data were presented in seven papers and without exception berry anthocyanin levels among several varieties were significantly reduced in leafroll infected vines compared with levels found in healthy vines (Mannini et al. 1994, 1996, 1999, Guidoni et al. 1997, Borgo et al. 2002, 2003, Kim et al. 2003). In a further study undertaken in Australia, combinations of unnamed leafroll variants from a range of cultivars were monitored and compared with healthy Emperor fruit (Krake 1993). This study differed from the results outlined above in that specific data related to anthocyanin levels from healthy and infected fruit were not presented. Instead, relying on visual assessments, Krake (1993) concluded that all leafroll variants severely reduced fruit colour compared with the fruit from healthy Emperor vines. The combined results of the above research found that berry anthocyanin levels accumulated at a faster rate in healthy vines than they did in leafroll infected vines. Indeed, the rate at which differences were detectable between healthy and infected vines was found to be evident as early as veraison (Guidoni et al. 1997, Mannini et al. 1999). Moreover, Guidoni et al. (1997) noted that in infected vines in Italy the berry anthocyanin levels never reached the quantity found in healthy vines; this result was despite the last sample from infected vines being collected one week after the last assessment from healthy vines. As with many factors associated with leafroll virus, the underlying reasons for the symptoms observed often remain obscure. For example, it remains unclear if leafroll virus slows down the movement of anthocyanins from leaves to berries, thereby causing their accumulation in the foliage, or if the virus interferes with anthocyanin synthesis in the berry (Guidoni et al. 1997). One possible explanation for reduced berry anthocyanin levels in leafroll infected vines could be related to modifications induced by the virus upon enzyme activity (Guidoni et al. 1997). Whatever the causal agents for reduced berry anthocyanin levels, any detrimental effects are likely to be of great importance to the quality of red and full bodied wines (Mannini et al. 1994, 1996). Hydrogen-ion concentration (pH) Eighteen studies directly measured juice pH. Eight reported finding significant differences between healthy and leafroll infected vines (Woodham et al. 1983, McCarthy et al. 1989, Credi & Babini 1997, Mannini et al. 1999, Cabaleiro et al. 1999, Guidoni et al. 2000, Kovacs et al. 2001, Borgo & Angelini 2002). Although there were several important variables between studies (e.g. age of vines, grape variety, rootstock, location and duration of study), all found juice pH was lower in leafroll infected fruit compared with the fruit of healthy vines. By contrast, the results of a further nine studies concluded that irrespective of virus status there was no significant difference in juice pH (Hale & Woodham 1979, Schoefling 1980, Woodham et al. 1984a, Clingeleffer & Krake 1992, Wolpert & Vilas 1992, Cabaleiro & Segura 1996, Mannini et al. 1996, Guidoni et al. 1997, Mannini et al. 1998). In an additional study (a Japanese language publication) pH was lower in leafroll-infected vines than in healthy vines, although it could not be determined if this difference was significant (Ueno et al. 1985). Irrespective of the level of statistical significance found, some interesting observations were recorded in several studies. In Italy, two clones of the cultivar Nebbiolo were assessed over four years (Mannini et al. 1996). Fruit quality indices, including pH, were assessed in leafrollinfected mother plants and then compared with in vitro cultures from these vines that had been subjected to heat therapy in an attempt to eliminate GLRaV-1 and 3. While no difference in pH was found between virused and heat treated vines, pH levels did differ between clones. In Nebbiolo clone B, where GLRaV-1 & 3 was present, pH levels differed significantly from clone A where GLRaV-3 was successfully eliminated (Mannini et al. 1996).

43 Two additional studies highlighted a further potentially important factor, that is, the presence of multiple grapevine viruses. In a study undertaken in the United States, Vidal Blanc vines were either healthy, infected with GLRaV-3 alone or infected with a combination of GLRaV3 and grapevine fleck virus (GFkV) (Kovacs et al. 2001). No difference in pH was observed between healthy vines (mean 3.10) and vines infected with only GLRaV-3 (3.08). However, in vines infected with GLRaV-3 and GFkV a significant reduction in the juice pH level (3.05) was observed (Kovacs et al. 2001). In Italy, the influence of identical multiple virus and virus-like diseases was shown to vary according to cultivar Albana or Trebbiano Romagnolo (Credi & Babini 1997). Juice pH from healthy vines of both cultivars was compared with that in vines infected with GLRaV-3 and grapevine fanleaf virus (GFLV) or in vines infected with GLRaV-3, Rugose Wood (Kober stem grooving and Rupestris stem pitting) and vein necrosis. Over six years the pH weighted mean (based on yield levels) in healthy Albana vines was higher than either of the infected treatments. In contrast, in Trebbiano Romagnolo no significant differences in pH levels were observed between healthy vines and either of the infected treatments (Credi & Babini 1997). About 50% of the reviewed studies found a significant difference in pH levels between healthy and leafroll-infected vines. The remaining studies concluded there was no effect of leafroll virus on fruit pH levels. Clearly though, the results from several studies suggest that pH levels may be related to (but not necessarily limited to) clonal differences or to the extent of virus infection. Nonetheless, on the basis of the evidence from these studies there remains some uncertainty about the extent to which leafroll virus affects fruit pH. Wine quality Only a small number of studies assessed the effects of leafroll virus on the end product wine. Based on spectrographic analysis (Over de Linden & Chamberlain 1970a) and sensory evaluations (Schoefling 1980, Ueno et al. 1985, Mannini et al. 1998), all studies found that the wine quality from leafroll infected vines was reduced compared with that from healthy vines. Mannini et al. (1998) argued that wine produced from healthy Nebbiolo vines was found to have a more complex bouquet and flavour as well as better colour intensity compared with wine from vines with GLRaV-3 and grapevine vitivirus A.

44 Table 2.4. Summary of the grape variety, cultivar, clone or rootstock examined during studies of the effects of leafroll virus on vine performance. * denotes studies not written in English. Grape variety(ies)/cultivar/clone/rootstock Chardonnay Lagrein Cabernet Franc, Cabernet Sauvignon, Merlot grafted on rootstocks Kober 5BB and 420 A Carmenre, Cabernet Sauvignon, Merlot grafted on rootstocks 420A and Kober 5BB Merlot Carmenre, Cabernet Sauvignon, Merlot Albario Albario Sultana clones H4 and H5 Rootstocks 420A, Kober 5BB and Teleki 5A Albana (clone AL 14T) and Trebbiano Romagnolo (clone TR 7T) on rootstocks Kober 5BB and S04 Zinfandel, French Colombard, Burger, Melon, Cabernet Sauvignon, Pinot St. George, all on St. George rootstock Albario Albario Nebbiolo clone 415 Grignolino and Nebbiolo Sultana (clone H5) Cabernet Sauvignon Kyoho Burger on Dogridge rootstock Vidal Blanc and St. Vincent Cabernet Franc, Mission, Sultana clone H5, Baco Blanc 22A & LN33 Burger grafted onto rootstock cv. Dogridge Muscadelle clones Nebbiolo Nebbiolo Nebbiolo clone 415, Grignolino clone 113, Nebbiolo Michet; rootstock Kober Country of study Italy Italy Italy Reference Bertamini et al. 2002 Bertamini et al. 2004 *Borgo 1991

Italy

*Borgo & Michielini 2000

Italy Italy Spain Spain Australia Italy Italy

*Borgo & Angelini 2002 *Borgo et al. 2003 *Cabaleiro & Segura 1996 Cabaleiro et al. 1999 Clingeleffer & Krake 1992 Credi & Babini 1996 Credi & Babini 1997

USA

Goheen & Cook 1959

Spain Spain Italy Italy Australia Bulgaria South Korea USA USA Australia USA Australia Italy Italy Italy

Gonzalez et al. 1995 Gonzalez et al. 1997 Guidoni et al, 1997 Guidoni et al. 2000 Hale & Woodham 1979 *Hristov & Abrasheva 2001 *Kim et al. 2003 Kliewer & Lider 1976 Kovacs et al. 2001 Krake 1993 Lider et al. 1975 McCarthy et al. 1989 Mannini et al. 1994 Mannini et al. 1996 Mannini et al. 1998

45

Grape variety(ies)/cultivar/clone/rootstock 5BB, S04 Dolcetto and Nebbiolo on rootstock Kober 5BB and S04 Nebbiolo clones 63, 66, 308 & 415 Mission and Baco 22A Malvasia on rootstock Paulsen 1103 Riesling clones Mission Zenkoji and Koshu Napolean clones 29-228, 39-29, 74-16 & 77-266 Riesling and Zinfandel Cabernet Franc clone AC 72.8186 Sultana clones H4 or H5 Sultana clone H5 Sultana selections

Country of study Italy Italy New Zealand Spain W. Germany Israel Japan Spain USA Australia Australia Australia

Reference

Mannini et al. 1999 Mannini et al. 2000 Over de Linden & Chamberlain 1970a, b Sampol et al. 2003 Schoefling 1980 Tanne et al. 1973 *Ueno et al. 1985 Valero et al. 2003 Wolpert & Vilas 1992 Woodham & Krake 1983 Woodham et al. 1984a Woodham et al. 1984b

46

47

SECTION 3: GLRAV-3 VECTORS/TRANSMISSION ECOLOGY.

INTRODUCTION
Mealybugs and soft scale insects are now the recognised vectors of viruses such as GLRaV-3 placed in Ampelovirus (Section 1). This recognition has come during an active period of research around the world when knowledge of both plant viruses and the ability of mealybugs (and other insects) to transmit them is being accumulated. In this section we provide a global context into which the species of mealybugs in New Zealand, and their transmission of GLRaV-3, can be placed. It begins with a brief examination of mealybugs themselves - where they are found around the world and some basic information on their biology relevant to disease transmission. This is followed by an overview of mealybug transmission of plant viruses starting with viruses of plants other than grapes, moving to grape viruses other than GLRaV-3, and concluding specifically with GLRaV-3. A short history of mealybugs on grapes in New Zealand The first Registrar of the University of New Zealand, in 1875, was William Maskell. He died in 1898, just four years before Romeo Bragato returned to New Zealand in 1902 as chief viticulturist of the Department of Agriculture. Had Maskell lived a few years longer, the two men would undoubtedly have discussed the control of mealybugs, and perhaps even kickstarted a wider research programme. In his spare time Maskell was a keen amateur microscopist, and, by the closing quarter of the 19th century, had become one of the worlds leading coccidologists. As a result, the taxonomy of mealybugs, and their early economic impacts in Victorian New Zealand are particularly well documented in both the scientific literature and newspapers of the day. By the time Bragato arrived, mealybugs were already recognised as serious pests of grapevines, as Broun (1896) had described the long-tailed mealybug (Pseudococcus longispinus) as a terrible nuisance in some northern vineries. It was Brouns recommendations that led to one of the earliest biological control programmes against an insect pest in New Zealand, with the introduction of the mealybug destroyer (the ladybird Cryptolaemus montrouzieri) in 1897 (Charles, 1989). Pioneer wine-makers in Auckland were also well aware of mealybug pests, and, in the days when vineyards were largely family concerns, cultural control was common. Cheap labour meant that stripping loose bark from dormant vines by hand to remove overwintering mealybugs was economically viable. This practice declined as labour costs increased, vineyards became bigger and pesticides became more reliable and readily available, but was still being carried out to some extent (especially by children) until the second world war (WW2) (J Corban, P Fredatovitch pers. comm.). Some early studies on the biology of mealybug pests of fruit crops were carried out in the 1930s, but were hampered by taxonomic confusion. After WW2, mealybugs were early targets for control by organochlorine insecticides, but resistance to DDT and then organophosphate insecticides was documented from the 1960s (Charles et al. 1993), and the potential for pesticide failure was largely behind a resumption of ecological studies in vineyards in 1978 (Charles 1981) For almost all this period, mealybugs were considered pests of fruit because of the sootymould fungi that grow on the honeydew they excrete. In vineyards it was thought that high numbers of mealybugs might debilitate the vines by reducing photosynthesis or removing sap, while large quantities of sootymould on the fruit might affect fermentation and/or taint the wine.

48 When plant pathogenic viruses were first identified, they were thought to be transferred between plants only by grafting, but insects, especially phloem feeding aphids, were soon implicated in their transmission. Mealybugs are also phloem feeders, and by the 1970s, Pseudococcus longispinus had been identified as a potential vector of cacao swollen shoot virus and alomae lethal disease of taro (Brunt 1970). At this time the mode of transmission of grapevine leafroll disease was considered to be aerial (W. Thomas, pers. comm.), and the possibility that it could be vectored by mealybugs in vineyards was first raised in the late 1970s (Charles 1979). Subsequent initiatives by government scientists to measure the rate of spread of grapevine leafroll disease in vineyards (e.g. Jordan 1993) were followed by experiments which showed that all three common species of mealybugs in New Zealand vineyards could transmit the disease between grapevines (Petersen and Charles 1997, Charles et al. 1997). Hence the association of mealybugs with GLRaV-3 in New Zealand was demonstrated well before the current taxonomic status of the causative ampelovirus and serotypes was recognised (see section 1). Pseudococcidae the mealybugs Mealybugs are terrestrial, plant feeding insects in two families (Pseudococcidae and Putoidae) within the superfamily Coccoidea of the order Hemiptera. There are 16-20 families in the Coccoidea - loosely termed scale insects - and the mealybugs consist of approximately 2000 species in 288 genera around the world (Ben Dov 1994). They are closely related to the armoured scale insects (Diaspididae) and the soft scale insects (Coccidae), with which they share many taxonomic and biological features; and slightly more distantly related to whiteflies and aphids, two other groups that vector plant viruses. A helpful website providing detailed information on the background and systematics of mealybugs can be found at: http://198.77.169.79/scalenet/scalenet.htm . The mealybugs are widely distributed in all of the zoogeographical regions of the world, especially in the tropics and sub-tropics at low or median elevations. They are of great economic importance in the tropics, sub-tropics, and in cooler climates especially in modified environments such as glasshouses. Mealybugs are present in all the geographical regions where grapes are grown (Table 3.1). Many articles on mealybugs and other scale insects are written in languages other than English, especially those produced for local industry and pest management journals (Table 3.2). Pest status of mealybugs in New Zealand There are no Putoidae in New Zealand, but at least 114 species of Pseudococcidae are known, of which about 16 are exotic and these range from inconsequential to major pests. The pests are broadly restricted to either bulbs or roots in the soil, or leaves, fruit and branches of shrubs and trees; the first arrived in New Zealand with early European immigrants, and there has been a steady trickle of additions ever since (Table 3.3). Nevertheless, New Zealand remains free of many species that are serious pests elsewhere, and which could survive in our environment and climate. Unfortunately, increasing world trade and human transport continues to threaten the status quo of mealybug species distribution. Mealybugs are known to be very invasive and the potential for new mealybug pests in New Zealand is high. Planococcus citri is a cosmopolitan pest that has been intercepted in New Zealand many times at ports of entry. Fortunately it has not established. The hibiscus mealybug (Maconellicoccus hirsutus) arrived accidentally in Grenada in the 1990s and rapidly extended its range throughout the Caribbean region, Southern California, Mexico, and Central America (Table 3.7). Phenacoccus parvus has spread through the Pacific Islands and

49 northern Australia in recent years. Some species may be too tropical to survive in New Zealand, but there are many other potential invaders. Table 3.1. Examples of known numbers of mealybug species in the main grape growing regions of the world. (See: http://198.77.169.79/scalenet/scalenet.htm) Country/Region No. of known Country/Region No. of known mealybug mealybug species species S. Hemisphere 116 Australia 202 9 Chile 18 97 New Zealand 115 12 27 N. America USA 351

W. Europe France Greece Italy Portugal Spain

Table 3.2. European language common names for different scale insect groups. English Italian French Spanish German Pseudococcidae mealybug cotonelli, or cocciniglie cotonose cochenilles farineuses, or pseudococcines chanchitos blancos Schmierluse, or Wolluse Diaspididae armoured scale (insect) cocciniglia cochenilles diaspines cochinilla Deckelschildluse Coccidae soft scale (insect) coccidi cochenilles coccines coccidos Schildlaus, Schuppe

or

In the early 1990s, more than 1300 mealybugs (and their natural enemies) were collected from six crops (apples, pears, nashi, citrus, persimmon and grapes) during a survey of 91 sites across the North Island (Charles 1993). The most widespread and economically important mealybugs were three species of the genus Pseudococcus. They were long-tailed mealybug Pseudococcus longispinus, citrophilus mealybug P. calceolariae, and obscure mealybug P. viburni. Pseudococcus longispinus and P. calceolariae were the most common species in all crops, except in pipfruit in Hawkes Bay, which was inhabited almost exclusively by P. viburni. These three species accounted for more than 99% of all mealybugs collected and are therefore regarded as the principal mealybug pests in New Zealand (Charles 1993). Pest status may alter with pest management systems and/or other ecological changes (Franco, Suma et al. 2003), and any of these species, or others, retain the potential to become pests on wine grapes.

50 Table 3.3. Dates of arrival of exotic mealybug species in New Zealand. Species Usually live on foliage Pseudococcus calceolariae (=similans) Nipaecoccus aurilanatus Pseudococcus longispinus Pseudococcus viburni Laminococcus flandersi Phenacoccus graminicola Chorizococcus oreophilus Eurycoccus antiscius Usually live on roots or bulbs Spilococcus mamillariae Rhizoecus graminis Vryburgia amaryllidis Rhizoecus californicus Rhizoecus falcifer Rhizoecus dianthi Rhizoecus cacticans Date first found 1878 1890 1890 1922 1950 1962 1965 1965 Reference (Cox 1987), (Charles, Froud et al. 2000) (Cox 1987) (Cox 1987)) (Cox 1987) NZAC* (Cox 1987) NZAC NZAC

1933 1941 1949 1956 1965 1976 1993

(Cox 1987) (NZAC) (Cox 1978) (de Boer 1967), (Cox 1987) (Cox 1978) (de Boer 1967) (Cox 1978) NZAC

*NZAC = New Zealand Arthropod Collection (Landcare Research).

Pseudococcus longispinus and P. calceolariae arrived in New Zealand from Australia in the early years of European colonisation, while P. viburni was first recorded in the 1920s (Table 3.3). They have had a chequered taxonomic history in New Zealand, and have been known by a number of different Latin and common names in the literature (Table 3.4). Some mealybugs show considerable phenotypic plasticity (i.e. they develop different morphological characteristics in response to different environments), and this has led to some taxonomic revisions (e.g., P. similans has been shown to be the same species as P. calceolariae (Charles et al. 2000). All three species are very polyphagous, and each one attacks dozens of species of a wide range of host plants. P. longispinus, P. calceolariae and P. viburni occur from the North of the North Island south to Nelson (P. longispinus and P. calceolariae) and mid Canterbury (P. viburni) (Cox 1987). On fruit crops, there appear to be regional or host influences on the species composition in orchards. For example, P. longispinus is the dominant mealybug in Auckland (Cox 1977a, Charles 1981), whereas P. viburni is most common in apples and other crops in Hawkes Bay (Charles 1982, Charles & Walker 1981, Ward 1966). In general, P. longispinus is most common in a sub-tropical climate, characterised by relatively high humidity, so tends to dominate in crops in Northland, Auckland, the Waikato and into Gisborne. Pseudococcus calceolariae is often more common across the central to southern parts of the North Island and the north of the South Island, while P. viburni may be best adapted to cooler climates. Pseudococcus longispinus and P. calceolariae may be especially numerous in together in vineyards in the Gisborne region. Pseudococcus maritimus and P. comstocki do not occur in New Zealand, and early references to these in the literature have been shown to refer to either P. calceolariae or P. viburni (Cox 1977b).

51 Table 3.4. The most usual taxonomic names recorded for three pest mealybugs of fruit crops in New Zealand. Present species name P. longispinus P. calceolariae Past synonymies in the Past common names NZ literature Dactylopius adonidum, Long-tailed mealybug P. adonidum P. fragilis Citrophilus mealybug P. gahani P. citrophilus P. obscurus Obscure mealybug, P. affinis Bakers mealybug P. viburni

P. viburni

Mealybugs and grapevines around the world Twenty six species of mealybugs from 16 genera have been recorded from grapes (Table 3.5), which is, perhaps, surprisingly few given the known number of mealybugs in these regions (Table 3.1). Most of the 26 species are polyphagous, and many are recognised pests. Some have a very different biology/ecology from those present in New Zealand, and some, should they ever arrive in New Zealand, are potentially more severe pests than those mealybugs already present. Some species are known vectors of viruses of grapes and other plants, and some cause plants (including grapes) to react physiologically by producing shoot deformations, galls etc. The presence of some species on Vitis may be more or less accidental, but, as the biologies of at least half of the species in Table 3.5 are hardly known, it may be that lack of recognition as a pest may be as much through ignorance as anything else. Based on literature records, the most important pest species of grapes globally are Planococcus ficus, Planococcus citri, Pseudococcus calceolariae, Pseudococcus longispinus and Pseudococcus viburni. However, as more and more species are shown to be vectors of plant viruses, a species pest status might change overnight, and the New Zealand industry must be aware of the potential threats from many new mealybug species. The geographical regions of origin of Vitis spp. are Eurasia and North America. It is reasonable to expect, through co-evolution, that the mealybug species native to those regions (c. 500 spp, Table 3.1) that are capable of exploiting the crop will already have done so. Although the time scale is shorter, a similar situation probably exists in those parts of the world where, through European colonisation, winegrapes have been established for several hundred years (e.g. S. Africa, S. America), as well as newer regions such as Australia and New Zealand (where more is known of the natural mealybug fauna in agroecosystems). Hence, although there are still many species of mealybug to be discovered around the world, brand new mealybug pests of grapes are perhaps rather unlikely to be found (with the possible exception of Asia, where grape plantings are expanding). It is, however, worth noting that some very efficient virus vectors in grapes (e.g. Ps. longispinus, and Ps. calceolariae, which are native to Australia) are new associations with grapes. This raises the scenario that the greatest threats from virus transmission in grapevines may come from completely new pests rather than existing ones.

52 Table 3.5 Species of mealybugs collected from Vitis spp. Recognised Known virus Known plant grape pest vector reaction (* on grapevine) X X X X X X * X

Genus Chorizococcus Coccus Dysmicoccus Ferrisia Formicococcus Geococcus Heliococcus Maconellicoccus Nipaecoccus Paraputo Peliococcus Phenacoccus Planococcus

Species shaferi microogenes brevipes virgata robustus coffeae bohemicus hirsutus nipae viridis leveri kimmericus turanicus aceris madeirensis lilacinus citri ficus minor calceolariae cryptus maritimus viburni jackbeardsleyi longispinus iceryoides vitis falcifer acropygae

X X *

* * * * X * X X

Pseudococcus

Rastrococcus Rhizoecus Xenococcus

Pest status of mealybugs on grapes in New Zealand Different mealybug species can damage different crops by depleting sap, by injecting toxins, and by transmitting plant virus diseases. In addition, and unlike armoured scale insects, mealybugs eject copious quantities of honeydew. This provides an excellent substrate for growth of a number of black sootymould fungi which, in turn, contaminate leaves and fruit and reduce photosynthesis or down-grade fruit quality. Mealybugs may also be significant quarantine pests on fresh produce exported between countries, but the conversion of fresh fruit to wine limits their significance as a grape pest to the vineyard. Mealybugs are probably present in every North Island vineyard that is old enough to produce grapes. P. longispinus is numerically dominant north of Gisborne, while P. calceolariae

53 usually takes over south of Gisborne. In the Gisborne region itself both species are very common. The relative numbers of each may vary from year to year (depending on prevailing weather), in different vineyards (depending on micro-environmental differences), within a single vineyard, or even within a single row of grapes. Both species occur in patches in Nelson and Marlborough, where P. calceolariae may be more common, as it is in midCanterbury. The only region where any vineyard may realistically be considered free from mealybug is Central Otago, but other GLRaV-3 vectors in the Coccidae may be more likely. There are similar regional/climatic differences in Australian vineyards, where P. longispinus and P. calceolariae cause damage in southern Australia, while P. viburni dominates in Queensland and warmer inland areas (Furness and Charles 1994). Developmental biology and life history characteristics of mealybugs Mealybugs are small, soft-bodied, plant feeding insects covered in a water-repelling, powdery wax (from which comes the term mealy in mealybug). Despite their small size and apparent fragility they are remarkably persistent and difficult to kill, and many have become successful pests. Some knowledge of their basic biology is required to understand both this success and how to control them. Mealybugs are hemimetabolous i.e., the immature stages resemble the adult female in general form, passing from first to second to third nymphal instars before finally moulting to adults. When young female mealybugs moult they appear simply to increase in size until maturity, and they do not pass through a pupal stage of structural re-organisation. Young male mealybugs are indistinguishable from females, but, at the end of the second instar, they typically spin a waxen cocoon in which they pass through a complete metamorphosis (including a fourth instar), emerging as small, fragile, two-winged adults. They have no functional mouthparts and cannot feed. They live for only a few days, during which they respond to female sex pheromone and mate. Most species of mealybug are obligate bisexuals, but a few including some pests are known from females only, which reproduce parthenogenetically. Mealybugs feed on phloem sap. Their thread-like proboscis is often longer than their body, and they insert this into the phloem stream of their host plant. They have remarkable control of the proboscis, and manipulate it in between plant cells on the way to the phloem. Once in place, mealybugs tend to remain at a feeding site for some time, but they remain mobile and can change feeding sites continuously, unlike armoured scale insects. Also unlike armoured scale insects, mealybugs have a continuous gut and eject copious quantities of honeydew from their anus. The honeydew is a nutritional waste product as far as the mealybug is concerned, but it can cause direct and/or indirect economic loss (depending on the crop and marketing issues) by spreading over the leaf or fruit and becoming a substrate for sootymould fungal growth. Furthermore, the high sugar content of honeydew is valued by ants, which may come to rely on it as a food source and may even farm the mealybugs to maximise its production. By protecting the mealybugs from attack by natural enemies, the mealybug/honeydew/ant interactions can have a dramatic impact on the pest status of mealybugs, and even on the crop ecosystem as a whole. In many tropical agroecosystems ant control is the most important prerequisite for mealybug control. In northern New Zealand vineyards, P. longispinus populations usually develop through 3 generations a year. First instar larvae move on to the foliage at budburst, and develop to maturity before moving to sheltered spots under the bark, or (as the season progresses) within grape bunches or in dense canopy, where they reproduce. Mortality during winter is very high, so young crawlers are always difficult to find in spring. The reproduction rate is so high, however, that numbers seem to explode in mid-summer and leaves and fruit may rapidly become covered with mealybugs and sootymould (Charles 1981).

54 Endosymbionts of mealybugs Some insects have cultivated intimate relationships with mutualistic bacteria since their early evolutionary history (Dohlen et al. 2001). Mealybugs, aphids, psyllids and whiteflies are plant sap-sucking insects with an obligate association with prokaryotic endosymbionts whose primary function appears to be the synthesis of essential amino acids that are lacking in plant sap (Douglas 1998, Moran et al. 2003). Particular endosymbionts are acquired through vertical, maternal transmission (Baumann et al. 2002, Baumann 2005) and stored within the mealybugs body cavity in aggregates of specialized cells called bacteriocytes (Baumann & Baumann 2005). This domestication is probably in all cases accompanied by a major reduction in endosymbiont genome size (Baumann 2005). Most, if not all, microbial endosymbiont relationships have a chemical basis in which compounds produced by one partner are useful for the other. But endosymbioses can also entail the transfer of genes from one partner to the other, which in some cases cements two cells into a bipartite, co-evolving unit. Evidence for this can be seen through phylogenetic analysis (Hoffmeister & Martin 2003). Phylogenetic analyses are consistent with an infection of a mealybug ancestor with a precursor of the endosymbiont followed by the vertical transmission of the endosymbiont to progeny (Baumann & Baumann 2005). Phylogenetic trees based on endosymbiont 16S-23S ribosomal DNA (rDNA) from psyllids, whiteflies, and aphids together with congruent trees based on host genes, suggest that the association arose from an infection of the insect host with a different prokaryote followed by co-speciation (Baumann et al. 2000, Thao et al. 2000, Martinez-Torres et al. 2001, Thao & Baumann 2004, Downie & Gullan 2005). Comparisons of the mealybug mitochondrial DNA fragments with those from other members of the Sternorrhyncha and other arthropods indicate that mealybug mitochondrial DNA has some unusual properties. Baumann et al. (2002) sequenced two fragments of the genome of Tremblaya princeps, the endosymbiont of mealybugs, which is a member of the beta subdivision of the Proteobacteria. Each of the fragments (35 and 30 kb) contains a copy of 16S-23S-5S rRNA genes. The genome of T. princeps has a number of properties that distinguish it from the genomes of Buchnera aphidicola and Carsonella ruddii, the endosymbionts of aphids and psyllids, respectively. Among these properties are a high G+C content (57.1 mol%), the same G+C content in intergenic spaces and structural genes, and similar G+C contents of the genes encoding highly and poorly conserved proteins. The high G+C content has a substantial effect on protein composition; about one-third of the residues consist of four amino acids with high-G+C-content codons (Baumann, Thao et al. 2002). The endosymbiotic bacteria of the whitefly Bemisa tabaci have been implicated in the insects ability to transmit tomato yellow leaf curl virus (Morin et al. 1999). This species has also been found to harbour novel bacteriocytic related chlamydial species within the family Simkaniaceae (Thao et al. 2003). Where virus transmission is of the persistent and circulative type (as above), then the symbiont-produced proteins assist in the transfer of virus from the insects haemocoel to the salivary glands. Virus transmission by mealybugs is the semipersistent type, and it may be that the endosymbionts either do not affect virus transmission, or do so using a different mechanism. So, knowledge of the molecular make-up of endosymbionts of mealybugs is developing rapidly, although their full function is not yet understood. Clearly, if studies show that endosymbionts are associated with mealybugs ability to vector GLRaV-3 there is an added economic imperative to examine them in greater detail. In addition, it would open the door wider to intriguing possibilities for GLRaV-3 control strategies by disrupting the mealybug/endosymbiont association. Application of appropriate antibiotics, for example, or infection with competing bacteria, may allow quite novel forms of mealybug pest control in the future.

55 Mealybug vectors of plant viruses Mealybugs are known to transmit virus diseases in several economically important crops. The virus-vector association is perhaps one of the most important biotic interactions in the epidemiology of plant virus diseases (Spence 2001). Transmission by vectors provides the primary method of spreading diseases that may lead to severe economic losses. Viruses, in order to survive and spread, have evolved specific associations with a variety of vectors that allow them to be transmitted from plant to plant. Pirone & Blanc (1996) characterised two types of association between viruses and vectors: in the first type, the virus is carried internally and must cross membrane barriers to enter and exit the vector. The best-known examples of this type of association are viruses that are circulatively transmitted by insects; the second type is an external association, i.e., the virus does not cross membranes barriers, and are non-circulatively transmitted by insects and nematodes. These viruses are nonpersistent or semi-persistent and are associated with the cuticular linings of the mouth parts or alimentary tracts of the vectors (Pirone & Blanc 1996). Of the former association the most important arthropods vectors of plant viruses belong to the sap-sucking Hemipterans particularly mealybug, scale, and aphid insects which are attributed with the transmission more than 380 viruses from 27 virus genera (Spence 2001). Viruses transmitted by mealybugs belong to the genus Ampelovirus in the family Closteroviridae (Ling et al. 2004) and include pineapple mealybug wilt-associated viruses 1 and 2 (Carter 1973, Ullman et al. 1989, German et al. 1992, Hu et al. 1996, Sether et al. 1998, Nickel et al. 2000, Santa Cecilia et al. 2001, Sether & Hu 2002), little cherry virus 2 (Raine et al. 1986, Eastwell & Bernardy 2001), and Grapevine leafroll-associated viruses (several serotypes, but especially 3) (Rosciglione & Castellano 1985, Engelbrecht & Kasdorf 1987, Engelbrecht & Kasdorf 1990, Boscia et al. 1993, Jordan 1993, Digiaro et al. 1994, Minafra & Hadidi 1994, Garau et al. 1995, Cabaleiro & Segura 1997a, b, La Notte et al. 1997, Petersen & Charles 1997, Borgo & Michielini 2000, Sforza & Greif 2000, Golino et al. 2002, Sforza et al. 2003) (Table 3.6).

56 Table 3.6 Examples of mealybug vectors of plant viruses Mealybug species Pseudococcus longispinus Pseudococcus calceolariae Pseudococcus viburni Pseudococcus maritimus Planococcus citri Planococcus ficus Heliococcus bohemicus, Phenacoccus aceris GVA & GVB Planococcus ficus Pseudococcus viburni Pseudococcus longispinus Pineapple mealybug wilt Dysmicoccus brevipes Dysmicoccus neobrevipes African Cassava Mosaic virus Phenacoccus manihoti ? Phenacoccus herreni ? Stictococcus vayssierei ? Cocoa Swollen Shoot Planococcoides njalensis badnavirus (CSSV) Planococcus citri Banana Streak Virus (BSV) Planococcus citri Saccharicoccus sacchari Dysmicoccus brevipes Little cherry virus 2 Phenacoccus aceris Taro bacilliform badnavirus Pseudococcus solomonensis Mealybug vectored viruses of plants other than grapes Knowledge of the mealybug/virus interactions in crops other than grapes may provide valuable insights into the mealybug/grapevine/GLRaV-3 relationships. The biologies of mealybugs and their natural enemies in agroecosystems other than grapes are often much better understood than in New Zealand as a result of many years of focused ecological research. It is clear that some of these interactions are important in determining rate of spread of virus, and it is almost certain that some are relevant to the spread of GLRaV-3 in New Zealand vineyards. In addition the changing nature of New Zealand vineyard ecosystems (with the adoption of Sustainable Winegrowing NZ), the increasing spread of Argentine ant into grape growing regions, and the threat from new mealybug species, all warrant an interest in these examples. Pineapple Mealybug Wilt-associated Closterovirus-2 (PMWaV-2) Mealybug wilt of pineapple (MWP) is one of the most destructive diseases of pineapple, Ananas comosus. Pineapple mealybug wilt associated virus (PMWaV) is a complex of closteroviruses forming two associated viruses: PMWaV-1 and V-2 (Hu, Sether et al. 2005). Both PMWaV. are acquired and transmitted by the pink and grey pineapple mealybugs (Dysmicoccus brevipes and D. neobrevipes respectively) and have caused major MWP epidemics in most tropical and subtropical regions throughout the world where pineapples are cultivated (German et al. 1992). Both PMWaV-1 and PMWaV-2 are distinct and share <50% nucleic acid homology with each other based on available sequences (Hu et al. 2005). A third associated virus (PMWaV-3) has Plant virus/disease GLRaV-3

57 been tentatively designated by Sether et al. (2005). Putative PMWaV-3, like PMWaV-1 and PMWaV-2, is transmissible separately or in combination with other PMWaV by D. brevipes and D. neobrevipes respectively. Moreover, mealybug feeding on pineapple plants in the absence of the PMWaV, or the presence of the PMWaV in the absence of mealybug feeding, does not induce MWP symptoms. Hu et al. (2005) have shown in transmission experiments that the presence of PMWaV-2 and mealybug exposure are necessary for the induction of MWP. Similarly, plants infected with PMWaV-3 that were continuously exposed to mealybugs did not develop symptoms of MWP in the absence of PMWaV-2 (Sether & Hu 2002, Sether et al. 2005) thus reinforcing the crucial role of V-2 in the etiology of MWP. The effect of PMWaV-1 infection in pineapple and other crops such as ratoon results in a reduction in fruit size, weight and yield per plant compared with by PMWaV-free plants (Sipes et al. 2002). More early-ripening fruit (30%) were produced by plants infected with PMWaV-1 than by PMWaV-free plants (P < 0.05). Infection with PMWaV-may be one of the reasons for asynchronous fruit ripening, which is a top limiting factor for pineapple production in Hawaii (Sipes et al. 2002). Methods of control of MWP include meristem propagation through tissue culture which has been shown to produce PMWaV-free plant material. Constructs of selected PMWaV-2 genes are being introduced into pineapple to develop MWP-resistant transgenic plants by inducing gene-silencing mechanisms (Hu et al. 2005). Alternatively, understanding the ecological and biological aspects of the mealybug vectors has helped in MWP management programmes. All stages of D. neobrevipes acquired PMWaV, although vector efficiency decreased significantly in older adult females (Sether et al. 1998). The probability of a single third-instar immature transmitting the virus was 0.04. Both species of mealybug acquired and transmitted PMWaV from infected pineapple material that had been clonally propagated for decades, and both species acquired PMWaV from sources previously infected with the virus by the other mealybug species (Sether et al. 1998). Much research has been carried out into understanding the interactive association between tending ants and PMWaV mealybug vectors and, moreover, their effect on MWP and natural enemy establishment (Rohrbach et al. 1988, Duodu & Thompson 1992, Petty & Tustin 1993, Jahn & Beardsley 1996, GonzalezHernandez et al. 1999, Jahn & Beardsley 2000, Pitaksa et al. 2000, Jahn et al. 2003, Gonzalez-Hernandez et al. 2005, Pandey & Johnson 2005a, b, Taniguchi et al. 2005). For example, Sether et al. (1998) demonstrated that the presence of ants was positively correlated with an increased rate of virus spread when caged with D. brevipes. African Cassava Mosaic Virus (ACMV) [cassava African mosaic bigeminivirus] Cassava (Manihot esculenta), one of the worlds major food crops is grown throughout the tropical regions of the world. Cassava originated in the neotropics but was introduced into West Africa from Brazil in the 16th century, and to Asia during the 17th century. Consequently the greatest diversity of cassava pests, as well as their natural enemies is found in the neotropics. In Africa, several native insect species, including the mealybug (Phenacoccus herreni), depress yield significantly primarily during dry periods. In the early 1970s, the cassava green mite, Mononychellus tanajoa and the cassava mealybug P. manihoti were inadvertently introduced to Africa from the neotropics. These pests have since spread throughout most of the cassava-growing regions of Africa, causing severe crop losses. The unauthorised movement of cassava germplasm, between and within continents, involves a risk of accidentally introducing additional pests (Bellotti, Braun et al. 1994). Mononychellus tanajoa and P. manihoti are by far the most economically important arthropod pests and have been well studied (Okoli 1982, Akinlosotu 1983, Hahn et al. 1990, Calatayud et al. 1996, Renard et al. 1998, Bento et al. 1999, Souissi 1999, Bento et al. 2000, Le Ru & Mitsipa 2000, Zeddies et al. 2001, Dorn et al. 2003). The long growing period and diverse agroecological conditions under which cassava cultivars are grown expose them to one or

58 more of these problems and the losses can be devastating (Mahungu et al. 1994). NPK fertilization can have a significant impact on symptom severity of African cassava mosaic virus (Ogbe et al 1993). Control strategies are based on host plant resistance, biological control and cultural practices. Adequate levels of resistance have been identified for mites and whiteflies and moderate levels for mealybugs (Bellotti et al. 1994). Pest populations of P. manihoti, were reduced successfully by encyrtid Hymenoptera, especially Apoanagyrus lopezi and A. diversicornis, throughout most of sub-Saharan Africa (Neuenschwander 2001, Zeddies et al. 2001) and North Eastern Brasil (Bento et al. 2000). Breeding cassava plants selected for different levels of antibiotic resistance to P. manihoti, M. tanajoa and other pests has proven to be a helpful addition to the various control measures that have been implemented (Okoli 1982, Hahn et al. 1990, Le Ru & Tertuliano 1993, Tertuliano et al. 1993, Mahungu et al. 1994, Tertuliano et al. 1999). Moreover, cassava mealybug parasitoids such as A. lopezi tended to perform better if cassava cultivars were selected for their strong antixenosis but low antibiotic characteristics (Souissi et al. 1997, Souissi et al. 1998a, b). Indeed, multiple pathogen resistance helps to ensure stability of crop performance (Mahungu et al. 1994). The spread and severity of Stictococcus vayssierei, a root mealybug of cassava was monitored from 1990 to 1998 with 5 cassava genotypes in five sites in the sub-humid forest region of Cameroon (Ngeve 2003). Pest frequency was low (12.5%) in 1990 but rapidly increased to 87.5% in 1994. Pest impact was more severe in the dry season than in the wet season. S. vayssierei adversely affected root yields and root counts. Improved resistance clones of cassava were more tolerant of S. vayssierei than the local variety. The pest was more severe when cassava was planted on the flat (30 mealybugs/hill) than when planted on ridges (16 adults/hill) and plants also sprouted better (91%) when cassava was planted on ridges than when planted on the flat (71%). Stictococcus vayssierei infestation was also more severe when cassava was intercropped with maize and groundnuts than when planted alone. Monocropping is recommended in areas where pest impact is very severe. Also, disinfestation of cuttings with insecticidal bioproducts should be exploited to reduce pest impact (Ngeve 2003). Chemical pesticide application to control cassava pests is discouraged and efforts are being made to develop Integrated Pest Management (IPM) programmes which do not incorporate pesticide use (Mahungu et al. 1994). Cocoa Swollen Shoot badnavirus (CSSV) Cocoa swollen-shoot badnavirus (CSSV) causes a disease that seriously affects the cocoa industry, particularly in Ghana. The virus is considered to have spread to cacao from forest trees of the order Tiliales (such as Cola spp. and Sterculia spp.) occurring throughout the cacao-growing areas, and both cacao and these trees provide a reservoir of infection for new plantings. Transmission is principally by mealybugs. An experiment carried out by Eguagie (1972) in Nigeria involved different durations of infection feeding (18-36 h) and transmission feeding (12-30 h) in order to assess the vector potential of Aphis gossypii. About 28,000 individual aphids were used, and virus symptoms failed to occur in all but one of 1038 test seedlings observed for virus symptom expression. It was concluded that A. gossypii is not a vector (of the Offa-Igbo 574 isolate), the single transmission obtained being attributable to mealybug contamination of test seedlings in the insectary (Eguagie 1972). Planococcoides njalensis was assumed to be the principal vector for many years, but work by Bigger (1981) indicated that the more mobile Planococcus citri was becoming dominant. Biggers (1981) census of the mealybug vectors of CSSV in Ghana, between 1971 and 1976, ranked the 6 mealybug species

59 encountered in descending order of abundance: Planococcoides njalensis, Planococcus citri, Phenacoccus hargreavesi, Ferrisia virgata, Pseudococcus concavocerarii, and P. calceolariae. Planococcoides njalensis occurred at a much higher density on infested trees than other species but was found on fewer trees than either Planococcus citri or Phenacoccus hargreavesi. Over 13-week periods, the probability of finding a mealybug on a particular tree was 0.87 for Planococcus citri, 0.77 for Phenacoccus hargreavesi, 0.36 for F. virgata, 0.32 for Pseudococcus concavocerarii, 0.23 for Planococcoides njalensis and 0.05 for Pseudococcus calceolariae. The results suggest that the first 4 species are more mobile than Planococcoides njalensis and could be more important in the spread of disease than has previously been supposed (Bigger 1981). Virus spread results mainly from viruliferous mealybugs moving between trees in physical contact, with relatively infrequent dispersion over greater distances by wind-borne first-instar nymphs. An integrated ecological approach to studying cocoa viruliferous pests and particularly in swollen shoot disease and has been a recurring feature of West African research from some of the earliest studies at Tafo and Ibadan to the present time (Bigger 1972, Bigger & Kumar 1975, Campbell & Kumar 1975, Legg & Kumar 1975, Bigger 1981, Legg & Lockwood 1981, Lockwood 1981, Firempong 1984, Thresh & Owusu 1986, Ollennu et al. 1989, Brew 1992, Ollennu et al. 1996, Padi et al. 1996, Adu Ampomah et al. 1999). Early ecological work by Roivainen (1976) centred on transmission of CSSV with several mealybug species, particularly Planococcoides njalensis when this species was considered the principal vector. Food uptake from virus infected plants was positively related to the infectivity of P. njalensis with CSSV isolate 1A. With this virus and P. njalensis, preliminary fasting increased the rate of settling down to feed, food uptake and infectivity of the mealybugs. The length of acquisition access time for maximum transmission was 48-72 hours. CSSV was not acquired during short feeds up to 1 hour but persisted in mealybugs up to 72-96 hours and loss of infectivity was exponential with a half-life of 12-13 hours. Moreover, this virus persisted in mealybugs through post-acquisition moulting and the vectors were more efficient after post-acquisition feeding than fasting. The virus had a short latent period in test plants and >98% showed symptoms in 3 months. Infectious mealybugs were used to test virus resistant and tolerant cacao selections. Transmission was not affected by light or darkness. Food uptake was highest and virus transmitted most frequently at 29-36C. The transmission of CSSV isolate 1M or cacao mottle leaf virus isolate 1C with P. njalensis as vector, or CSSV 1A with Ferrisia virgata, was similar to that of CSSV 1A with P. njalensis (Roivainen 1976). Field studies were carried out to assess the resistance to CSSV in young, bearing trees by monitoring the natural spread by the mealybug vectors from adjacent graft-inoculated line sources. Apparent rates of virus spread were estimated by regressing proportions of infected trees on time from inoculation. In some hybrids between parents derived from Upper Amazonian material, spread was only 25-40% of that in currently recommended varieties. Such resistance was considered to be the most effective and feasible means of reducing economic losses, and resistant hybrids could be made available quickly for large-scale use (Legg & Kumar 1975, Legg & Lockwood 1981, Lockwood 1981). A large-scale programme was begun in Ghana in 1946 to control the disease by identifying and destroying infected cacao trees. The programme, which is still in operation, has been partially successful in controlling the disease, but less successful in eradicating infection. Some of the difficulties of controlling the disease in this way were shown by data on the incidence and distribution of latent infection associated with apparent infection, which are being obtained by a coppicing technique. In contrast Thresh & Owusu (1986) found there was little evidence to justify coppicing trees around treated outbreaks; routine eradication measures would be more effective if a suitable insecticide were introduced for controlling the mealybug vectors. Removing only trees with symptoms was inefficient unless outbreaks were

60 very small; it was better to also remove all apparently healthy adjacent trees. More drastic measures were justified in large outbreaks and where these were numerous it was more effective to clear the whole areas and replant in small compact blocks (Thresh & Owusu 1986). Preliminary data presented by Legg & Kumar (1975) indicated that where the disease is prevalent, infection can be widespread and apparently isolated outbreaks complex. Moreover, it appears that even in the virus-sensitive Amelonado cacao, infection can remain latent for periods exceeding 2 years, or reinfection of outbreak areas occurs from other plant species. It is considered that ultimately, the cultivation of cacao types resistant to infection will provide a more satisfactory solution to the problem; work has been in progress along these lines at Tafo since 1969 (Legg and Kumar 1975). Lockwood (1981) provides an account of studies carried out to assess seed inoculation experiments. Differences in infection levels following seed inoculation by the mealybug vector Planococcoides njalensis (Laing) were used as a shortterm test of virus resistance in a range of cacao progenies. The results were highly correlated with virus resistance determined in field tests. The best assessment of seed inoculation experiments allowed for progeny-dependent variation in the latent period of the virus (Lockwood 1981). Advances in mild strain cross-protection of cocoa against CSSV have been identified (Ollennu et al. 1996). These comprise three mild strains had more recently been identified in addition to four that had been maintained since the 1950s and 1 recent strain (NI) has being evaluated in the field. Five of the mild strains were transmitted to inter-Amazon and Amazon x Amelonado hybrids by both patch grafts and a mealybug vector. Mild strain infection was detected by virobacterial agglutination. In gauzehouse and field trials mild strains protected young plants against the CSSV severe 1A-New Juaben isolate (Ollennu et al. 1996). Continuing progress on breeding for CSSV resistance in Ghana includes: variation in mealybug vectors; variation in CSSV isolates; resistance breeding, including varieties in use in 1969, development of Inter-Amazon hybrids (1969-81), methods of screening for resistance (manual inoculation of seeds and mealybug inoculation), searching for pollen parents to replace Amelonado, searching for female parents, and mutation breeding; mild strain cross protection; and identification of markers linked to loci controlling resistance to CSSV (Adu Ampomah et al. 1999). Banana Streak Virus (BSV) Banana streak virus (BSV) is typically vectored by mealybugs. Several species of mealybug appear to be responsible for BSV transmission particularly Planococcus citri and Saccharicoccus sacchari. However in Ugandan Plantations where Planococcus citri have not yet been recorded, Dysmicoccus brevipes (pineapple mealybug), S. sacchari (sugarcane mealybug) and Pentalonia nigronervosa (banana aphid) found to be the most abundant insect species (Kubiriba et al. 2001), however, their role and that of other agents in BSV transmission is not well documented. Insect samples were collected from banana farms in sites with low, moderate and high BSV infections in Uganda. Subsequently, live mealybugs and aphids were again collected and used in acquisition, retention and transmission tests, and BSV diagnosed using TAS-ELISA from banana fields sampled. Abundance of D. brevipes was positively and significantly correlated with BSV incidence unlike that of Pentalonia nigronervosa. Transmission studies in the screenhouse showed that mealybugs acquired BSV one day after feeding on virus sources and approached optimum acquisition after the third day. Pineapple and sugarcane mealybugs retained BSV up to 5 days from the day of transfer from the virus source. BSV was first detected in the recipient banana plants 4 weeks after transmission using pineapple mealybug and 6 weeks after inoculation using sugarcane mealybug. Under screenhouse conditions, both mealybugs therefore appear to transmit BSV semipersistently (Kubiriba et al. 2001).

61 Taro bacilliform badnavirus (TaBV) Laboratory transmission of TaBV by Pseudococcus solomonensis was reported for the first time in Samoa (Macanawai et al. 2005). Mealybug vectored grape viruses other than GRLaV-3 Grapes are the most widely grown fruit crop in the world, with grapevine fanleaf virus (GFLV) grapevine leafroll virus (GLRV) and Grapevine Corky bark diseases caused by GVA and GVB being among the most damaging and widespread of the array of viruses that affect them. As an example of the severity and presence of grape viruses, surveys were carried out in different vineyards of Apulia, Italy, to assess the incidence of grapevine A virus [grapevine stem pitting associated closterovirus (GSPaV)], grapevine leafroll associated closterovirus 1 (GLRaV-1), GLRaV-3, grapevine fanleaf nepovirus (GFLV) and grapevine fleck virus (GFV) and establish possible correlations of rugose wood with closteroviruses. Of 1828 vines tested individually, 88.3% were infected by at least 1 of the 5 viruses. GLRaV-3 (67.3%), GSPaV (55.7%) and GFV (59.3%) were the most widespread. GLRaV-1 (11.8%) and GFLV (7.1%), which prevailed in older vineyards, were detected less frequently. GSPaV appeared to be the main causal agent of wood rugose symptoms, enhanced by the presence of GLRaV-1 and GLRaV III. Many grape viruses are spread primarily through infected propagating material. However, increasing evidence indicates the consistent presence of closteroviruses in vineyards established with certified material now appears to be associated with their possible spread in the field by insect and nematode vectors. The most prevalent insect vectors are mealybugs (Pseudococcidae) and to a lesser extent softscale (Coccidae) insects. Grapevine fanleaf virus (GFLV) GFLV is a nepovirus of great economic importance and is transmitted by a soil borne nematode. Fortunately the nematode is not present in New Zealand (see Section 1). Grapevine diseases caused by GVA and GVB Grapevine virus A (GVA) (the putative agent of Kober stem grooving) and Grapevine virus B (GVB) (the putative agent of corky bark) are vitiviruses (see Table 1.2) that share some biological and physicochemical properties (Boscia et al. 1993). Both GVA and GVB are successfully transferred from infected grapevines to herbaceous hosts by several species of mealybugs, including Planococcus ficus, Pseudococcus affinis (Garau et al. 1995), P. longispinus (Rosciglione & Castellano 1985, La Notte et al. 1997), and the coccid Ceroplastes rusci from Tunisia (La Notte et al. 1997), thus confirming that transmission by mealybugs of grapevine vitiviruses is not species-specific. A preliminary survey of populations of different mealybug species collected by La Notte et al. (1997) in the vineyards of different Mediterranean countries showed that 77% contained GVA. In Italy, transmission of GVA by the mealybug P. longispinus was studied by La Notte et al. (1997). Phases of the transmission process (acquisition access time, retention and inoculation access time) were checked by bioassays using Nicotiana clevelandii and RT-PCR. Virus was acquired either from infected N. clevelandii or from a purified preparation through a stretched membrane. Evidence was obtained that P. longispinus instars transmit GVA in a semi-persistent manner: they acquired GVA in as little as 15 min when feeding on N. clevelandii or 12 h when feeding on purified virus preparations through a membrane; they retained the virus for up to 48 h when fasting, but no longer than 15 h when allowed to feed on herbaceous hosts following serial transfers; they were able to transmit GVA to healthy plants with no latent period, after a 30 min feeding (the shortest inoculation access time tested). GVA occurs extensively in Stellenbosch, South Africa (Engelbrecht & Kasdorf 1987). It was not found to be associated with either stem pitting or fleck symptoms. It occurred mostly in association with undecorated closterovirus (CV)-like particles. Limited ISEM with decoration

62 of the CV-like particles using antiserum to a Swiss isolate of CV 2 200 nm, (CV type 1) confirmed the presence of a second CV in local grapevine sources. In addition, it indicated the presence of a third CV-like particle longer than GVA (GVB: see Boscia et al. 1993). GVA together with undecorated CV-like particles were also detected in initially GVA-free LN-33 grapevines growing under field conditions and naturally infected with leafroll. GVA and CV type I were also present in the vine mealybug, Planococcus ficus, following exposure to a grapevine source carrying these viruses. However, because of lack of CV type 1 antiserum, only GVA was confirmed in grapevine following controlled transmissions with P. ficus although undecorated CV-like particles were present (Engelbrecht and Kasdorf 1987). GLRaV-1, -2, -5 The viruses transmitted to grapes by Planococcus ficus and their symptoms were studied by Engelbrecht & Kasdorf (1990b). Controlled transmissions of grapevine leafroll closterovirus with P. ficus resulted in typical symptoms on the recipient LN-33 indicator vines in subsequent seasons. Symptoms on individual vines varied from slight to strong reddening and rolling of the leaves, the most severe being nearly as strong as those on vines chip-budded with the donor Vitis vinifera Waltham Cross 22/3 clonal source. With only 2 of the 12 vines included in the transmission trials remaining symptomless, P. ficus appeared to be an efficient vector of grapevine leafroll disease. Although present in the Waltham Cross 22/3 source, P. ficus failed to transmit both GLRaV-1 and -2. GLRaV-1 was the only closterovirus not detected in extracts of P. ficus following an acquisition access period of 2 weeks on the Waltham Cross 2/3 source (Engelbrecht & Kasdorf 1990). More recently, a survey was conducted in the northern wine-growing regions of France where GLRaV-1 and -3 are the most widespread grape Ampelovirus species. Mealybugs, Heliococcus bohemicus, Phenacoccus aceris and the soft scale coccid, Parthenolecanium corni, were confirmed vectors, for the first time, of both GLRaV-1 (and -3) with a transmission efficiency of 14%, 23% and 29% respectively (Sforza et al. 2003). An experiment conducted in California, USA to examine the ability of mealybugs to transmit grapevine leafroll-associated viruses, found, for the first time, that GLRaV-5 was transmitted by Pseudococcus longispinus (Golino et al. 2002). These results are consistent with the recent taxonomic analysis of Ampelovirus and Closterovirus (see section 1). Transmission of GLRaV-3 by mealybugs Grapevine leafroll associated closterovirus serotype 3 (GLRaV-3) is transmitted by at least 8 species of mealybugs (Table 3.6). The transmission of GLRaV-3 from grapevine to grapevine by Pseudococcus longispinus and P. calceolariae was demonstrated by Petersen & Charles (1997). Controlled transmission experiments using first and third instars of each species revealed GLRaV-3 transmission to recipient vines by P. longispinus and P. calceolariae first instars only. An increase in virus titre within the season did not significantly alter the transmission rate of GLRaV-3 by either P. longispinus or P. calceolariae first instars (Petersen & Charles 1997). Pseudococcus longispinus is also one of four species in California shown to transmit GLRaV-3, the others being P. viburni, P. maritimus and Planococcus citri (Golino et al. 2002). A recent study in the northern wine-growing regions of France, where GLRaV-1 and -3 are the most widespread grape Ampelovirus species, showed that Heliococcus bohemicus and Phenacoccus aceris were capable of a transmission efficiency of 14% and 23% respectively. This is the first report of GLRaV-1 and -3 transmission by mealybugs in France (Sforza et al. 2003). Grapevine leafroll closteroviruses are now present in most, if not all, winegrowing regions of the world. A preliminary survey of populations of different mealybug species collected in the vineyards of different Mediterranean countries showed that 33% contained GLRaV-3 (La

63 Notte et al. 1997). In New Zealand vineyards GLRaV-3 can spread at a rate that doubled the number of infected vines each year (Jordan 1993) and is causing increasing levels of economical damage (Bonfiglioli et al. 2002). In order to estimate the recent spread of the disease, a large number of vineyards has been monitored. The results of this work demonstrate a high spread rate of GLRaV-3. Two types of spread were observed: one was typical of spread by insects and the other demonstrated randomly scattered disease patches, which might indicate planting of infected propagation material. In South Africa the vine mealybug, Planococcus ficus, is the principal vector (Walton 2003) and GLRaV-3 infection is widely recognised as the biggest strategic threat to the longevity of vineyards. Commercial wine and rootstock cultivars in the Western Cape were tested in 1970 for a variety of viruses. An extremely high percentage (99.9%) of vines had been infected with viruses (Nel & Engelbrecht 1972). Unspecified grape leafroll viruses represented 68.4% of these virus infections. A much lower rate of infection was detected in rooted scion cultivars. The high percentage virus infection in South Africa is attributed to the use of rootstocks since the turn of the century. Field spread of graft-transmissible diseases and associated viruses in a Vitis vinifera Tinta Barocca vineyard to healthy interplanted LN-33 (a Vitis hybrid) vines was studied by Engelbrecht & Kasdorf (1990a, b) in Stellenbosch, South Africa. In the 7th season of planting, 71% of the vines showed symptoms of grapevine leafroll (GLR) disease, and all infected vines tested positive for GRLaV-3. Approximately 30% of diseased LN-33 also contained GVA which repotedly depended on GRLaV-3 for transmission (Englebrecht & Kasdorf 1990a). One of these vines also carried GRLaV-2, while another contained isometric virus-like particles which were serologically similar to Castellano's viruslike particle, and which produced typical grapevine fleck symptoms on V. rupestris St George. On indexing to V. vinifera cultivars Shiraz and Gamay, a further 2 LN-33 vines with GLR symptoms were found to be latently infected with grapevine Shiraz Disease. Field spread of grapevine corky bark disease to an LN-33 vine, found infected with GRLaV-3, was observed. GRLaV-1 was not detected in the LN-33 vines despite its presence in the Tinta Barocca vines. Random field spread of several graft-transmissible diseases and viruses into the interplanted LN-33 vines suggested active participation of Planococcus ficus (Engelbrecht & Kasdorf 1990a, b). More recent evidence shows that GVA is consistently associated with Shiraz Disease (Goszczynski and Jooste 2003). In a recent survey of vineyards in Washington and Oregon (USA) Martin et al. (2005) found significant GLRaV-3 infection rates of 6.5% and 4.4% respectively and concluded that these were similar to other wine grape production areas in North America. The grape mealybug Pseudococcus maritimus is known to vector GLRaV-3 in both Washington and Californian vineyards. In Canada GLRaV-3 was detected in 10.8% of wine and grape juices (McKenzie et al. 1996). The incidence of grapevine leafroll associated viruses is high in all winegrape production areas in northern Spain. GLRaV-3 infection is widespread with an incidence of about 40% (Cabaleiro 1997). The spread of the disease is quite rapid in northern Spain with 21.3% infection be reached in a healthy block within three years of planting. Most new plantings in Spain are grown on virus-free certified rootstocks but the spread of virus to new healthy vineyards is a recognised concern for their certification programme. The known and suspected Pseudococcidae vectors of GLRaV-3 have a wide host range including many herbaceous and woody plant species. Mealybugs are considered to be uncommon in both the colder winegrape production regions of Europe (Germany) and in the warm humid regions of Europe e.g., northwest Spain (Cabaleiro 1997). However, high population densities are not required for virus transmission and this may be the case where natural spread of GLRaV-3 has occurred but no vectors have been identified e.g., Yugoslavia (Dimitrijevic 1973) and Australia (Habili et al. 1995).

64 Some characteristics of the acquisition and transmission of GLRaV-3 by Planococcus citri were determined using ELISA and transmission assays (Cabaleiro & Segura 1997). Groups of 5 mealybugs were used, as determined by the results of ELISA of insect groups of various size. The virus was transmitted to only 1/10 test plants (Vitis vinifera Albarino and Cabernet), each of which had been exposed to a group of insects fed on GLRaV-3 infected plants for at least 3 days, even though more than 80% of the insect groups were expected to contain viruliferous individuals under these conditions. Viruliferous mealybugs transferred to potato plants could retain the virus for up to 24 hours but lost the capacity for effective transmission to vines within 1 hour after transfer. In newly infected vines, the virus remained latent or undetectable by ELISA for at least 13 months (Cabaleiro & Segura 1997). More recently, Krger et al. (2006) have experimentally shown that the acquisition access period for Planococcus ficus and Pseudococcus longispinus is 1 hour, the inoculation access period is 30 minutes and the retention period for GLRaV-3 is 72 hours. Transmission of GLRaV-3 by other insects In a vineyard of northern Italy where natural spread of leafroll was found, Belli et al. (1994) were unable to record the presence of mealybugs, but found the soft scale coccids Parthenolecanium corni and Pulvinaria vitis. Transmission experiments carried out in insect proof chambers with young nymphs both Parthenolecanium corni and Pulvinaria vitis (both of which can be found in New Zealand vineyards) gave positive results in 2 out of 5 vines. La Notte (1997) detected GLRaV-3 in Ceroplastes rusci, but did not claim that it was a vector. As noted above, the soft scale coccid, Parthenolecanium corni was confirmed as a vector for both GLRaV-1 (and -3) (Sforza et al. 2003). Hence transmission of GLRaV-3 around the world by other species of mealybugs and coccids would appear likely, and will probably be confirmed over time as further research in different countries is carried out. Experiments which show failure of potential vectors to transmit GLRaV-3 may not be reported, especially if all species in the trial fail. In New Zealand, spread of GLRaV-3, in the apparent absence of mealybugs (possibly because numbers were too low to be readily observed), has sometimes been attributed to phylloxera. A small trial in New Zealand collected phylloxera from GLRaV-3 infected vines, but no GLRaV-3 could be detected in them, so the planned transmission trial was aborted (Charles et al. 1997, 1998). The recent advances in virus taxonomy have linked GLRaV-3 transmission ever more closely to mealybugs and coccids as opposed to other insects (Section 1). Hence, transmission in the absence of mealybugs or coccids appears increasingly likely to be via grafting or previously infected plants than by other types of insects (e.g. Phylloxera) or animals (e.g. nematodes). For example, data from a replanted vineyard in the Waikato showed that transient sprouts from vine prunings or roots remaining in the soil after vine removal were infected with GLRaV-3. These shoots were present during spring and early summer at the same time as mealybugs and new, virus-free vines, and provided a ready source for re-infection either directly or via mealybug infested weeds within the rows (Charles et al. 1998). Transmission ecology of GLRaV-3 GLRaV-3 cannot be transmitted mechanically (e.g. by pruning or machinery). It is graft transmissible and the planting of infected but initially symptomless young plants remains a potential source of disease. At least two other plant-originated mechanisms for spread have been suggested transmission to neighbours by self-root grafting, and residual point sources of infection through poor hygiene following pruning or re-planting. Once present in a vineyard, GLRaV-3 in New Zealand is spread by at least 3 species of mealybug vectors (Pseudococcus longispinus, P. calceolariae and P. viburni (Petersen &

65 Charles 1997, Charles et al. 1997) and/or two species of soft scale insects (Hemiptera: Coccidae), Parthenolecanium corni and Pulvinaria vitis (Belli et al. 1994, Cabaleiro & Segura 1997a, Golino et al. 2002, Sforza et al. 2003, Charles et al. 2005). Practical difficulties in measuring virus transmission within a vineyard include the seasonal, patchy distribution of mealybugs, the delayed appearance of disease symptoms in vines until the year following infection, and the relative lack of disease symptoms in white compared with red varieties. Nevertheless, GLRaV-3 can spread rapidly. In one Auckland vineyard all newly planted, disease-free Cabernet Sauvignon vines became infected within 6 years of planting (Jordan et al. 1993). In another, infection in a block of Pinot Noir increased from 12 to 92% over 5 years (Petersen 1996). These and other epidemiological studies have shown that infection usually spreads more rapidly along a row than between rows, that infected vines are often clustered both uni-dimensionally (along a row) and two-dimensionally (across rows) (Habili & Nutter 1997, Cabaleiro & Segura 1997b). It has been presumed that the rapid spread of disease in vines is associated with high numbers of mealybugs, but no quantitative relationship has yet been shown, and at least some mealybugs can usually be found in vineyards in which the disease is either absent or present at very low levels. First instar mealybugs (crawlers) appear to be the most important life-stage for transmission (Cabaleira & Segura 1997a, Petersen & Charles 1997, Sforza et al. 20003), but the ecology of transmission is not well understood. Hence the observed spread of grapevine leafroll disease by mealybugs may be due to a combination of random dispersal, natural crawling and passive assistance from humans (e.g. on machinery or animals). In a recent study in the Waikato (Charles et al. unpublished data), GLRaV-3 spread was apparently influenced by grape variety. GLRaV-3 spread more rapidly in Chardonnay than in Merlot vines, despite usually lower mealybug numbers in the Chardonnay (and in the adjacent mature Breidecker) than in the Merlot (and in the mature Sauvignon Blanc adjacent to them). It remains unknown whether the difference was due to differing resistance factors in grapevines or feeding/transmission factors in mealybugs. Under integrated fruit production management systems (such as Sustainable Winegrowing New Zealand) insecticides should only be used once a defined threshold has been exceeded. The much lower rate of increase in infected vines (at least in Chardonnay) in low mealybug years provided a first indication that a mealybug threshold of perhaps 5 mealybugs/leaf in the third generation may be sufficient to minimise infection and perhaps allow disease management at a block level by cultural activities (Charles et al. unpublished data). A threshold at this time of year may, however, be of limited use if the damage (disease transmission) has been done by the time it is exceeded. To be of real practical value, an economic threshold must be predictable from mealybug samples taken earlier in the season, in time for remedial action (not necessarily limited to insecticides) to be taken. Yet the distributions of mealybug populations within vineyards are notoriously patchy, and they are difficult to sample economically. In addition, the predictions will have to incorporate an understanding of transmission ecology, and so may well be site and variety specific. Determining a threshold that includes GLRaV-3 dynamics for multiple varieties in a vineyard may be too difficult for existing sampling methods. Research on other insect/plant virus associations have illustrated a number of issues that are very relevant to the mealybug/ GLRaV-3 association. In particular it can be recognised that: (a) The ability of mealybugs to transmit a virus does not equate to the efficiency with which it is transmitted. So, the spread and/or impact of disease may be faster or slower if different mealybugs are present. (b) Differences in efficiency of transmission may be because different species, or life-stages within a species, may take up and re-infect the virus particles differently. The differences

66 may be due to inherent differences between the mealybugs, or be more a consequence of the relationship between the virus and plant. (c) Differences among mealybug biologies that may affect transmission include: numbers of generation/year; preference for different species or cultivars of Vitis; dispersal rates and distances; and feeding on different parts of the grapevine. It is clear that there is a dynamic relationship between mealybug vector and virus and grapevine. This dynamic is ultimately responsible for the expression, spread and severity of the disease caused by GLRaV-3. New Zealand is a cool climate viticultural region. It is recognisably different from most other wine producing regions in the world. In addition, it is new; and, being new, the industry is undertaking many experiments with new cultivars and clones in different micro-climates. It seems realistic to expect the vector/vine/virus dynamic, and the expression of that relationship as grapevine leafroll disease in New Zealand, to be unique or at least as recognisably different as the wines it produces. The New Zealand wine industry is fortunate that mealybugs such as Plancoccus ficus and P. citri are not present. Yet the absence of P. ficus means that that research results on this species from South Africa (in particular) may not be directly relevant to New Zealand because our mealybugs have different biologies, and transmission efficiency. In addition, many countries with pest mealybugs in vineyards (e.g. France, Italy, South Africa, Australia) also have longer periods in which to ripen grapes, such that the consequences of virus/vine interactions are perhaps less severe, or at least different. Hence, for example, the development of complex mathematical models (Jeger, Van Den Bosch et al. 1998, Grilli and Holt 2000) to help improve the selection of more effective management strategies for virus diseases and vector control elsewhere (Spence 2001), may be of limited value in New Zealand.

67

SECTION 4: MANAGEMENT OPTIONS WITHIN THE VINEYARD TO LIMIT GLRAV-3 INFECTION AND SPREAD.
INTRODUCTION
The indiscriminate exchange and marketing of uncertified grape graftwood and rootstocks is the single major reason that grape virus diseases became widespread throughout wine growing regions of the world. Eventually the situation became so bad that a number of countries introduced certification programmes for the production of virus-free clonal plant material (see review by Bovey et al. 1975). Extensive international literature is available on the importance of high health planting materials and certification programmes that now ensure that high health, virus-free, planting material is supplied to winegrape growers in many major wine producing countries. Some countries have a relatively long history of certification programmes e.g. since the 1970s in California and Washington State, USA (Martin et al. 2005) while others are yet to establish virus-free certification programmes e.g. Czech Republic (Komnek & Holleinov 2003). Like many other countries, New Zealand has a virus free certification programme (Nimmo-Bell 2005) but also has a relatively recent history of use of uncertified planting materials, which has provided the basis for virus establishment and spread in vineyards. It is not the intention of this review to analyse and describe the characteristics and risk management processes within these various certification programmes, but rather to focus on the management of the GLRaV-3 infection risks and management of insect vectors of this virus in both new and established vineyards. Insects as vectors Much of what we know about viruliferous insect vectors on grapes and transmission ecology of the diseases by these insects on the plants has been reviewed in the previous section. To recap: until the 1980s, the spread of grapevine leafroll diseases was assumed to occur only through infected vines, mainly through the use of asymptomatic American grapevine rootstocks (Golino 1993). In 1983, however, it was found that some mealybugs (Pseudococcidae) were vectors of grape leafroll associated viruses (Rosciglione et al. 1985). Since then the role of mealybugs in the transmission of grapevine leafroll viruses, including GLRaV-3 has been confirmed (see section 3). Some species of scale insects are also reported be vectors of grape leafroll associated viruses but their abundances in vineyards, and therefore significance as vectors, are probably somewhat less than mealybug species. Published literature on the management options to limit the spread of GLRaV-3 within new vineyards is very limited and an extensive search of international literature identified very few scientific publications on the subject of strategies to reduce or eliminate spread (other than use of certified planting material). In some instances this is because of the relatively recent recognition of GLRaV-3 as a problem and identification of the mealybug species involved with virus transmission e.g. Pacific Northwest, USA, (Martin et al. (2005). The most complete information comes from the South African wine industry where GLRaV-3 has long been recognised as a serious problem (Pietersen 2004). Despite the extent of the grape leafroll virus problem in the South African industry there are few, if any, scientific publications on GLRaV-3 management and control (G. Pietersen pers. comm.). The subject is, however, extensively covered in Technical Guides produced for South African winegrape producers.

68 This was the only comprehensive information available and therefore forms the basis of much of the information presented in this section. Managing the spread of GLRaV-3 in vineyards Assumptions The general considerations for this section are based on the following assumptions and current knowledge (Pietersen 2004) about GLRaV-3 and the manner by which it may be spread in winegrape vineyards. 1. GLRaV-3 is not known to infect any plant genus other than Vitis and spread is always considered to have occurred from infected Vitis plants. 2. The virus itself cannot be transmitted by mechanical means, for example in the sap of an infected plant during pruning. 3. The virus can be spread by grafting infected plant material or by vegetative propagation of infected planting material. 4. The virus can also be transmitted from vine to vine by a number of mealybug and scale insect species. 5. Mealybugs are the most important and efficient vectors of GLRaV-3. 6. While reports of scale insects as vectors exist in the scientific literature, mealybugs are, by far, the most commonly reported insect vectors of GLRaV-3. Managing the spread of GLRaV-3 originating within vineyards The most common pattern observed in studies in South African vineyards was the statistically significant occurrence of two or more GLRaV (unspecified) infected vines directly adjacent to each other within rows (Pietersen 2004). Newly infected vines were typically found to spread from these vines along the rows in either direction, before spreading to adjacent plants across rows. This means of spread from an infection focus in a vineyard is termed secondary spread by epidemiologists, as it takes place within a vineyard. It may be caused by non-vector spread, for example by adjacent root graft union, but is more likely to be due to the movement of the virus-infected mealybug crawlers: 1. 2. 3. 4. through mealybug dispersal behaviour on implements on vineyard workers moving along rows by various combinations of these activities.

Roguing The spread from these foci is relatively slow (in plant pathology terms) and is thought to be the consequence of having an initial infected plant at a specific position in the vineyard. Removal of infected vines (roguing) is therefore likely to be a potential means of control of secondary spread, irrespective of the mechanism of spread, as this removes the source of the virus. This control method was attempted in four Mother blocks where infected vines were removed annually after a positive ELISA test for GLRaV-1, or -2 or -3. In these blocks the number of new vines which became infected was dramatically reduced compared with Mother blocks in which roguing was not applied. Some new infections in the rogued blocks did occur however, and were often one or two vines away from where infected vines had been formerly. In this case roguing dramatically reduced the secondary spread of GLRaV (unspecified) but it did not completely eliminate it. Spread may have occurred from the infected vines shortly before the virus testing was done, resulting in still sub-detectable concentrations of virus in these newly infected vines, or: 1. by virus-infected mealybugs left behind after removal of the infected vine that moved to adjoining vines where they transmitted the disease

69 2. because the vine was not completely removed and virus-infected mealybugs that survived on remaining roots and then moved to adjacent vines, which were then subsequently infected. Roguing is not expected to be effective under all conditions. It is only feasible to apply roguing in vineyards where primary infection (that which brings the disease into the vineyard) is low, where the incidence of (secondarily) infected plants is relatively low, and where latent infections (no symptoms visible or negative virus test results but plants are infected already) are not excessive. Roguing may also not be effective in cultivars with long latent infection periods (e.g. green cultivars if using visual assessments). Under these conditions roguing may need to be supplemented by application of a systemic insecticide to make it more effective. Treatment of infected vines with a systemic insecticide before removing them will kill virus infected mealybugs still feeding on the vine and prevent their movement to adjacent plants. Latent infections have been shown during a South African survey to occur mainly on vines directly adjacent to infected vines in rows. Therefore, treating potentially infected plants adjoining infected vines with a systemic insecticide at the time of roguing, will allow any virus present in them to replicate to detectable amounts without the vine serving as a source of further infection, as mealybugs will not survive on them. Subsequent removal of such vines can then be done to eradicate the virus from those foci. If total physical removal of vines is not possible, the use of a systemic herbicide to kill all parts of the infected vine may be required, as re-growth of the infected vine (after the effective insecticide period has passed) can serve as a source from where virus can again be acquired and spread by mealybugs. The number of adjoining vines needing to be treated with a systemic insecticide can be gauged by: 1. the amount of secondary spread (clumps of infected plants) already detectable 2. the age of the vines (older vines are likely to have a longer latent period, allowing a longer period in which sub-detectable spread of the virus can occur) 3. the numbers of mealybugs present. Managing spread of GLRaV-3 originating in other vineyards Another common spatial distribution pattern observed in South African studies is an abundance of GLRaV (unspecified) infected vines at the edges of a vineyard with a diminishing gradient of infected plants towards the middle or other side of the vineyard. This spatial pattern suggests that the virus has been introduced to the vineyard after its establishment from a source external to it and from a specific direction. In most instances the number of infected vines in these gradients increase in the direction of infected, older proximal vineyards. These probably serve as the sources of virus from where mealybugs acquire the virus and spread to neighbouring vineyards by their own motility, by wind, by vineyard workers or implements, or by various combinations of these. Studies on the epidemiology of GLRaV (unspecified) are planned to correlate these gradients with the potential mechanisms through which they arise (e.g. wind dispersal of mealybugs, dispersal of mealybugs on implements, wandering of mealybugs etc.). The possibility also exists that gradients and edge effects detected are not solely due to virusinfected mealybugs being introduced from external sources but that non-infected mealybugs may be introduced or present at higher levels at the edges of vineyards, from where spread of existing GLRaV (unspecified) disease foci would be more rapid. In the absence of direct evidence for the causes of the gradients found some control strategies are proposed: 1. Plant as far away from old, infected vineyards as practical 2. Plant upwind (prevailing summer winds) of old infected vineyards where possible (wind dispersal of vector not quantified on vines) 3. Establish large blocks (as close to a square shape as possible) rather than small blocks with high edge/inside vine ratios

70 4. Consider shelter belts (non-host) between vineyards 5. Ensure that all activities involving vineyard workers and implements are done in the youngest, least infected blocks first, and then only in older, infected blocks (especially during summer months, when mealybug crawlers are present) 6. Take steps to remove mealybugs on vineyard workers and implements between working in infected vineyards and moving to healthy vineyards including change of clothing or washing down equipment. In applying the above strategies, make the assumption that older green cultivar vineyards are GLRaV (unspecified) infected because distinguishing between infected and healthy vines based on symptoms only is not easy. Aerial dispersal of mealybugs In New Zealand, it may be impractical to plant vineyards sufficiently isolated from other mealybug infested vineyards or orchards. Here, the most common mealybug is Pseudococcus longispinus, and long-distance aerial dispersal of young instars of the species has been recorded in Australia. Warm and windy conditions favoured dispersal of first and second instars, which might be blown in excess of 50km/day. In addition, 75% of young crawlers survived desiccation for 48h without food at 32%RH, suggesting that a high proportion of dispersing nymphs could survive a 24-48h dispersal period from vine to vine (Barrass et al. 1994). As young nymphs are the principal vectors of GLRaV-3, it would appear that most New Zealand vineyards could be at risk from aerially borne, infected mealybugs. Crop husbandry practices All crop husbandry practices should be firstly completed in newly planted/younger blocks and then progressively onto older more infected vineyard blocks. The movement of vineyard workers should be limited to a minimum in and around newly planted blocks. In New Zealand this is likely to be more seasonally important i.e. from mid summer to leaf fall when mealybug infestations are high relative to spring populations. South African recommendations suggest that where vineyard workers needing to move from old to new blocks, they have to ensure that their implements are clean by hosing them down with water. They also recommend that the clothing and shoes of workers should also be inspected for the presence of mealybug egg sacs. Managing infested planting material A third spatial pattern observed in South African studies was the presence of randomly distributed GLRaV (unspecified) infected vines (or foci of vines) within young vineyards. This suggests primary spread due to the establishment of infected planting material. In view of the lag phase between virus infection and the ability to detect the virus it can be expected that, while virus infection occurs at Foundation-, Mother-block, and Nursery stages, some plant material from the Certification Scheme will be infected. However, primary infection can only be ascribed with certainty to infected planting material when found in newly established vineyards in instances where these vineyards were established on formerly virgin soils and where the presence of these random foci of infection can be correlated with a specific clone/rootstock combination at that site (there is also the possibility of the infection being due to an external factor at the site). While this mechanism of spread of GLRaV (unspecified) exists, its relative importance in South Africa remains unknown. In order to control this method of spread in red cultivars it is important that nurseries selling vines also improve their control of GLRaV (unspecified) spread at the nursery. It is recommended that only certified planting material be purchased from reputable nurseries, and that this planting material be pre-treated at the nursery with a systemic insecticide effective against mealybugs. Failing this, newly established vineyards should be treated with such

71 systemic insecticides at the time of planting. The systemic insecticide may need to be reapplied in such a manner that it protects the young vine for at least the first two seasons. In this way, if such a vine was latently infected with GLRaV-3, time is given for either the virus concentration to rise to detectable levels or for symptoms to develop, without serving as a source for secondary spread by mealybugs to adjoining healthy vines. Diseased vines must then be rogued during the first two or three seasons. Heat treatment of dormant cuttings As an alternative to systemic insecticides, P. ficus can be killed by immersing dormant cuttings in hot water. A five minute immersion at 52oC killed >99% of mealybugs, and had the added advantage of killing other pests (e.g. Phylloxera, nematodes) and diseases (e.g. Pierces disease, Flavescence dore, Phytophthora cinnamomi) (Haviland et al. 2005). Re-planting strategies There is also circumstantial evidence that a new vineyard may become infected with GLRaV (unspecified) from a previously highly infected one through the survival of virus infected mealybugs between removal of the old vineyard and re-establishment of the new one at that site, possibly on some residual vine roots. Theoretically, this mode of virus spread could be controlled by: 1. eliminating mealybugs (by treating vines with a systemic insecticide) in the old vineyard prior to removal 2. lengthening the period that the site lies fallow between plantings 3. the very thorough removal of old, infected vines and their roots; or herbicide treatment of them; also killing and removing volunteer vines, and then: 4. establishing the new block with certified planting material treated with a systemic insecticide followed by roguing for the first two or three seasons (as for the control of spread of infected planting material). This has potential implications for the New Zealand wine industry, as vineyards are regularly re-established at sites previously planted to grapevines. However, the relative importance of this mode of spread in South Africa is unknown. All vineyard material that might still be in the soil must be removed before new grapevines are established. Mealybugs are able to survive on all living plant material that is found above and below the surface of the soil. No mealybug will be likely to survive if all old plant material is destroyed. As an alternative, any remaining old plant material should be treated with a weed-killer. Do not under-plant beneath infected vines as this leads to the rapid transfer of mealybug from old vines to new plants i.e. it is a very effective way to transfer GLRaV-3 infection. Always use certified material when new blocks are planted. Studies in New Zealand have also shown the importance of vineyard hygiene between vine removal and replanting, and that adventive shoots from old prunings or surviving roots, and under-vine planting can all result in the spread of GLRaV-3 between an old and new crop (Charles et al. 1997). Cultivar susceptibility to mealybug Certain cultivars are more susceptible to mealybug than others, illustrated in Table 4.1 (from Le Roux, 1996).

72 Table 4.1. Cultivar sensitivity to mealybug. Mealybug sensitive cultivars Chardonnay Cabernet Sauvignon Cape Riesling Pinotage Shiraz Merlot Less sensitive cultivars Colombar Sauvignon blanc Semillon Hanepoot Chenin blanc

Before planting a new vineyard, the history of mealybug infestation in the area should be ascertained, especially if one of the more susceptible cultivars is being considered. If high average temperatures occur in a specific area, there is a bigger chance of mealybug problems. Favourable micro-climates, e.g. slopes that get excessive northerly sun, may also occur in certain blocks and encourage mealybug infestation. Top grafting existing grapevines to other cultivars Because of changing wine consumption trends worldwide, a need for alternative cultivars has developed and with this a tendency to top-work existing grapevines to other cultivars in order to save time and costs to establish new vineyards. Wine producers should realise that GLRaV (unspecified) is transmitted by grafting. They should also be aware of the danger attached to top-working existing vineyards to other cultivars, except if testing done on the existing vineyards show them to be free of leafroll and other harmful viruses (Schliefert 2001). Apart from leafroll virus, other diseases could also occur in existing vineyards, e.g. Shiraz disease for which there is no rapid detection method at present. The disease can be latent in existing grapevines and could then be transmitted to the cultivar onto which it is grafted. Management practices for leafroll infected vineyards South African recommendations suggest that the following practices could result in the production of higher quality grapes from leafroll infected vines or blocks: 1. Pruning severely for a lower yield (Lider et al. 1975). When only isolated cases of leafroll virus are reported in a block, these vines should be marked and then pruned so as to ensure a lower yield, or these vines could be harvested at a later stage 2. Reduction of stress to the vine from pea-size stage for more active growth later in the season 3. Irrigation and fertilisation for more active late season growth 4. Cutting back shoots in late November results in side-shoots of 30-40 cm growth in December 5. Removal of old leaves from certain red wine cultivars, e.g. Shiraz. Mealybug control in vineyards Mealybugs are the most important vectors of leafroll viruses, and therefore effective mealybug control is a prerequisite for containing the spread of leafroll infection. Mealybugs can survive on all live grapevine material on or under the soil surface. Before new vines are established on old vineyard soil, all live old vines and roots must be removed so that no mealybugs can survive. Studies suggest that P. ficus can be found on grapevine roots up to 60

73 cm below the surface, which has implications for control strategies (Carstens 2001). Old vines should be treated with herbicide directly after the last harvest and vines should not be removed for at least 6-8 weeks to ensure all roots are killed. Certified planting material should always be used to establish new vineyards. Implements coming from infested vineyards should be sprayed clean of mealybugs before entering uninfested vineyards. Try to work in newly planted, uninfested blocks first and then move on to the older, more infested blocks. Limit movement of workers in new plantings to a minimum. If workers need to move from infested to uninfested blocks, ensure that mealybugs or egg sacks are not carried on workers' clothing. Sampling for mealybugs and thresholds for action In the course of the season, thorough monitoring should be carried out, especially in November. An infection rate of 2-3% warrants chemical control. South African recommendations to determine the level of mealybug infection in vineyard blocks include: 1. Drafting a plan of the specific vineyard with a clear indication of each row and the number of sections per row 2. Selecting twenty sections with five grapevines each, proportionately spread throughout the block 3. Monitoring each of the five grapevines in each section in those areas where new growth is found 4. Recording the presence or absence of mealybug on each grapevine 5. The total number of infected grapevines will indicate the percentage of mealybug infection for that vineyard block. Mealybug is monitored by two methods: either regular vine inspections alone or by using pheromone traps (where the sex pheromones for the mealybugs in the vineyard are available) in conjunction with vine inspections. Physical monitoring of vines Vine inspections are recommended every two weeks and should commence from the beginning of October and are based on: 1. Drawing a plan of the vineyard, indicating the rows and number of plots in each row 2. Choosing twenty plots of five vines each evenly distributed through the block. Larger blocks are subdivided into units of 2 ha maximum, ensuring that areas with a history of mealybug infestation are included in the monitoring regime 3. Inspecting each of the five vines, especially the new growth around the crown of vines stems 4. Recording only the presence or absence of mealybug females (crawlers, nymphs and/or adult females) on each vine, i.e. is the vine infested or uninfested. Even if there is only one mealybug female on a vine, the vine is noted as infested 5. The presence of ants is also a good indicator that mealybug may be present. The total number of infested vines out of the hundred monitored, indicates the estimated percentage mealybug infestation for that block or unit of a block. Grapevine stem (or trunk) infestation rates above 2% justifies intervention with insecticides, or mass releases of commercially available natural enemies. In South Africa this includes the hymenopteran parasitoid Coccidoxenoides peregrinus. Monitoring with pheromone traps Pheromone traps are a new tool to aid the integrated management of P. ficus in vineyards (Walton et al. 2003). Pheromone traps are extremely sensitive in detecting low populations of P. ficus and can be used as a quarantine and early warning tool for the mealybug in vineyards.

74 Studies have shown that pheromone traps can be used to direct physical stem monitoring (the primary sampling method) to problem areas, making P. ficus monitoring less labour intensive and therefore more easily implemented in commercial vineyards. It is stressed that physical stem monitoring of vines should be carried out as a verification tool in blocks where sufficiently high P. ficus male counts are found and before any control actions are taken. South African recommendations for the monitoring of the grapevine mealybug (P. ficus) suggest that monitoring of mealybug males should be done every two weeks from October. The optimum distance of efficiency for a pheromone trap is 50 metres (i.e. covers an area of 1 hectare). The recommendations include: 1. Sub-dividing large vineyard blocks into units of approximately 1 ha each 2. Placing one trap or monitoring unit (yellow delta trap with sticky bottom and pheromone caps) per hectare at the beginning of October. Monitoring is continued until the end of March 3. Where more than one trap is placed in a block larger than 1 ha, traps in the sub-units should be 100 m apart so that they do not interfere with each other 4. Traps should be placed in or above the cordon region 5. The open ends of the trap should be left open and unimpeded (remove leaves or shoots if required) for unimpeded pheromone release 6. Sticky bases should be replaced every two weeks 7. Record two-weekly counts of each monitoring unit or trap 8. P. ficus pheromone caps should be changed after 8 weeks (= 2 months). (This may differ according to species pheromone characteristics.) It is worth stressing that these recommendations for using pheromone traps are unique for P. ficus in South Africa. They are indicative of future management requirements for pheromone trapping systems when introduced into New Zealand vineyards, but they will not be the same. Correlation between pheromone trap counts and vineyard infestation South African research has found a clear correlation between pheromone trap counts of P. ficus and percentage stem infestation. This work indicated that the action threshold of 2% P. ficus stem infestation (Walton 2003) will be reached at an adult male vine mealybug pheromone trap count of above 65 per biweekly trap count. During two seasons pheromone traps caught adult males before any physical signs of P. ficus infestation could be found. This suggests that pheromone traps are more sensitive in detecting the development of P. ficus populations in specific vineyards than physical monitoring of stems and can be an extremely useful detection tool in quarantine situations. It is important to verify high pheromone trap counts with physical P. ficus stem counts to prevent over-reaction due to perceived 'high infestations'. The only reliable monitoring method which has action thresholds attached to it is physical stem monitoring for mealybug populations. Experiments in Australia showed that dispersing male Pseudococcus longispinus were rarely caught in sticky traps compared with dispersing first instar nymphs, and that males only flew in very light winds (Barrass et al. 1994). Hence it is conceivable that pheromone traps for New Zealand species (when developed) will provide a good measure of resident populations, but an alternative trapping system may be required to measure airborne immigration. Threshold values for action Treatment thresholds in South Africa are based on the premise that 2 % mealybug infestation is equivalent to 65 males/trap over a period of 2 weeks. A. If a trap count is less than 65 males/trap over a 2-week period then control is not recommended.

75 B. If a trap count is more than 65 males/trap over a period of 2 weeks: 1. Vine inspections should be completed in the hectare or sub-unit of the block covered by that particular trap (20 plots/ha, 5 vines/plot) 2. If infestation exceeds 2%, control should be applied in that part of the block 3. If infestation in the block is less than 2%, but there are some heavily infested vines, then spot treatment is recommended to prevent infestation from spreading. C. If a trap registers high counts (45-64 per trap) twice in a row: 1. It is recommended that vine inspection should be done immediately (20 plots/ha, 5 vines/plot) 2. If infestation exceeds 2%, an insecticidal control is recommended. Two-weekly pheromone trapping should start in October and continue until before harvest. Pheromone trapping could continue on a monthly basis after this period in commercial blocks with a history of high P. ficus infestation. Pheromone trapping is recommended throughout the year at vine propagation and quarantine vineyard units. Traps should further be placed in areas where human movement takes place at regular intervals, such as near roads, end rows and packsheds. Traps should also be placed in new vineyards planted on soil where old established and P. ficus infested vineyards have been removed. Insecticidal control of mealybugs The key to effective chemical control of mealybug is timely action, but decision-making should be justified based on a mealybug monitoring programme. 1. Blocks or areas in blocks where high levels of mealybug infestation (2% and more) occurred during the previous season, must be treated during dormancy (after leaf drop and before budding). 2. Dormant sprays are recommended because in South Africa these are thought to have a lower impact on natural enemies. 3. Use of hand lances is recommended in South Africa to achieve thorough wetting through high volume application. Wet vines thoroughly (2 - 3 L spray volume per vine) using high application pressures. 4. If subsequent monitoring shows that infestation already exceeds 2% before the end of November, infested vines as well as the two adjacent vines should be treated, or spray only the infested spots in the block. 5. If outbreaks of more than 2% infestation occur later in the growing season, use a chemical with a short withholding period that breaks down rapidly, so that natural enemies have enough time during the remainder of the growing season to re-establish. 6. Sprays after harvest should preferably be limited to instances where infestation is so severe that vines lose their leaves prematurely and vines may not be able to ripen the canes properly or may die. 7. Postharvest sprays are only suggested if monitoring records indicate that the infestation early in the season did not exceed 2% and that the outbreak really only occurred later. These applications may be made with hand lances only. Spot applications must be made, unless monitoring indicates that the infestation is so widespread that spot treatments are not practical. 8. If 75% or more of mealybugs are parasitized after harvest, spraying is not required, since most mealybugs will be dying anyway. 9. Soil application of systemic products can be considered as an alternative to cover sprays. The key to effective chemical control of the mealybug is to take timely action. Treatments should be preceded by monitoring and the information gathered from this can be used to

76 decide whether control has to be exercised or not. If a high mealybug infection (of 5% or higher) had been reported in the previous season, dormant treatments should be conducted in the problem areas. These treatments should be done by thorough application (i.e. drenching) of vines using hand held spray lances. The safest insecticides to control mealybugs during the season are those with short residual life (i.e. those that break down within two weeks). The advantage is that the fast breakdown of the toxins enables the natural enemies to occupy the vineyard shortly after the insecticide has been administered. Since mealybug occurs in random spots in vineyards, an attempt should be made to give spot treatments in order to reduce the impact on the ecosystem. Chemical control done during the harvest season does not have the desired effect because the majority of vine mealybugs are parasitised. A large proportion of the remaining mealybugs die with the onset of winter. Mealybug populations are also protected against insecticides by means of wax and honeydew secretions. A recent survey in South Africa has shown that mealybug should be controlled as early as possible in the growth season. Some insecticides are only applied in the growing season, while others are used both in winter and the growing season. Dosages (per 100 L water) also differ, depending on the growth stage. Dormant treatments (before the new growth begins) should only be applied if more than 5% infestation occurs. This reduces hibernating mealybug populations to such an extent that natural enemies can control mealybug effectively the next season. It is also recommended that routine dormant spraying should be avoided (because of the possible development of insecticide resistance). Try to spray only the vine that has been marked and the two vines on either side with hand-held spray guns and a high pressure pump. Atomiser sprays do not give sufficient coverage. In South Africa the following chemical products and application methods are still being used at present (Nel et al, 1999, Vermeulen 1999): 1. Lorsban or Chlorpyrifos EC (480 g/L) at 200 ml/100 L is recommended twice, 14 days apart, before budding. After budding there are problems with phytotoxicity on young shoots and leaves. Chlorpyrifos at this dosage also suppresses ant species in South Africa (the Argentine and cocktail ants). 2. Profenofos EC (500 g/L) at 100 ml/100 L is also recommended twice, 14 days apart, before budding. Thorough drenching is recommended. 3. Tokuthion or Prothiofos EC (960 g/L) at 50 ml/100 L is recommended once just before buds begin to swell. A withholding period of 100 days is recommended. During the growing season the following chemical products and application methods are allowed in South Africa (Nel et al. 1999, Vermeulen 1999): 1. Losrban or Chlorpyrifos EC (480 g/L) is recommended at 75 ml/100 L from four weeks after budding up to 28 days before harvest. It is suggested that higher rates will result in phytotoxicity on leaves. 2. Nuvan or Dichlorvos EC (1000 g/L) is recommended at 75 ml/100 L up to 7 days before harvest. This treatment is only supplementary to winter treatments. 3. Rogor or Dimethoate EC (400 g/L) is recommended at 125 ml/100 L up to 28 days before harvest. 4. Formothion EC (250 g/L, 330 g/L) is recommended at 150 ml/100 L up to 10 days before harvest. 5. Supracide or Methidathion EC (420 g/L) is recommended at 50 ml/100 L up to 8 days before harvest. This treatment is prescribed as a late corrective one-off treatment. 6. Mevinphos SL (500 g/L) is recommended from 37.5-45 ml/100 L up to 7 days before harvest. Many of these products are broad-spectrum, neurotoxic pesticides that are increasingly unacceptable in New Zealands horticultural sector; some are no longer available and others

77 are targeted for reduction under NZ Winegrowers Sustainable Wine Production programme. Tokuthion and, to a lesser extent chlorpyrifos, are commonly used in New Zealand but use of the selective insecticide Applaud (buprofezin) has become more widespread, especially in many North Island vineyards affected by mealybugs. Postharvest spraying Postharvest spraying for mealybug control is not recommended in South Africa. During this period populations of natural enemies are at their highest and these are more susceptible to some of the insecticides (e.g. chlorpyrifos) than are the mealybugs. Spraying at this time also interferes with the potential for effective biological control next season. Insecticide resistance Insecticides have given good control of mealybug for many years in South Africa but widespread control failures suggest that populations resistant to insecticides are increasingly common (De Wet & Moores 2003). Recent research has indicated that both target-site and metabolic resistance mechanisms are present, in the form of insensitive acetylcholinesterase and enhanced esterase activity. The insensitive acetylcholinesterase mechanism gives resistance to organophosphate and carbamate insecticides and is very specific to certain insecticides in these groups. Enhanced esterase activity gives resistance to a broader spectrum of insecticides and different chemical groups as well. This should cause concern for growers, as many conventional pesticides will prove ineffective against these resistance mechanisms. A deeper understanding and characterization of these mechanisms will give researchers a clearer picture of the situation, and should allow an informed choice of chemical to ensure more sufficient control if these mechanisms are present in the population. Some populations of Pseudococcus viburni in New Zealand apple orchards are resistant to both Tokuthion and Lorsban (Charles et al. 1993), but (currently) not Applaud (P. Lo HortResearch, unpublished data). Claims of poor insecticidal control in some New Zealand (Hawkes Bay) vineyards also raises important questions over the species involved in these vineyards and their relative susceptibility to the primary insecticides Tokuthion and Applaud. There is no information on the susceptibility (or otherwise) for either citrophilus (P. caleolariae) or longtailed (P. longispinus) mealybugs - the most common species in North Island vineyards (Bell et al. 2005, Charles 1993). Degree-day models for timing mealybug insecticides Heat accumulation is widely used in Integrated Pest Management (IPM) programmes to predict the outbreak of pest populations and time control measures against susceptible life stages. It is expressed in degree-days (DD), and is determined by the rate of development of the insect at different temperatures. Information resulting from the use of DD models can be used as additional inputs in a pest management system for a key pest, including P. ficus, to predict the timing of subsequent generations and life stages (Walton 2004). The number of degree days required for P. ficus to complete one generation was 235 DD. In many cases this number of DD coincided with the visible generational increases of P. ficus in South African vineyards. The number of DD for the development of P. ficus accumulated rapidly from early October in all areas. This was also the period during which P. ficus populations increased rapidly and there also appears to be an indirect qualitative relationship between bunch infestation and cumulative DD. DD can be used as an additional tool for P. ficus management in vineyards. The completion of an estimated one P. ficus generation (235 DD) in a specific grape growing area should serve as a warning tool for pending P. ficus infestation. This benchmark (235 DD), together with monitoring with pheromone traps and physical monitoring could aid growers in the decision

78 processes for timely P. ficus management. The use of DD is still experimental in South Africa but 235 DD is used as an early warning system for P. ficus management (Walton 2004). Biological control of mealybug Biological control of P. ficus is encouraged because it can be very effective and can keep mealybug populations under the 1% level of infection under optimum circumstances. Natural enemies can effectively control the pest but, if absent, new colonies of the mealybug may establish and spread. Optimum circumstances for biological control include the absence of dust and ants, and sensible chemical control. Beneficial insects are advantaged by ground cover plants that flower early in the season, and a flowering period that lasts as long as possible through the season, because many natural enemies use pollen as an alternative food source. The removal of vine leaves early in the season also contributes to more effective biological (and chemical) control. Biological control by means of augmentative releases of commercially available natural enemies can be applied if mealybug populations are low enough (infestation level <2%). Biological control alone is not usually effective under mealybug outbreak conditions. Several species of both predators and parasitoids of mealybugs are common in New Zealand vineyards too (Charles 1993, Clearwater 2002). Management of ant populations Ant control is of utmost importance in South African vineyards. Biological control agents become totally ineffective without ant control, and without this, dramatic increases in the mealybug population are seen. Cultivation practices Plastic sheeting can, in some cases, provide an ideal environment for nests of some ant species, particularly in winter, as they increase soil temperature and moisture immediately beneath the sheets. If large numbers of actively moving ants are noticed under the sheets, they should be removed as soon as possible. Weed control Keep weeds under control throughout the year. Tall weeds provide alternate pathways into the vine canopy for the ants, which should be avoided during the growing season. Weeds also harbour other honeydew-excreting insects, such as aphids, which provide an additional food source for ants. The use of a cover crop system can help in controlling weeds, while having other beneficial effects for the natural enemies of the mealybug. Many broad-leaf weeds act as hosts for mealybugs, the honeydew of which provide a food source for ants. The choice of an appropriate cover crop is important to eliminate the risk of it becoming source of mealybug infestation within the vineyard. Mealybug control Honeydew excreted from mealybugs provides the primary food source for ants in vineyards. The higher the mealybug population, the more difficult it is to control ants. Monitoring of ants should start early in the growing season Monitoring and identification and should continue periodically until ants are first noticed moving into the vine canopy. Monitoring for ants in South African vineyards is done in conjunction with mealybug monitoring as follows: 1. Monitoring twenty sections with five grapevines each, proportionately spread throughout the block (up to 2 ha) 2. Recording the presence or absence of ants in the vine canopy of each grapevine 3. The total number of infested grapevines gives an indication of the percentage of ant infestation for that vineyard block

79 4. A preliminary threshold of 25% is used to justify insecticidal control. Three types of ants that are disruptive to mealybug biocontrol are commonly found in vineyards. They are Argentine ant, pugnacious ants and cocktail ant. They can be distinguished by size, colour and nesting site. Of these, only Argentine ant is found in New Zealand. Worker ants are 3 mm, light brown, and all of equal size. Nest entrances are mainly in the soil and inconspicuous. At least 8 species of ants have been identified in New Zealand vineyards (Lester et al. 2003) and the Argentine ant is now known to occur in some major North Island winegrape production regions. Relatively little is known of the impact of ant species on mealybug biological control in New Zealand (Charles 1993). The impact of Argentine ant is only likely to increase disruption to existing mealybug biological control activity in vineyards. Insecticidal control of ants Effective ant control is a prerequisite in South Africa for mealybug control because ants protect mealybugs against their natural enemies. Ueckermann (1998) found the most effective means of ant control to be circular spraying around the trunk. This method of control is also more environmentally friendly than chemical treatments applied to the soil. The purpose of this application method is to keep ants out of the vines, but still allow them on the soil surface. Here ants may act as predators of beetle, fruit fly and moth larvae and pupae. At the moment chlorpyrifos EC is the only registered pesticide against ants (Nel et al., 1999). Chlorpyrifos has also been tested at 41 ml/L (not a registered concentration) with reasonably good results (Ueckermann, 1998). Trunk treatments for ants should be dependent on ant activity, which usually increases from October onwards if mealybugs are present. In South Africa insecticidal stem treatments are recommended, as these Soil nesting ants have the least impact on the natural enemies of the mealybug. Insecticide should be applied as soon as ants move into the vine canopy (October/November). Earlier applications are considered to be less effective. For pugnacious ant, alphacypermethrin SC is registered in South Africa and use is recommended at 200 ml/10 L water. For Argentine ant, alphacypermethrin SC at 100 ml/10 L, Chlorpyrifos 75 WG at 250 g/10L water and SuSCon Blue Ribbon (chlopyrifos granules) are registered and recommended. For insecticidal stem treatments, 50 ml of the spray mixture must be applied evenly around the vine stems above irrigation lines. One application is considered sufficient for the entire growing season, except where high ant pressure exists. Continued monitoring is recommended to assess efficacy of treatments. Insecticidal control immediately after harvest is not recommended. To control cocktail ants in South Africa, a dormant application of Vine nesting ants chlorpyrifos EC at 400 mL/100 L water is recommended as a full cover (canopy) application. Application method Trunk barriers are applied with a knapsack spray pump with ring-spray attachment for good coverage around stems. Treatment of all trellis poles is also recommended. Spot treatments for ants are not recommended as ants can move along the wires. Treating entire rows where ant infestations are present is recommended. Weeds must be kept low to prevent alternative access into vines. Weed control within vineyards Weeds are the hosts to a variety of mealybug species in both S. Africa and New Zealand. Weeds can also act as a bridge for ants to reach the grapevine canopy and hence can be detrimental to biological control. Mealybugs occur on the roots of the following weeds in South Africa (Fourie et al. 1996):

Common blackjack (Bidens pilosa) Khaki weed (Tagetes minuta)

80

Small mallow (Malva parviflora) Flax-leaf fleabane (Conyza bonariensis) Black nightshade (Solanum nigrum) Thornapple (Datura stramonium) Sowthistle (Sonchus oleraceus) Storksbill (Erodium moshantum) Fat hen (Chenopodium album).

A correlation has been found between the occurrence of these weeds and mealybug problems in vines. The best method of control is the planting of cover crops as recommended by Fourie et al. (1997). Long flowering cover crops that do not host mealybug may reduce ant problems, may help to reduce the forming of dust, may serve as supplementary nutrition for natural enemies and may bind nitrogen. Weeds may also be controlled physically (shrub beaters) and chemically (herbicides) (Fourie et al. 1996). Weeds have to be controlled from early in the season, since they act as access routes to the vine for ants and do not contribute much to the quality of the soil. Ant control is impossible if weeds grow into the vines.

81

SECTION 5: SUMMARY AND RECOMMENDATIONS.

SUMMARY AND INTEGRATION OF KEY FINDINGS It became quite clear from the review that although the symptoms of grapevine leafroll disease have been recognized in New Zealand (and elsewhere) for more than 100 years, an understanding of the causes of the disease came slowly. Recognition that the disease was caused by a virus provided some early focus on the type of problem, even though initial emphasis on roguing (with no knowledge of a vector) did little to achieve control. The small size of the New Zealand wine industry in the 1960s to early 1980s, centred on American hybrid cultivars, provided little impetus to develop further understanding. The first implication of mealybugs as vectors of grapevine viruses overseas was followed by further laboratory experiments to identify the viruses they could transmit and to measure transmission efficiency. These studies were hampered substantially by a lack of tools to (a) accurately identify the viruses themselves, and (b) measure low virus titres in both vines and mealybugs. Rapid advances in molecular biology and improved ELISA techniques then provided some rapid advances, but it is really only in the past 2-5 years that a reasonably clear understanding of the identity of the causal agents and vectors of grapevine leafroll disease have been elucidated. Even so, our understanding remains piecemeal. Molecular biologists have worked on virus taxonomy; virus epidemiologists have measured the spread of the disease in vineyards; entomologists have confirmed transmission in the laboratory; and grapevine physiologists have measured the effects of the disease on grapevines. But nowhere, in New Zealand or internationally, have we found an interdisciplinary research programme on the inter-relationships between the virus, vector and vine (V3). It is now clear that the disease known as grapevine leafroll disease in New Zealand is predominantly caused by the ampelovirus GLRaV-3, which is vectored between plants by three species of mealybugs. Other closteroviruses and insects may also be involved, at least sometimes. However, despite this basic knowledge of the mechanics of GLRaV-3 transmission and hence spread of grapevine leafroll disease, many questions with a direct bearing on the practical issue of how to manage the disease in New Zealands vineyards remain to be answered. Such fundamental questions include: What are the effects of GLRaV-3 on vine yield and wine quality in different cultivars and regions? How does the virus titre ebb and flow during the season? Is an observed plant resistance due to a particular strain of virus, a different species/cultivar/clone of vine, or the ability of mealybug to transmit the virus from one plant to another? How can we best measure the transmission ecology of different mealybugs in different regions on different cultivars/clones? What is the best or most appropriate strategy for vector control? How do ants, natural enemies and ground cover influence mealybug populations and hence the spread of GLRaV-3?

The search of the global scientific literature shows that there are no great advances elsewhere in the world that New Zealand can immediately appropriate. In fact, the combination of mealybug species, climate and vineyard environment that comprises the New Zealand wine industry is unique. New Zealand appears to be one of the very few countries that is warm enough for mealybugs to regularly reach high population densities, yet not warm enough to provide sufficient additional ripening to compensate for the effects of GLRaV-3 on grape quality and harvest.

82 It seems that the way to provide successful solutions to grapevine leafroll disease in New Zealand is to establish an holistic, multi-disciplinary approach, as adopted in South Africa. This is particularly so if the goal is to provide long-term management over the course of decades, rather than simply from year to year. The virus vine vector relationship (V3) should be viewed as a single entity, each component of which is interdependent on the other two. Research to understand V3 should be tackled by a multi-disciplinary team of plant virologists, entomologists, vine physiologists, pest controllers, vineyard managers, grapevine breeders/improvers and winemakers. Even though there would be many (some quite narrowly focussed) research projects, this approach would provide the framework within which the many participants would be able to view all aspects of the V3 relationships. This would inevitably lead to better priority setting and design of research projects, better communication and understanding of what individual projects were contributing, better and more innovative projects incorporating appropriate ideas from overseas, and ultimately, better and more robust long-term solutions to the grapevine leafroll disease problem in New Zealand. The design of V3 should take into account some of the key findings from this review: Section 1: GLRaV-3 the virus GLRaV-3 has probably been in New Zealand for more than 100 years, although it was first positively documented in 1964. Leafroll symptoms were seen in European Vitis vinifera, but not in US rootstock species. Virus-free vines resulting from thermotherapy were released by DSIR during the early 1970s. Vine pathology was first measured in the 1970s. Recognition of the similarity of disease symptoms to potassium deficiency provided an indication of the mode of pathology of the virus. Leafroll disease was first associated with closterovirus-like particles in the mid-1980s. Monoclonal antibodies and improved techniques allowed sensitive detection of GLRaV-3 from 2000. By 2003, 55 viruses known to infect grapevines had been identified globally. RT-PCR is the most sensitive, but difficult, technique for virus identification. Recent advances in virus taxonomy (based on molecular structure) have clarified the relationships between different grapevine viruses: GLRaV-3 is placed in the Ampelovirus group. Different groups of viruses are characterised, in part, by their transmission by different groups of insects. Ampelovirus is transmitted by mealybugs and scale insects. The existence of strains of GLRaV-3 with different pathogenicity has been postulated, but not yet demonstrated. Section 2: impact of GLRaV-3 on vine growth and productivity. Early work in New Zealand supported overseas data that leafroll disease adversely affected vine growth, yield, fruit colour and sugar content. Degeneration of phloem cells in leaves, stems, and fruit petioles has been reported in GLRaV-3 infected vines. This is usually accompanied by an accumulation of starch in leaves, which may be the feedback mechanism whereby photosynthetic activities are shut down in infected leaves. GLRaV-3 can depress photosynthesis by 25-65% (depending on cultivar and environment), which is likely to directly affect all aspects of growth and cropping. The appearance of disease symptoms depends on many factors, but GLRaV-3 depresses photosynthesis even in vines that do not reveal visual symptoms. Foliar nutrients may alleviate the virus effects on photosynthesis. The effect of GLRaV-3 on root growth of infected vines is unclear.

83 It is unclear whether GLRaV-3 plays a role in graft incompatibility, although other viruses have been implicated. Viruses (including GLRaV-3) reduce weight and girth of canes, but large variation caused by environmental conditions makes the impact difficult to quantify. Increased vigour of virus-free vines may necessitate changes in vine management. A GLRaV-3 plus fanleaf virus infection reduced number, area and mineral content of leaves. GLRaV-3 infection may affect seasonal phenology through delayed budbreak, delayed flowering and delayed berry maturity, although these effects are not well quantified especially in New Zealand. GLRaV-3 infection may increase vine longevity due to reduced lifetime productivity. Evidence from around the world suggests that where present, GLRaV-3, either on its own or in combination with other virus and virus-like diseases, severely reduces grape yield among a wide range of white and red varieties. Similarly, among white and red varieties, evidence from a majority of studies suggests that in leafroll infected vines fruit sugar levels at harvest are lower, titratable acidity is higher and berry anthocyanin levels are reduced, compared with the juice from healthy vines. In almost half the studies where data were presented, fruit pH differed significantly from that found in healthy vines. Wine quality from GLRaV-3 infected vines is reduced. The effects of GLRaV-3 on quantitative and qualitative parameters of vine performance can impose severe limitations on wine production and quality. Section 3: GLRaV-3 vectors and ecology of transmission Three (3) species of exotic mealybugs feed on leaves, shoots, fruit, and occasionally roots of grapevines in New Zealand. None of the c.120 native species of mealybugs are known to feed on grapes. Twenty three (23) additional species worldwide are also known to feed on grapes. Seventeen (ex 26) species are considered to be pests of winegrapes. At least 10 of these are known plant virus vectors, and at least 8 transmit GLRaV-3. Mealybugs are small, hemimetabolous, phloem feeding insects. They excrete large quantities of honeydew (on which grow sooty mould fungi). New Zealand species develop through 2-3 generations a year. Numbers increase greatly between spring and autumn. Mealybugs contain endosymbiotic bacteria, which may synthesise amino-acids essential for mealybug nutrition, and may also facilitate virus transmission. Transmission experiments indicate that mealybug transmission of plant viruses is usually, if not always, semi-persistent. Both mealybug and virus are required to induce mealybug wilt of pineapple (MWP) symptoms. MWP control strategies are either through the plant (meristem propagation of virus-free plants or development of transgenic plants through gene-silencing mechanisms) or the vector (especially through increased biological control). Presence of Argentine ant may be positively correlated with an increased rate of MWP virus spread. Unauthorised movement of cassava germplasm within and between continents has led to spread of pests. Control of cassava mealybugs (vector of ACMV) has been achieved by combinations of host plant resistance, biological control and cultural practices. Mealybug parasitoids were

84 more effective if plants were selected for strong antixenosis but low antibiotic characteristics. Environmental conditions (e.g. wet v. dry), mono- v. polycultures and IPM methods all can have significant impacts on mealybugs and virus control in different crops. Plant viruses are typically able to be transmitted by several species of mealybugs, with differing efficiency. Virus spread can occur through airborne dispersal of young crawlers. Virus infections in some cocoa plants can remain latent for >2 years, and some cultivars/hybrids are more resistant to disease than others. Mild-strain cross protection is also developed in cacao. GFLV is transmitted only by a soil nematode. Development of resistant rootstocks using genes resistant to nepoviruses is under investigation. GVA and GVB (trichoviruses) are transmitted by the same species of mealybugs that transmits GLRaV-3. P. longispinus transmits GLRaV-5, but not -1 or -2. Transmission of GVA appeared to depend on the presence of GLRaV-2. Mealybug vectors of plant viruses typically acquire virus in periods varying from 0.25-12 h and retain virus for 12 h - 5 days. Virus may persist in a mealybug after post-acquisition moulting; but mealybugs may quickly lose the ability to transmit virus after transfer. Transmission efficiency is variable (typically 15-25%). Efficiency may be greater after post-acquisition feeding than fasting. Temperature may be an important variable in transmission efficiency.

Section 4: Managing GLRaV-3 infection and spread within the vineyard The international distribution of graft transmissible grapevine viruses is largely due to early indiscriminate exchange of graftwood and rootstocks. There has been little international interest in measuring or limiting the extent of the virus in national vineyards at least until recently. There is virtually no literature on management strategies to reduce or eliminate the spread of GLRaV-3, other than by reliance on certificated planting material. In some countries (e.g. USA, Spain) the role of mealybugs as vectors of GLRaV-3 has only recently been realised, and control strategies are not yet widely in place. By far the most advanced vineyard management systems in place for managing the spread of GLRaV-3 are in South Africa. Even here the plans are in technical guides, rather than scientific literature. Although the South African management strategies are useful in principle to New Zealand, fundamental differences mean that they are not entirely transferable. Key differences are: different mealybug vectors; heavy use of broad-spectrum insecticides; influence of several species of disruptive ants. In South Africa, the most common pattern of GLRaV-3 spread occurs initially within rows between immediately adjacent infected vines before moving across rows and vineyards through various combinations of: mealybug dispersal and vineyard machinery and personnel. Roguing may reduce the rate of spread of GLRaV-3 where the incidence of primary and secondary infected plants is low and where latent infection is not excessive, but it is not expected to be effective under all conditions. Latent infection in asymptomatic vines is managed by spraying targeted insecticides. Edge-effects can indicate the spread of disease from other vineyards.

85 Management strategies include planting large, isolated vineyards with shelterbelts, upwind from existing virus sources; and working daily in youngest, least infected blocks first. Good hygiene, especially when replanting an old infected vineyard, is important to minimise point sources of infection. Some grape cultivars are more susceptible to mealybugs than others. Beware of transmitting GLRaV-3 through top-grafting vines to new cultivars. Mealybug control is the key to GLRaV-3 control, especially when vineyards are susceptible to sustained mealybug immigration. Mealybugs are monitored by a combination of physical examination of vines and pheromone traps (for P. ficus only). Pheromones are new tools that help to reduce the labour intensive monitoring and determine the need for control actions. Pesticide application are recommended to targeted vines during dormancy, to minimise negative effects on natural enemies. Common insecticides in S. Africa are either not registered in New Zealand, or are not compatible with SWGNZ. There are widespread control failures due to insecticide resistance. Biological control of P. ficus can provide effective control. Ant control (including Argentine ant) is crucial for mealybug control in S. Africa. Appropriate cover crop and weed control strategies are an important part of managing ants, mealybugs, natural enemies and hence the spread of GLRaV-3.

Consideration of these findings allows some obvious research topics to be suggested (and see table ES1): Section 1 Possible research topics Determine why different cultivars/species of grapevine infected with GLRaV-3 exhibit a range of symptoms (from mild to very severe). Investigate whether this is due to: o Genetic tolerance within the plant to the presence of GLRaV-3 o Presence of more than one type of virus, or strain of GLRaV-3 o Presence of other viruses or infectious agents o Presence of a number of strains of GLRaV-3 differing in severity Determine whether different grapevine viruses in New Zealand are additive with respect to symptoms. Section 2 Possible research topics Identify and develop management tools for GLRaV-3 infected vines to achieve high must quality and yield and to manage timing of harvest (e.g. foliar fertilisation, crop loading) Identify and develop management tools for GLRaV-3 free (increased vigour) vines in the New Zealand environment (nutrition, summer pruning, botrytis control, etc.) to support virus free plantings Determine the susceptibility of rootstocks to GLRaV-3 and the effect of the virus on graft compatibility and vine longevity Identify GLRaV-3 tolerant/resistant Vitis species, identify method of tolerance/resistance (can it be translocated from rootstock to scion? identify gene level?), breed GLRaV-3 tolerant/resistant rootstocks/vines Quantify the impacts of GLRaV-3 infection on timing of budbreak, flowering and berry maturity Quantify the long-term impacts of GLRaV-3 on vine yield in New Zealand cultivars and regions Quantify the effects in New Zealand of GLRaV-3 on berry quality (especially sugars, titratable acidity and anthocyanins) and resulting wine (particularly sensory attributes).

86 Section 3 Possible research topics Define the ecological relationships between mealybugs (especially P. longispinus and P. calceolariae) in different regions and their grapevine cultivar preferences and efficiency of GLRaV-3 transmission (i.e. transmission ecology and V3 interactions) Determine the role of mealybug endosymbionts in virus transmission, and seek ways to disrupt the endosymbiosis and hence improve GLRaV-3 management. Section 4 Possible research topics Design and implement GLRaV-3 management strategies based on New Zealand fauna, industry practices (especially SWNZ) and environments Develop new monitoring strategies using pheromone traps as key resources to rapidly locate small patches of mealybugs in large or small vineyards Develop new biological control strategies through the introduction of new natural enemies or manipulation of existing species (including options of augmentative control) Determine the impact of Argentine ant (and other ant species in New Zealand vineyards) on mealybug control and spread of GLRaV-3 in New Zealand Develop new mealybug control options using safe chemistry or non-chemical means Quantify dispersal behaviour and distances travelled by mealybug males and crawlers, and the benefits from shelterbelts or other physical barriers Design the best (most appropriate) vineyard plant biodiversity for maximum disruption to mealybugs and hence spread of GLRaV-3 in New Zealand.

87

REFERENCES
Adu Ampomah, Y., Adomako, B., Owusu, G.K., Ollennu, L.A.A., Bekele, F., End, M. and Eskes, A. (1999). Breeding for resistance to the cocoa swollen shoot virus in Ghana. Proceedings of the International Workshop on the Contribution of Disease Resistance to Cocoa Variety Improvement, Salvador, Bahia, Brazil, 24 26 November, 1996. 1999, 173 179; 20 ref., International Group for Genetic Improvement of Cocoa University of Reading; Reading; UK. Akinlosotu, T.A. (1983). "Studies on the control of the cassava mealybug (Phenacoccus manihoti) and the green spider mite (Monoychellus tanajoa) in south-western Nigeria." Journal of Root Crops 9(1-2): 33-43. Antcliffe A.J., Woodham R.C. and Cellier K.M. (1979). A comparison of 182 Sultana clones selected for yield. Aust J. Agric Res 30:1111-1122. Ayllon M.A., Lopez C., Nav.-Castillo J., Garnsey S.M., Guerri J., Flores R.and Morino P. (2002). Polymorphism in the 5 terminal region of Citrus Tristeza virus (CTV) RNA: Incidence of three sequence types in isolates of different origin and pathogenicity. Arch. Virol. 146:27-40. Barrass, I.C., Jerie, P., Ward, S.A. 1994. Aerial dispersal of first- and second instar longtailed mealybug, Pseudococcus longispinus (Targioni Tozzetti) (Pseudococcidae: Hemiptera). Australian Journal of Experimental Agriculture 34: 1205-8. Baumann, P. (2005). Biology of bacteriocyte-associated endosymbionts of plant sap-sucking insects. Annual Review of Microbiology 59: 155pp Baumann, L. and P. Baumann (2005). "Cospeciation Between the primary endosymbionts of mealybugs and their hosts." Current Microbiology 50(2): 84-87. Baumann, P., Moran, N.A. and Baumann L. (2000). Bacteriocyte-associated endosymbionts of insects. New York, Springer http://link.springer.de/link/service/books/10125. Baumann, L., Thao, M.L., Hess, J.M., Johnson, M.W. and Baumann, P. (2002). "The genetic properties of the primary endosymbionts of mealybugs differ from those of other endosymbionts of plant sap-sucking insects." Applied and Environmental Microbiology 68(7): 3198-3205. Bell, V.A., Walker, J.T.S., Charles, J.G., Cole, L.M., Suckling, D.M. and Miller, J.G. (2005). Mealybug monitoring and the use ofa sex pheromone in Hawkes Bay vineyards. HortResearch Client Report No. 18312. Belli, G., Fortusini, A., Casati, P., Belli, L., Bianco, P. A. and Prati, S. (1994). Transmission of a grapevine leafroll associated closterovirus by the scale insect Pulvinaria vitis L. Rivista de Patologia Vegetale 4: 105-108. Bellotti, A.C., Braun, A.R., Arias, B.,Castillo, J.A.,Guerrero, J.M.,Thresh, J.M.,Otim Nape, G.W.,Fabres, G.,Yaninek, Y.S. and Adipala, E. (1994). "Origin and management of Neotropical cassava arthropod pests." African Crop Science Journal 2(4): 407-417. Belton, M., Edwards F. and Bonfiglioli, R. (2006). Improving diagnostics for leafroll virus type 3 (GLRaV-3). 6th Int Cool Climate Symposium for Viticulture and Oenology, Christchurch, NZ. Ben Dov, Y. (1994). A systematic catalogue of the mealybugs of the world (Insecta: Homoptera: Coccoidea: Pseudococcidae and Putoidae) with data on geographical distribution, host plants, biology and economic importance. 1994, 686 pp.; 83 pp. of ref. Andover; UK, Intercept Limited. Bento, J.M.S., de Moraes, G.J., Bellotti, A.C., Castillo, J.A., Warumby, J.F.Lapointe, S.L. and de Moraes, G.J. (1999). "Introduction of parasitoids for the control of the cassava mealybug Phenacoccus herreni (Hemiptera: Pseudococcidae) in north-eastern Brazil." Bulletin of Entomological Research 89(5): 403-410.

88 Bento, J.M.S., de Moraes, G J., Matos, A.P., de Bellotti, A.C., de Moraes, G.J. and de Matos, A.P. (2000). "Classical biological control of the mealybug Phenacoccus herreni (Hemiptera : Pseudococcidae) in northeastern Brazil." Environmental Entomology 29(2): 355-359. Bertamini M, Nedunchezhian N, Tomasi F and Grando M S (2002). Phytoplasma [Stolbursubgroup (Bois Noir-BN)] infection inhibits photosynthetic pigments, ribulose-1,5biphosphate carboxylase and photosynthetic activities in field grown grapevine (Vitis vinifera L. cv. Chardonnay) leaves. Physiological and Molecular Plant Pathology 61: 357-366. Bertamini, M., Muthuchelian, K. and Nedunchezhian, N. (2004). "Effect of grapevine leafroll on the photosynthesis of field grown grapevine plants (Vitis vinifera L. Cv. Lagrein)." Journal of Phytopathology 152(3): 145-152. Bigger, M. (1972). "Recent work on the mealybug vectors of cocoa swollen shoot disease in Ghana." Pans 18(1): 61-70. Bigger, M. (1981). "The relative abundance of the mealybug vectors (Hemiptera: Coccidae and Pseudococcidae) of cocoa swollen shoot disease in Ghana." Bulletin of Entomological Research 71(3): 435-448. Bigger, M. and R. Kumar (1975). Recent work on the mealybug vectors on cocoa swollen shoot virus. Proceedings of the 4th Conference of West African Cocoa Entomologists, Zoology Department, University of Ghana, Legon, Ghana, 9th 13th December 1974. 1975, recd. 1977, 62 66; 22 ref. Legon; Ghana, Zoology Department University of Ghana. Bonfiglioli, R., Hoskins, N. and Edwards, F. (2002). "Grapevine leafroll virus type 3 spreading in New Zealand." Australian & New Zealand Grapegrower and Winemaker 457: 58-61. Bonfiglioli, R., Hoskins, N. and Edwards, F. (2001) Grapevine leafroll viruses in New Zealand Viticulture Fact Sheet and research update. Riversun Technical Bulletin 5; 8pp Bonfiglioli, R. and Stewart, D. (2005). Learning about leafroll 3. New Zealand Winegrower 9 (3): 88-90. Borgo, M. (1991). "Effect of grapevine leafroll virus on some production parameters." Rivista di Viticoltura e di Enologia 44(2): 21-30. Borgo, M. and Angelini, E. (2002). "Influence of grapevine leafroll (GLRaV3) on Merlot cv. grape production." Bulletin de l'O.I.V. 75(859-860): 611-622. Borgo, M., Angelini, E. and Flamini, R. (2003). "Effects of grapevine leafrool associated virus 3 on main characteristics of three vineyards." L'enologo 3: 99-110. Borgo, M., Calo, A. and Bonotto, A. (1992). "Study on the relationship between virus diseases and productive features: Preliminary results concerning the elimination of leafroll and fleck." Rivista di Viticoltura e di Enologia 45(1): 3-10. Borgo, M. and Michielini, C. (2000). "Natural spread of grapevine leafroll on varieties and biotypes of Vitis vinifera." Rivista di Viticoltura e di Enologia 53(4): 3-13. Boscia, D., Greif, C., Gugerli, P., Martelli, G.P., Walter, B. and Gonsalves, D. (1995). "Nomenclature of grapevine leafroll-associated putative closteroviruses." Vitis 34(3): 171-175. Boscia, D., Savino, V., Minafra, A., Namba, S., Elicio, V., Castellano, M.A., Gonsalves, D. and Martelli, G.P. (1993). "Properties of a filamentous virus isolated from grapevines affected by corky bark." Arch. Virol. 130(1/2): 109-120. Bovey,R., Rochaix, M. et Simon, J.L. 1975. Rsultats et perspectives davenir de la selection sanitare de la vigne. Review of Suisse Viticulture, Arboriculture and Horticulture. 7 161-166. Bragato R. 1902. Report to the New Zealand Department of Agriculture: 451-5. Bragato R. 1903. Report to the New Zealand Department of Agriculture: 451-61.

89 Brew, A.H. (1992). "Towards the use of pathogenic micro-organisms in the control of some cocoa insect pests in Ghana." Cocoa Growers' Bulletin(45): 26-30. Brunt A.A. (1970). Cacao swollen shoot virus. CMI/AAD desriptions of plant viruses No 10. Cabaleiro, C., Pieiro, A. and Segura, A. (1997). Photosynthesis in grapevines infected with leafroll virus (GLRaV-3). 12th Meeting of the international council for the study of viruses and virus-like diseases of the grapevine (ICVG), Lisbon, Portugal: 153-154. Cabaleiro, C. and Segura, A. (1996). "Effect of grapevine leafroll-associated virus-3 in a commercial cv. Albarino vineyard." Investigacion Agraria: Produccion y Proteccion Vegetales 11(3): 451-463. Cabaleiro, C. and Segura, A. (1997a). "Some characteristics of the transmission of grapevine leafroll associated virus 3 by Planococcus citri Risso." European Journal of Plant Pathology 103(4): 373-378. Cabaleiro, C., Segura, A. and Garcia-Berrios, J.J. (1999). "Effects of grapevine leafrollassociated virus 3 on the physiology and must of Vitis vinifera L. cv. Albarino following contamination in the field." American Journal of Enology and Viticulture 50(1): 40-44. Cabaleiro, C. and Segura, S. (1997b). "Field transmission of grapevine leafroll associated virus 3 (GLRaV-3) by the mealybug Planococcus citri." Plant Disease 81(3): 283-287. Calatayud, P.A., Boher, B., Nicole, M. and Geiger, J.P. (1996). "Interactions between cassava mealybug and cassava: cytochemical aspects of plant cell wall modifications." Entomologia Experimentalis et Applicata 80(1): 242-245. Campbell, C.A.M. and Kumar, R (1975). The distribution of mealybug vectors of CSSV within trees. Proceedings of the 4th Conference of West African Cocoa Entomologists, Zoology Department, University of Ghana, Legon, Ghana, 9th 13th December 1974. 1975, recd. 1977, 67 71; 6 fig. Legon; Ghana, Zoology Department University of Ghana. Carstens, R. (2001). Biologie en beheerstrategie in Wynland. Article Wingerdwitluis. March 2001: 75-78. Carter, W. (1973). "Mealybug wilt of pineapple." FAO Plant Protection Bulletin 21(5): 115. Castellano M.A., Martelli G.P. and Savino 1983 Virus-like particles and ulstrastructural modifications in the phloem of leafroll affected grapevines. Vitis 22:23-29. Chabbouh N., Martelli G.P., Savino V., Greco N., and Lafortezza R. (1993). Potato virus X in Tunisian grapevines 1993 Vitis 32:165-169. Chamberlain E. E. 1967. Leafroll virus in the grapevines. Wine Review 4(3): 29-32. Chamberlain E.E., Over de Linden A.J. and Berrysmith F. (1970). Virus disease of grapevines in New Zealand. New Zealand Journal of Agricultural Research 13: 338-358. Charles, J.G. (1979). The ecology and control of mealybugs in Auckland vineyards. MSc Thesis, university of Auckland: 110pp. Charles, J.G. (1981). Distribution and life-history of the long-tailed mealybug, Pseudococcus longispinus (Homoptera: Pseudococcidae) in Auckland vineyards. New Zealand Journal of Zoology 8: 285-293. Charles, J.G. (1982a). Economic damage and preliminary economic thresholds for mealybugs (Pseudococcus longispinus T.-T., Homoptera: Pseudococcidae) in Auckland vineyards. NZ. J.Agric.Res. 25(3): 415-420. Charles, J.G. (1982b). Commercial control of mealybugs in apple orchards. Orchardist of New Zealand 55:298-299 Charles, J.G. (1989). Pseudococcidae, mealybugs (Homoptera). pp223-236. In: Cameron, P.J. et al. (eds): A review of Biological Control of invertebrate pests and weeds in New Zealand 1874 -1987. Tech. Comm. CAB International, UK. 424pp. Charles, J. G. (1993). "A survey of mealybugs and their natural enemies in horticultural crops in North Island, New Zealand, with implications for biological control." Biocontrol Sci. Technol. 3(4): 405-418.

90 Charles, J.G., Cole, L., Mundy, D. and Walker, J. (2005). Soft scales in New Zealand vineyards. Report to Winegrowers of New Zealand Ltd. February 2005. HortResearch Client Report No. 15419: 4pp. Charles, J.G.., Froud, K.J. and Beck, D. (1998). Limiting the spread of Grapevine Leafroll Disease (GLR). Report to Winegrowers of New Zealand. July 1998. HortResearch Client Report No. 98/164. Charles, J.G., Froud, K.J. and Henderson, R. C. (2000). "Morphological variation and mating compatibility within the mealybugs Pseudococcus calceolariae and P. similans (Hemiptera: Pseudococcidae), and a new synonymy." Systematic Entomology 25(3): 285-294. Charles, J.G.., Petersen, C. and Froud, K.J. (1997). Limiting the spread of grapevine leafroll disease (GLR). Report to Winegrowers of New Zealand Inc. July 1997. HortResearch Client Report No. 97/88. Charles, J.G.. and Walker, J.T.S. (1981). Mealybug control in apples. Orchardist of New Zealand 54: 252-253. Charles, J.G., Walker, J.T.S. and White, V. (1993). Resistance to chlorpyrifos in the mealybugs Pseudococcus affinis and P. longispinus in Hawkes Bay and Waikato pipfruit orchards. Proceedings 46th Plant Protection Conference: 120-125. Clearwater, J.R. 2002. The Ecology of Mealybugs in Gisborne Vineyards. Report to NZ Winegrowers. 9pp. Clingeleffer, P.R. and Krake, L.R. (1992). "Responses of Cabernet franc grapevines to minimal pruning and virus infection." American Journal of Enology and Viticulture 43: 31-37. Cohen D. and van den Brink R. (2005). Detection of leafroll virus in newly infected grapevines. HortResearch Client report No. 12673. Final report to Winegrowers of New Zealand. Cohen, D. and van den Brink R. (2003). Leafroll virus movement in grapevines. HortResearch Client Report 11871 for Winegrowers of New Zealand. Cohen, D. and van den Brink R., (2000). Improved sensitivity for ELISA detection of GLRaV-1 and GLRaV-3. Extended Abstracts 13th ICVG Conference, Adelaide. Pp 146-147. Cohen D., van den Brink R.C., MacDiarmid R.M., Beck D.L. and Forster R.L.S. 2000. Resistance to Tamarillo mosaic virus in transgenic tamarillos and expression of the transgenes in F1 progeny. Acta Horticulturae 521 43-49 Cox, J.M. (1977a). Survey of mealybug species (Homoptera: Pseudococcidae) in Auckland orchards and vineyards. New Zealand Journal of Agricultural Research. 20(2): 25961. Cox, J.M. (1977b). Status of New Zealand records of Pseudococcus comstocki and P. maritimus (Homoptera: Pseudococcidae). New Zealand Journal of Zoology 4: 165166. Cox, J. (1978). "Revision of the Rhizoecus species (Homoptera: Pseudococcidae) known from New Zealand." New Zealand Journal of Zoology 5: 623-638. Cox, J. M. (1987). Pseudococcidae (Insecta: Hemiptera). Fauna of New Zealand 11. Science Information Publishing Centre, DSIR, Wellington New Zealand: 230pp. Credi, R. and Babini, A. R. (1996). "Effect of virus and virus-like infections on the growth of grapevine rootstocks." Advances in Horticultural Science 10(2): 95-98. Credi, R. and Babini, A. R. (1997). "Effects of virus and virus-like infections on growth, yield and fruit quality of Albana and Trebbiano Romagnolo grapevines." American Journal of Enology and Viticulture 48: 7-12. de Boer, J.A. (1967). "New Records of Pseudococcidae from New Zealand (Homoptera: Coccoidea)." New Zealand Entomologist 3(5): 8-10.

91 de Wet, O. and Moores, G. (2003). Vine mealybug and chemical control: Is this the end of the road? Wynboer, 2003. http://www.wynboer.co.za/recentarticles/0803bugchem.php3 Digiaro, M., Bedzrob, M. P., D' Onghia, A.M., Boscia, D. and Savino, V. (1994). "On the correlation between grapevine virus A and rugose wood." Phytopathologia Mediterranea 33(3): 187-193. Dimitrijevic D. 1973. Some observations on natural spread of grapevine leafroll disease in Yugoslavia. Rivista di Patologia Vegetale. SIV 9: 114-119. Dohlen, C. D. von, Kohler, S., Alsop, S.T. and McManus, W.R. (2001). "Mealybug betaproteobacterial endosymbionts contain gamma-proteobacterial symbionts." Nature London 412(6845): 433-436. Dorn, B., Mattiacci, L., Bellotti, A.C. and Dorn, S. (2003). "Effects of a mixed species infestation on the cassava mealybug and its encyrtid parasitoids." Biol. control 27(1): 1-10. Douglas, A.E. (1998). "Nutritional interactions in insect-microbial symbioses: aphids and their symbiotic bacteria Buchnera." Annual Review of Entomology 43: 1737. Downie, D.A. and Gullan, P. J. (2005). Phylogenetic congruence of mealybugs and their primary endosymbionts. Journal of Evolutionary Biology 18(2): 315-? Duodu, Y.A. and Thompson, W. (1992). "Management of ant-mealybug complex in pineapple fields in Guyana." FAO Plant Protection Bulletin 40(2): 82-88. Eastwell, K.C. and Bernardy, M.G. (2001). "Partial characterization of a closterovirus associated with apple mealybug-transmitted little cherry disease in North America." Phytopathology 91(3): 268-273. Eguagie, W.E. (1972). "Non-transmittability of Offa-Igbo isolate of cocoa swollen shoot virus (CSSV) by Aphis gossypii Glover (Homoptera: Aphididae)." Nigerian Entomologists' Magazine 2(2): 49-55. Engelbrecht, D.J. and Kasdorf, G.G.F. (1987). "Occurrence and transmission of grapevine virus A in South African grapevines." South African Journal for Enology and Viticulture 8(1): 23-29. Engelbrecht, D.J. and Kasdorf, G.G.F. (1990a). "Field spread of corky bark, fleck, leafroll and Shiraz decline diseases and associated viruses in South African grapevines." Phytophylactica 22(3): 347-354. Engelbrecht, D.J. and Kasdorf, G.G.F. (1990b). "Transmission of grapevine leafroll disease and associated closteroviruses by the vine mealybug, Planococcus ficus." Phytophylactica 22(3): 341-346. Esau, K. (1948). Phloem structure in the grapevine, and its seasonal changes. Hilgardia 18: 217-296. Faoro F., Tornaghi R., Fortusini A. and Belli G. (1981). Association of a possible closterovirus with grapevine leafroll in Northern Italy. Rividsta di Patologia Vegetale 17:183-189. Firempong, S. (1984). "Laboratory and field evaluation of cocoa progenies for resistance to mealybug vectors (Hemiptera: Pseudococcidae) of swollen shoot virus." Bulletin of Entomological Research 74(1): 97-102. Fourie, J.C. 1996. Uitkenning en chemiese beheer van belangrike onkruide in Wingerde van Suid-Afrika. Nooitgedacht Pers, Kaapstad. Fourie, J.C., Louw, P.J.E. and Agenbach, G.A. 1997. The effect of different cover crop species and cover crop management practices on the available N and N-status of young Sauvignon blanc vines on a sandy soil in Lutzville. Abstract, South African Society for Enology and Viticulture Congress, 27-28 November 1997, Cape Town. Franco, J.C., Suma, P., da Silva, E.B., Blumberg, D. and Mendel, Z. (2003). Management strategies of mealybug pests of citrus in Mediterranean countries. 'Integrated control in citrus fruit crops'. Proceedings of the IOBC WPRS Working Group, Valencia, Spain, 6 8 November, 2002. Bulletin OILB SROP. 2003, 26: 6, 137.

92 Fuchs M., Walter B. and L. Pink 2000. Evaluation of transgenic grapevine rootstocks expressing the coat protein gene of grapevine fanleaf virus inder vineyard conditions. Extended Abstracts 13th ICVG Conference, Adelaide. Pp 50-51 Furness, G.O. and Charles J.G. 1994. Mealybugs. pp 54-55 in Diseases and Pests (eds P. Nicholas, P. Magarey, M. Watchel). Grape Productions Series No. 1. Winetitles, Adelaide, SA. 106pp. Garau, R., Prota, V. A., Boscia, D., Fiori, M. and Prota, U. (1995). "Pseudococcus affinis Mask., new vector of grapevine trichoviruses A and B." Vitis 34(1): 67-68. German, T. L., Ullman, D.E. and Gunashinghe, U.B. (1992). "Mealybug wilt of pineapple." Advances in Disease Vector Research 9: 241-259. Goheen, A.C. (1981). Grape Virus Diseases. In: D.L. Flaherty (Ed.): Grape pest management. Berkeley, Agricultural Sciences Publications, University of California: 84-92. Goheen, A.C. (1988). Leafroll. In: R. C. Pearson and A. C. Goheen (eds): Compendium of Grape Diseases. The American Phytopathological Society Press 52 Goheen A.C. and Cook J.A. (1959). Leafroll (red-leaf or Rougeau) and its effect on vine growth, fruit quality and yields. American Journal of Enology and Viticulture 10: 173181. Goheen A.C. Harmon F.N. and Weinberger J.H. (1958). Leafroll (White Emperor Disease) of grapes in California. Phytopathology 48: 51-54. Golino, D.A. (1993). "Potential interactions between rootstocks and grapevine latent viruses." American Journal of Enology and Viticulture 44(2): 148-152. Golino, D.A., Sim, S.T., Gill, R. and Rowhani, A. (2002). "California mealybugs can spread grapevine leafroll disease." California Agriculture 56(6): 196-201. Gonsalves, D. (1998). Control of papaya ringspot virus in papaya: A case study. Annual Review of Phytopathology, 36, 415-437. Gonzlez E, Daz T, Mosquera MV (1995). Effects of various types of virus on Vitis vinifera L. cv. Albario cultivated in vitro. Vitis 34 (4): 243-244. Gonzlez, E., Mosquera, M.V., San Jos M.C. and Daz, T. (1997). Influence of virus on the chlorophyll, carotenoid and polyamine contents in grapevine microcuttings. J. Phytopathology 145: 185-187. Gonzalez-Hernandez, H., Johnson, M.W. and Reimer, N.J. (1999). "Impact of Pheidole megacephala (F.) (Hymenoptera : Formicidae) on the biological control of Dysmicoccus brevipes (Cockerell) (Homoptera : Pseudococcidae)." Biological Control 15(2): 145. Gonzalez-Hernandez, H., Pandey, R.R. and Johnson, M.W. (2005). "Biological characteristics of adult Anagyrus ananatis Gahan (Hymenoptera : Encyrtidae), a parasitoid of Dysmicoccus brevipes (Cockerell) (Hemiptera : Pseudococcidae)." Biological Control 35(2): 93. Gonzalez-Hernandez, H., Reimer, N.J. and Johnson, M.W. (1999). "Survey of the natural enemies of Dysmicoccus mealybugs on pineapple in Hawaii." Biocontrol 44(1): 47. Goszczynski D.E., Jooste, A.E.C. 2003. Shiraz disease is transmitted by the mealybug Planococcus ficus and associated with Grapevine virus A. Extended Abstracts 14th Meeting of ICVG, Locorotondo 2003: 219 Grilli, M. P. and Holt, J. (2000). "Vector feeding period of variability in epidemiological models of persistent plant viruses." Ecological Modelling 126: 49-57. Gugerli P., Brugger J.J. and Bovey, R. (1984). Lenroulement de la vigne: misse en vidence de particule virales et dveloppement dune mthode immunoenzymatique pour le diagnostic rapide. Rev. Suisse Viticult. Aboricult. Horticult. 16: 299-304. Guidoni, S., Ferrandino, A., Argamante, N., Mannini, F. and Di Stefano, R. (1997). "The effect of grapevine leafroll and rugose wood sanitation on agronomic performance and berry and leaf phenolic content of a nebbiolo clone (Vitis vinifera L.)." American Journal of Enology and Viticulture 48(4): 438-442.

93 Guidoni, S., Mannini, F., Ferrandino, A., Argamante, N. and di Stefano, R. (2000). "Effect of virus status on leaf and berry phenolic compounds in two wine grapevine Vitis vinifera cultivars." Acta Horticulturae 526: 445-452. Habili, N., Ewart, A.J.W., Fazeli, C.F., Scott, N.S., Krake, L.R. and Rezaian, M.A. (1996). Virus types associated with leafroll disease in Australia. Australian & New Zealand Grapegrower and Winemaker: 25-28. Habili N. and Nutter F. (1997). Temporal and spatial analysis of Grapevine leafroll-associated virus 3 in Pinot Noir grapevines in Australia. Plant Disease 81:625-628. Habili N., Symons R.H. 2000. Grapevine viruses detected by Waite Diagnostics in Australia. Extended Abstracts 13th ICVG Conference, Adelaide. Pp 124-126. Hahn, S.K., Caveness, F.E., Lema, K.M. and Theberge, R.L. (1990). Breeding cassava and sweet potato for pest and disease resistance in Africa. Integrated pest management for tropical root and tuber crops: proceedings of the workshop on the global status of and prospects for integrated pest management of root and tuber crops in the tropics held in Ibadan, Nigeria on 25 30 Oct. 1987. 1990, 66 72., International Institute of Tropical Agriculture (IITA); Ibadania; Niger. Hale, C.R and Woodham, R.C. (1979). Effect of leafroll disease on the acid and potassium composition of sultana grapes. American Journal of Enology and Viticulture 30(2): 91-92. Haviland, D.R, Bentley, W.J., Daane, K.M. 2005. Hot-water treatments for control of Planococcus ficus (Homoptera: Pseudococcidae) on dormant grape cuttings. Journal of Economic Entomology 98(4): 1109-1115. Herschler, A. (1936). Ernhrungsstrungen an Reben durch Bodenverhltnisse mit Besonder Bercksichtigung von Kalimangelschden. Ernhr. Pflanze 32:197-204. Hoefert, L.L. and Gifford, E.M. (1967). "Grapevine leafroll virus - history and anatomic effects." Hilgardia 38: 403-426. Hoffmeister, M. and Martin, W. (2003). "Interspecific evolution: microbial symbiosis, endosymbiosis and gene transfer." Environmental Microbiology 5(8): 641-649. HortResearch (2004). New Zealand Horticulture Facts and Figures: Fruit, Flowers and Vegetables. The Horticulture and Food Research Institute of New Zealand Limited, Auckland. Pp. 33. Hristov, I. and Abrasheva, P. (2001). "Effect of grapevine fanleaf virus and grapevine leafrollassociated virus 3 on vine plants under conditions of in vitro cultivation." Rastenievudni Nauki (Plant Science), Bulgaria 38(5-6): 269-274. Hu, J.S., Sether, D.M., Metzer, M.J., Perez, E., Gonsalves, A., Karasev, A.V. and Nagai, C. (2005). "Pineapple mealybug wilt associated virus and mealybug wilt of pineapple." Acta Horticulturae 666: 209-212. Hu, J.S., Sether, D.M. and Ullman, D.E. (1996). "Detection of pineapple Closterovirus in pineapple plants and mealybugs using monoclonal antibodies." Plant Pathology 45(5): 829-836. Hull, R. (2002). Matthews Plant Virology, 4th Edition, Academic Press. Jahn, G.C. and Beardsley, J.W. (1996). "Effects of Pheidole megacephala (Hymenoptera: Formicidae) on survival and dispersal of Dysmicoccus neobrevipes (Homoptera: Pseudococcidae)." J. econ. Entomol. 89(5): 1124-1129. Jahn, G.C. and Beardsley, J.W. (2000). "Interaction of ants (Hymenoptera: Formicidae) and mealybugs (Homoptera: Pseudococcidae) on pineapple." Proc. Hawaii Entomol. Soc. 34: 181-185. Jahn, G.C., Beardsley, J.W. and Gonzalez-Hernandez, H. (2003). "A review of the association of ants with mealybug wilt disease of pineapple." Proceedings of the Hawaiian Entomological Society 36: 9-28.

94 Jeger, M.J., Van Den Bosch, F., Madden, L.V. and Holt, J. (1998). "A model for analysing plant-virus transmission characteristics and epidemic transmission." IMA Journal of Mathematics Applied in Medicine and Biology 15: 1-18. Jooste A.E.C. and Goszczynski D.E. (2005). Single-strand conformation polymorphism (SSCP), cloning and sequencing reveal two major groups of divergent molecular variants of grapevine leafroll-associated virus 3 (GLRaV-3). Vitis 44: 39-43. Jordan, D. (1993). "Leafroll spread in New Zealand vineyards." Australian and New Zealand Wine Industry Journal 8(4): 322-324. Kim, H.R., Chung, J.D., Kim, K.R., Choi, Y.M., Yiem, M.S. and Park, J.W. (2003). Effects of grapevine leafroll-associated virus 3 infection on vine growth and fruit quality in Kyoho grapevines. Journal of the Korean Society for Horticultural Science 44(3): 335-339. Kliewer W.M and Lider L.A. (1976). Influence of leafroll virus on composition of Burger fruits. American Journal of Enology and Viticulture 27(3): 118-124. Kovacs, L.G., Hanami, H., Fortenberry, M. and Kaps, M.L. (2001). "Latent infection by leafroll agent GLRaV-3 is linked to lower fruit quality in French-American hybrid grapevines Vidal blanc and St. Vincent." American Journal of Enology and Viticulture 52(3): 254-259. Krake L.R. (1993). Characterisation of grapevine leafroll disease by symptomatology. Australia and New Zealand Wine Industry Journal 8: 40-44. Krastanova, S., Ling, K.S., Zhu, H.Y., Xue, B., Burr, T.J. and Gonsalves, D. (2000). "Development of transgenic grapevine rootstocks with genes from grapevine fanleaf virus and grapevine leafroll associated closteroviruses 2 and 3." Acta Horticulturae 528: 367-372. Krastanova S., Perrin M., Barbier P. Demangeat G., Cornuet P., Bardonnet N., Otten L., bPinck L and Walter B. 1995. Transformation of grapevine rootstocks with the coat protein gene of grapevine fanleaf nepovirus. Plant Cell Reports 14: 550-554. Krger K., Saccagi D. and Douglas N. 2006. Grapevine leafroll-associated virus 3-vector interactions: transmission by the mealybugs Planococcus ficus and Pseudococcus longipsinus (Hemiptera: Pseudococcidae). Extended Abstracts 15th Meeting of ICVG, Stellenbosch: 130-131. Kubiriba, J., Legg, J.P., Tushemereirwe, W. and Adipala, E. (2001). "Vector transmission of Banana streak virus in the screenhouse in Uganda." Annals of Applied Biology 139(1): 37-43. Lahogue, F. and Boulard, G. 1996. Recherche de genes de rsistance naturelle deux viruses de la vigne: le court-nou et lenroulement. Vitis 35: 43-48. Le Roux, P. (1996). Die chemiese beheer van wingerdwitluis, in Wingerdwitluis paneelbespreking, SAWWV Kongres, Kaapstad, 8 November 1996. Le Ru, B. and Mitsipa, A. (2000). "Influence of the host plant of the cassava mealybug Phenacoccus manihoti on life-history parameters of the predator Exochomus flaviventris." Entomol. Exp. Applic. 95(2): 209-212. Le Ru, B. and Tertuliano, M. (1993). "Tolerance of different host-plants to the cassava mealybug Phenacoccus manihoti Matile-Ferrero (Homoptera: Pseudococcidae)." International Journal of Pest Management 39(4): 379-384. Legg, J.T. and Kumar, R. (1975). Some aspects of swollen shoot disease. Proceedings of the 4th Conference of West African Cocoa Entomologists, Zoology Department, University of Ghana, Legon, Ghana, 9th 13th December 1974. 1975, recd. 1977, 58 61; 20 ref. Legon; Ghana, Zoology Department University of Ghana. Legg, J.T. and Lockwood, G. (1981). "Resistance of cocoa to swollen-shoot virus in Ghana. I. Field trials." Annals of Applied Biology 97(1): 75-89.

95 Lester, P.J., Baring, C.W., Longson, C.G., and Hartley, S. (2003). Argentine and other ants (Hymenoptera: Formicidae) in New Zealand horticultural ecosystems: distribution, hemipteran hosts, and review. New Zealand Entomologist 26: 79-89. Lider, L. A., Goheen, A.C. and Ferrari, N.L. (1975). "A comparison between healthy and leafroll-affected grapevine planting stocks." American Journal of Enology and Viticulture 26(3): 144-147. Ling K-S., Zhu H-Y., Petrovic N. and Gonsalves D. (2001). Comparitive effectiveness of Elisa and RT-PCR for detecting Grapevine leafroll-associated closterovirus-3 in field samples. Am. J. Enol. Vitic. 52: 21-27. Ling, K.S., Zhu, H.Y. and Gonsalves, D. (2004). "Complete nucleotide sequence and genome organisation of Grapevine leafroll-associated virus 3, type member of the genus Ampelovirus." Journal of General Virology 85: 2099-2102. Lockwood, G. (1981). "Resistance of cocoa to swollen-shoot virus in Ghana. II. Assessment of seed inoculation experiments." Annals of Applied Biology 97(1): 91-97. Macanawai, A.R., Ebenebe, A.A., Hunter, D., Devitt, L.C., Hafner, G J. and Harding, R.M. (2005). "Investigations into the seed and mealybug transmission of Taro bacilliform virus." Australasian Plant Pathology 34(1): 73-76. McCarthy M.G, Cirami R.M and Van Velsen R.J. 1989. Virus thermotherapy effects on the performance of a Muscadelle selection. Vitis 28: 13-19. McKenzie, D.J., Johnson, R.C. and Warner, C. 1996. Incidence of four important viral pathogens in Canadian vineyards. Plant Diseases 80: 955-958. McKissock A. (1964). Two important virus diseases of grapevines in New Zealand. New Zealand Journal of Agricultural 108: 332-339. Mahungu, N.M., Dixon, A.G.O. and Kumbira, J.M. (1994). "Breeding cassava for multiple pest resistance in Africa." African Crop Science Journal 2(4): 539-552. Mannini, F., Argamante, N. and Credi, R. (1996). "Improvements in the quality of grapevine Nebbiolo clones obtained by sanitation." Acta Horticulturae 427: 319-324. Mannini, F., Argamante, N. and Credi, R. (1999). "Contribution of virus infections to clonal variability of some *Vitis vinifera* L. cultivars." Bulletin de l'O.I.V. 72(817-818): 144-160. Mannini, F., Argamante, N. and Credi, R. (2000). "Leaf morphological modifications induced by different viruses in clones of Vitis vinifera cultivars." Acta Horticulturae 528: 765768. Mannini, F., Credi, R. and Argamante, N. (1994). Changes in field performance of clones of the grapevine cv. Nebbiolo after virus elimination by heat therapy. 6th International Symposium on Grape Breeding, Yalta, Crimea, Ukraine: 117-122. Mannini, F., Credi, R., Gerbi, V., Lisa, A., Minati, J.L. and Argamante, N. (1994). "Effect of viral infections on agronomical and enological behaviour of clones of the grapevine cv. Ruche and Dolcetto." Quaderni della Scuola di Specializzazione in Viticoltura ed Enologia, Univ. Torino 18: 55-71. Mannini, F., Gerbi, V. and Credi, R. (1998). "Heat-treated virus-infected grapevine clones: Agronomical and enological modifications." Acta Horticulturae 473: 155-163. Mannini, F., Guidoni, S., Ferrandino, A., Argamante, N. and Credi, R. (1997). Photosynthesis and grape composition of a Vitis vinifera clone after virus sanitation. 12th Meeting of the international council for the study of viruses and virus-like diseases of the grapevine (ICVG), Lisbon, Portugal: 155-156. Martelli, G. P., Agranovsky, A.A., Bar-Joseph, M., Boscia, D., Candresse, T., Coutts, R.H. A., Dolja, V.V., Falk, B.W., Gonsalves, D., Jelkmann, W., Karasev, A.V., Minafra, A., Namba, S., Vetten, H.J., Wisler, G.C. and Yoshikawa, N. (2002). "The family Closteroviridae revised." Archives of Virology 147(10): 2039-2044. Martelli, G.P. 2003. Grapevine virology highlights 2000-2003. Extended abstracts, 14th Meeting of ICVG, Locorotondo 2003: 3-10.

96 Martin, R.R., Eastwell, K.C., Wagner, A., Lamprecht, I. and Tzanetakis, I.E. 2005. Suvery for viruses of grapevine in Oregon and Washington. Plant Disease 89: 763-766. Martinez-Torres, D., Bauades, C., Latorre, A. and Moya, A. (2001). "Molecular systematics of aphids and their primary endosymbionts." Molecular Phylogenetics and Evolution 20: 437449. Matthews, C. (1996). Epidemiology of grapevine leafroll disease within New Zealand vineyards. MSc thesis University of Auckland. Minafra, A. and Hadidi, A. (1994). "Sensitive detection of grapevine virus A, B, or leafrollassociated III from viruliferous mealybugs and infected tissue by cDNA amplification." Journal of Virological Methods 47(1-2): 175-187. Monis J. and Bestwick R.K. (1996). Detection and localization of grapevine leafrollassociated closteroviruses in greenhouse and tissue culture grown plants. Am. J. Enol. Vitic. 47:199-205. Moran, N. A., Plague, G.R., Sandstrm, J.P. and Wilcox, J.L. (2003). "A genomic perspective on nutrient provisioning by bacterial symbionts of insects." Proceedings of the National Academy of Science, USA 100: 1454314548. Morin, S., Ghanim, M., Zeidan, M., Czosnek, H., Verbeek, M. and van den Heuvel, J.F.J.M. (1999). A GroEL homologue from endosymbiotic bacteria of the whitefly Bemisia tabaci is implicated in the circulative transmission of tomato yellow leaf curl virus. Virology 256(1):75-84. Mossop, D.W., Elliott, D.R. and Richards, K.D. (1985). Association of clostervirus-like particles and high molecular weight double-stranded RNA with grapevines affected by leafroll disease. N.Z. J. Agric. Res. 28: 419-425. Mullins, M.G, Bouquet, A, and Williams, L.E. (1992). Biology of the Grapevine. Cambridge Univerisyt Press, UK: 239 pp. Namba, S., Yamashita, S., Doi, Y. and Yora, K. (1979). Grapevine leafroll virus, a possible member of the closteroviruses. Annals Phytopath Society Japan 45: 497-502. Nel, A.C. and Engelbrecht, D.J. (1972). Grapevine virus diseases in South Africa and the influence of latent viruses in the nursery. Ann. Phytopathol., n' horssrie: 67-74. Nel, A., Krause, M., Ramautar, N. and Van Zyl, K. (1999). A guide to the control of plant pests. Directorate: Agricultural Production Input, National Department of Agriculture, Republic of South Africa. Neuenschwander, P. (2001). "Biological control of the cassava mealybug in Africa: a review." Biol control 21(3): 214-229. Ngeve, J.M. (2003). "The cassava root mealybug (Stictococcus vayssierei Richard) (Homoptera: Stictococcidae): a threat to cassava production and utilization in Cameroon." International Journal of Pest Management 49(4): 327-333. Nickel, O., Chagas, C.M. and Vasconcelos, A.P.A. (2000). "Association of pineapple mealybug wilt with closterovirus-like particles and dsRNA in Bahia, Brazil." Fitopatologia Brasileira 25(2): 200-202. Nimmo-Bell, (2005). Development of the New Zealand Grafted Grapevine Standard. A Nimmo-Bell Publication, June 2005: 32p. La Notte, P., Buzkan, N., Choueiri, E., Minafra, A., and Martelli, G.P. and (1997). "Acquisition and transmission of grapevine virus A by the mealybug Pseudococcus longispinus." Journal of Plant Pathology 78(1): 79-85. Ogbe, F.O., A.C. Ohiri and E.C. Nnodu (1993). Effect of NPK fertilization on symptom severity of African cassava mosaic virus. International Journal of Pest Management, 39(1), 80-83. Okoli, O.O. (1982). Prospects of breeding for host plant resistance to cassava mealybug (Phenacoccus manihoti) and to green spider mites of cassava (Mononychellus tanajoa). Proceedings, international workshop on control of cassava mealybug and

97 green spider mite, Umudike, Nigeria, 1981. 1982, 72 75; 16 ref. Umudikeia; Niger, National Root Crops Research Institute. Ollennu, L.A.A Hughes Jd, A., Owusu, G.K. and D' A. Hughes J. (1996). "Mild strain crossprotection of cocoa against cocoa swollen shoot badnavirus." Tropical Science 36(2): 116-128. Ollennu, L. A. A., Owusu, G. K. and Thresh, J. M. (1989). "The control of cocoa swollen shoot disease in Ghana." Cocoa Growers' Bulletin 42: 25-35. Over de Linden A.J. and Chamberlain E.E. (1970a). Effect of grapevine leafroll virus on vine growth and fruit yield and quality. New Zealand Journal of Agricultural Research 13: 689-698. Over de Linden A.J. and Chamberlain E.E. (1970b). Production of virus-free grapevines in New Zealand. New Zealand Journal of Agricultural Research 13: 991-1000. Padi, B., Hollander, J. den and Den Hollander, J. (1996). "Morphological variation in Planococcoides njalensis occurring on cocoa in Ghana." Entomologia Experimentalis et Applicata 79(3): 317-328. Pandey, R.R. and Johnson M.W. (2005a). "Effect of pink pineapple mealybug hosts on Anagyrus ananatis Gahan size and progeny production." Biological Control 35(1): 1. Pandey, R. R. and Johnson M.W. (2005b). "Effects of cool storage on Anagyrus ananatis Gahan (Hymenoptera : Encyrtidae)." Biological Control 35(1): 9. Petersen, C.L. and Charles, J.G. (1997). "Transmission of grapevine leafroll-associated closteroviruses by Pseudococcus longispinus and P. calceolariae." Plant Pathology 46(4): 509-515. Petersen, C. and Jordan, D. (1992). ELISA works well for the identification of leafroll. Proceedings of the New Zealand Grape and Wine Symposium: profit in the market, in the winery, in the vineyard. Christchurch, New Zealand. New Zealand Society for Viticulture and Oenology: 54-55. Petty, G.J. and Tustin, H. (1993). "Ant (Pheidole megacephala F.)--mealybug (Dysmicoccus brevipes Ckll.) relationships in pineapples in South Africa." Acta-hortic 334: 387-395. Pietersen, G. (2004). Spread of grapevine leafroll disease in South Africa - a difficult, but not insurmountable problem. Wynboer. June 2004. http://www.wynboer.co.za/recentarticles/0406leaf.php3 Pirone, T.P. and Blanc S. (1996). "Helper-dependent vector transmission of plant viruses." Annual Review of Phytopathology 34: 227-247. Pitaksa, C., Chantarasuwan, A. and Kongkanjana, A. (2000). "Ant control in pineapple field." Acta horticulturae 529: 309-314. Raine, J., McMullen, R.D. and Forbes, A.R. (1986). "Transmission of the agent causing little cherry disease by the apple mealybug Phenacoccus aceris and the dodder Cuscuta lupuliformis." Can. J. Plant Pathol Rev. Can. Phytopathol. 8(1): 6-11. Renard, S., Calatayud, P.A., Pierre, J.S. and Le Ru, B. (1998). "Recognition behavior of the cassava mealybug Phenacoccus manihoti Matile-Ferrero (Homoptera: Pseudococcidae) at the leaf surface of different host plants." Journal of Insect Behavior 11(3): 429-450. Rohrbach, K.G., Beardsley, J.W., German, T.L., Reimer, N.J. and Sanford, W.G. (1988). "Mealybug wilt, mealybugs, and ants on pineapple." Plant-Dis. 72(7): 558-565. Roivainen, O. (1976). Transmission of cocoa viruses by mealybugs (Homoptera:Pseudococcidae). Journal of the Scientific Agricultural Society of Finland. 48: 3, 203 304; 8 fig., 5 diag., 20 graphs, 36 tab. Rosciglione, B. and Castellano, M. A. (1985). "Further evidence that mealybugs can transmit Grapevine virus A ( GVA) to herbaceous hosts." Phytopathol. Mediterr. 24(1/2): 186188.

98 Sampol, B., Bota, J., Riera, D., Medrano, H. and Flexas, J. (2003). "Analysis of the virusinduced inhibition of photosynthesis in malmsey grapevines." New Phytologist 160: 403-412. Santa Cecilia, L. V. C., Goncalves-Gervasio, R. de C. R., Souza, B. and Silva, J. R. da (2001). "Phenolic contents associated with the bug Dysmicoccus brevipes (Cockerell, 1893) (Hemiptera: Pseudococcidae) versus pineapple plant wilt." Ciencia e Agrotecnologia 25(2): 467-470. Scheu G. (1936). Mein Winzerbuch Reichsnahrstand Verlags Ges. M. b. H. Berlin. Scheu G. (1950). Mein Winzerbuch. Ed. 2. D. Meininger, Neustadt. Schliefert, L. (2001). Before top-working your red grapes to whites or vice versa, consider the serious problems with top-working existing vines. The Australian Grapegrower & Winemaker, August 2001: 15-17. Schoefling H. (1980). First results of a field-trial on the performance of the heat-treated and non heat-treated white Riesling clones. 7th Meeting of the International Council for the Study of Viruses and Virus-Like Diseases. (Ed. A.J. McGinnis) Niagara Falls, Canada, Sept 8-12: 311-320. Sether, D.M. and Hu J.S. (2002). "Closterovirus infection and mealybug exposure are necessary for the development of mealybug wilt of pineapple disease." Phytopathology 92(9): 928-935. Sether, D.M., Melzer, M.J., Busto, J., Zee, F. and Hu, J.S. (2005). "Diversity and mealybug transmissibility of ampeloviruses in pineapple." Plant Disease 89(5): 450-456. Sether, D.M., Ullman, D.E. and Hu, J.S. (1998). "Transmission of pineapple mealybug wiltassociated virus by two species of mealybug (Dysmicoccus spp.)." Phytopathology 88(11): 1224-1230. Sforza, R., Boudon Padieu, E. and Greif, C. (2003). "New mealybug species vectoring Grapevine leafroll-associated viruses-1 and -3 (GLRaV-1 and -3)." European Journal of Plant Pathology 109(9): 975-981. Sforza, R. and Greif, C. (2000). "Mealybugs and grapevine leafroll." Phytoma 532: 46-50. Sipes, B.S., Sether, D.M. and Hu, J.S. (2002). "Interactions between Rotylenchulus reniformis and Pineapple mealybug wilt associated virus-1 in pineapple." Plant Disease 86(9): 933-938. Souissi, R. (1999). "The influence of the host plant of the cassava mealybug Phenacoccus manihoti on the plant and host preferences of its parasitoid Apoanagyrus lopezi." Biol. control 15(1): 64-70. Souissi, R., Nenon, J.P. and Le Ru, B. (1998a). "Tritrophic interactions between host plants, the cassava mealybug Phenacoccus manihoti Matile-Ferrero (Hom., Pseudococcidae) and its parasitoid Apoanagyrus lopezi De Santis (Hym., Encyrtidae)." Journal of Applied Entomology 122(9-10): 561-564. Souissi, R. and le Ru, B. (1997). "Effect of host plants on fecundity and development of Apoanagyrus lopezi, an endoparasitoid of the cassava mealybug Phenacoccus manihoti." Entomologia Experimentalis et Applicata 82(2): 235-238. Souissi, R. and le Ru, B. (1998b). "Influence of the host plant of the cassava mealybug Phenacoccus manihoti (Hemiptera: Pseudococcidae) on biological characteristics of its parasitoid Apoanagyrus lopezi (Hymenoptera: Encyrtidae)." Bulletin of Entomological Research 88(1): 75-82. Spence, N.J. (2001). Virus-Vector Interactions. In: M.J. Jeger and N.J. Spence (eds) Plant Virus Disease Transmission and Epidemiology. Biotic Interactions in Plant-Pathogen Associations, CABI Publishing, Wallingford, UK: 15-26. Taniguchi, G., Thompson, T. and Sipes, B. (2005). "Control of the big-headed ant, Pheidole megacephala, (Hymenoptera : Formicidae) in pineapple cultivation using Amdro in bait stations." Sociobiology 45(1): 1.

99 Tanne, E. and Spiegel-Roy P (1973). Mineral content of healthy and leaf-roll infected leaves of the grape (cv. Mission). Rivista di Patologia Vegetale 9(3): 227-234. Tanne, D., Sela, I., Klein, M., and Hsrpaz, I. (1977). Purification and characterizationof a virus associated with the grapevine leafroll disease. Phytopathology 67: 442-447. Tanne, E., Naveh, L. and Sela, I. (1989). Molecular-hybridization evidence for the presence of a potyvirus in leafroll-infected grapevine. Plant Pathology 38:183-189. Teliz, D., Tanne, E., Gonsalves, D. and Zee, F. (1987). Field serological detection of viral antigens associated with grapevine leafroll disease. Plant Disease 71:704-709. Tertuliano, Calatayud, P.A. and le Ru, B.P. (1999). "Seasonal changes of secondary compounds in the phloem sap of cassava in relation to fertilization and to infestation by the cassava mealybug." Insect Science and its Application 19(1): 91-98. Tertuliano, M., Dossou Gbete, S. and Le Ru, B. (1993). "Antixenotic and antibiotic components of resistance to the cassava mealybug Phenacoccus manihoti (Homoptera: pseudococcidae) in various host-plants." Insect Science and its Application 14(5): 657665. Thao, M.L. and Baumann, P. (2004). "Evolutionary relationships of primary prokaryotic endosymbionts of whiteflies and their hosts." Appl. Environ. Microbiol. 70: 3401 3406. Thao, M.L., Moran, N.A., Abbot, P., Brennan, E.B., Burckhardt, D.H. and Baumann, P. (2000). "Cospeciation of psyllids and their prokaryotic endosymbionts." Applied Environmental Microbiology 66: 28982905. Thomas, W.T. (1976). The impact of virus diseases on quality and yield in the vineyard. Wine Review 13 (2): 17-31. Thresh, J.M. and Owusu, G.K. (1986). "The control of cocoa swollen shoot disease in Ghana: an evaluation of eradication procedures." Crop Protection 5(1): 41-52. Turturo, C., Saldarelli, P., Dong, Y., Digiaro, M., Minafra, A., Savino, V. and Martelli, G.P. (2005). Genetic variability and population structure of Grapevine leafroll-associated virus 3 isolates. J. Gen. Vrol. 86: 217-224. Ueno, K., Kinoshita, K., Togawa, H. and Iri, M. (1985). "Improvement of wine quality by elimination of grapevine leafroll virus." Journal of the Brewing Society of Japan [Nippon Jozo Kyokai Zasshi] 80(7): 490-495. Ueckermann, P. (1998). Ant control in vineyards. Wynboer Tegnies 105: 8-9. Ullman, D.E., German, T.L., Gunasinghe, U.B. and Ebesu, R.H. (1989). "Serology of a closteroviruslike particle associated with mealybug wilt of pineapple." Phytopathology 79(12): 1341-1345. Valero, M., Ibanez, A. and Morte, A. (2003). "Effects of high vineyard temperatures on the grapevine leafroll associated virus elimination from Vitis vinifera L. cv. Napoleon tissue cultures." Scientia Horticulturae 97(3-4): 289-296. Vermeulen, A.K. (1999). A Guide to the use of Registered Fungicides and Pesticides against Grapevine Diseases and Pests: Wine Grapes. ARC-Fruit Vine and Wine Research Institute, Private Bag X5026, Stellenbosch, 7599, South Africa. Walker, J.T.S, Charles, J.G, Froud, K.J and Connolly, P. (2004). Leafroll virus in vineyards: modelling the spread and economic impact. Report to New Zealand Winegrowers Limited: 19pp. Walton, V.M. (2004). Degree day estimation as a warning tool for the vine mealybug. Wynboer, December 2004. http://www.wynboer.co.za/recentarticles/0903degree.php3 Walton, V.M., (2003). Development of an integrated pest management system for vine mealybug, Planococcus ficus (Signoret), in vineyards in the Western Cape Province, South Africa. PhD dissertation, University of Stellenbosch.

100 Walton, V.M., Pringle, K.L., Danne, K.M. (2003). Vine mealybug pheromone trapping and monitoring protocol in South African vineyards. Wynboer, June 2003. http://www.wynboer.co.za/recentarticles/0803bugproto.php3 Ward, A. (1966). Mealybugs (Hemiptera: Pseudococcidae) in Hawkes Bay orchards. New Zealand Journal of Agricultural Research 9(2):127-135. Wolpert, J.A. and. Vilas, E.P (1992). "Effect of mild leafroll disease on growth, yield and fruit maturity indices of Riesling and Zinfandel." American Journal of Enology and Viticulture 43: 367-369. Woodham, R.C., Antcliff, A.J., Krake, L.R. and Taylor, R.H (1984b). "Yield differences between sultana clones related to virus status and genetic factors." Vitis 23: 73-83. Woodham, R.C, Emmett, R.W and Fletcher, G.C. (1984a). Effects of thermotherapy and virus status on yield, annual growth and grape composition of Sultana. Vitis 23: 268-273. Woodham, R.C., Krake, L.R. and Cellier, K.M (1983). "The effect of grapevine leafroll plus yellow speckle disease on annual growth, yield and quality of grapes from Cabernet Franc under two pruning systems." Vitis 22: 324-330. Xue B., Ling K-S., Reid C.L., Krastanova S., Sekiya M., Momol E.A., Sle S., Mozsar J., Gonsalves D. and Burr T.J. 1999. Transformation of five grapevine rootstocks with plant virus genes and a virE2 gene from Agrobacterium tumefaciens. In Vitro Cell. Dev. Biol.- Plant 35:226-231 Yoshikawa, N., Iida, H., Goto, S., Magome, H., Takahashi, T. and Terai,Y. (1997). Grapevine berry inner necrosis, a new trichovirus: comparative studies with several known trichoviruses. Arch. Virol. 142 (7): 1351-1363. Zeddies, J., Schaab, R.P., Neuenschwander, P. and Herren, H. R. (2001). "Economics of biological control of cassava mealybug in Africa." Agric. econ. 24(2): 209-219. Zimmermann, D., Bass, P., Legin, R. and Walter, B. (1990). "Characterization and serological detection of four closterovirus-like particles associated with leafroll disease on grapevine." Journal of Phytopathology 130: 205-218. Zimmermann, D., Sommermeyer, G., Walter, B. and Van Regenmortel, M.H.V. (1990). "Production and characterization of monoclonal antibodies specific to closteroviruslike particles associated with grapevine leafroll disease." Journal of Phytopathology 130: 277-288.