205Ann. appl. Biol. (2004), 144:205-218Printed in UK

*Corresponding Author E-mail: fiona_constable@hotmail.com

© 2004 Association of Applied Biologists

The incidence, distribution and expression of Australian grapevine yellows,restricted growth and late season leaf curl diseases in selected Australian

vineyards

By F E CONSTABLE1*, J JONES2, K S GIBB3, Y M CHALMERS4 and R H SYMONS1

1Department of Plant Science, Waite Campus, University of Adelaide, Glen Osmond, South Australia5064, Australia

2BiometricsSA, University of Adelaide/South Australian Research and Development Institute, GPO Box397, Adelaide SA 5001, Australia

3School of Science, Charles Darwin University, Darwin, NT 0909, Australia4Department of Primary Industries, PO Box 905, Mildura, Victoria, 3502, Australia

(Accepted 17 October 2003; Revised version received 29 September 2003)

Summary

Surveys were conducted in four Chardonnay vineyards for 3 to 6 years and one Shiraz vineyard for3 years to determine the yearly percentage of grapevines affected by Australian grapevine yellowsdisease (AGYd), restricted growth disease (RGd) and late season leaf curl disease (LSLCd). In each ofthe Chardonnay vineyards in each year, all three diseases were characterised by remission of diseasein some grapevines, recurrence of disease in other grapevines and new observations of disease inpreviously unaffected grapevines. The pattern of temporal incidence of each disease was differentbetween vineyards for the survey period. Although Koch�s postulates have not been fulfilled,phytoplasmas are considered to be the most likely cause of AGYd. While some grapevines exhibiteda combination of AGYd and RGd or AGYd and LSLCd, both RGd and LSLCd can occur independentlyof AGYd. Statistical analyses using log-linear models also indicated that RGd and LSLCd were notalways associated with AGYd. Thus, it is possible that phytoplasmas are not the cause of RGd orLSLCd and their association is coincidental. Expression of AGYd in Shiraz grapevines occurred laterin the season compared to Chardonnay. Very little recurrence of AGYd was observed in the Shirazgrapevines indicating that the variety Shiraz responds differently to phytoplasma infections, assumingthat AGYd in Shiraz is a phytoplasma caused disease. RGd and LSLCd were not observed on anygrapevines in the Shiraz vineyard.

Key words: Spatial distribution, incidence, Australian grapevine yellows, temporal, restricted growth,late season leaf curl

Introduction

Surveys have shown that Australian grapevineyellows disease (AGYd) and phytoplasmas arefound in most viticultural regions of Australia(Bonfiglioli et al., 1996; Magarey & Wachtel,1986b). Electron microscopy (Magarey et al., 1988),antibiotics (Magarey & Wachtel, 1986b) and PCRtechniques (Bonfiglioli et al., 1995; Gibb et al., 1999;Padovan et al., 1995; Constable et al., 2003) haveconfirmed an association between phytoplasmas andAGYd. Three different phytoplasmas have beenassociated with AGYd and include the Australiangrapevine yellows (AGY) phytoplasma, tomato bigbud (TBB) phytoplasma and the Buckland Valleygrapevine yellows (BVGY) phytoplasma (Constableet al., 1998, 2002; Gibb et al., 1999). The AGYphytoplasma is more frequently detected in AGYdaffected grapevines compared to the TBB

phytoplasma (Constable et al., 2003, 1998; Gibb etal., 1999). The BVGY phytoplasma has only beenfound in two vineyards in the Buckland Valley grapegrowing region of Victoria (Constable et al., 2002;Gibb et al., 1999).

AGYd occurs more often in the warmer inlanddistricts of Sunraysia in New South Wales andVictoria, Riverina in New South Wales and theRiverland in South Australia compared to otherviticultural regions. Chardonnay and Riesling appearto be affected by AGYd more often than othervarieties (Magarey & Wachtel, 1986a). However,AGYd symptoms have been observed andphytoplasmas have been detected in other white andred varieties (Bonfiglioli et al., 1996). Previousresearch has shown that the incidence of AGYd maybe temporarily high in one year in some Chardonnayvineyards (Magarey & Wachtel, 1986b).

Restricted growth disease (RGd) also commonly

206 F E CONSTABLE ET AL.

affects Chardonnay grapevines in Sunraysia(Bonfiglioli et al., 1995; Padovan et al., 1995).Symptoms of RGd include retarded growth resultingin shortened shoots and smaller leaves. The affectedgrapevines have an overall appearance of stuntingor lack of vigour throughout the season comparedto unaffected grapevines. Some grapevines with RGdmay have also displayed uneven or no buddevelopment resulting in canes and cordons thatwere bare in places or entirely bare with little or nobunch development. No significant associationbetween phytoplasmas and shoots of grapevinesaffected by RGd alone has been shown (Bonfiglioliet al., 1995; Gibb et al., 1999; Padovan et al., 1995).However in another study, phytoplasmas were morefrequently detected in grapevines that had displayedboth AGYd and RGd compared to grapevinesdisplaying AGYd alone (Constable et al., 2003).Phytoplasmas are considered as one possible causeof RGd.

Late season leaf curl disease (LSLCd) can alsoaffect Chardonnay grapevines in the Sunraysiadistrict (Bonfiglioli et al., 1995; Gibb et al., 1999).Symptoms associated with LSLCd include theappearance of symptoms from late summer whereleaves on affected shoots roll tightly downward,remain green, and overlay one another in a shingledappearance. The leaves are often tough, leathery andquite brittle. In one study phytoplasmas weredetected in 48/59 LSLCd affected shoot samples(late AGY; Bonfiglioli et al., 1995). In another studyphytoplasmas were detected in 12/126 shoot samplesfrom LSLCd affected grapevines and of these 12,four were AGY phytoplasma and eight were the TBBphytoplasma (Gibb et al., 1999). An additional studyshowed that phytoplasmas were more frequentlydetected in grapevines affected by LSLCd and AGYdcompared to grapevines affected by either diseasealone (Constable et al., 2003). Consequently theassociation between phytoplasmas and LSLCd isuncertain.

In addition to the PCR studies, two surveys havebeen conducted to determine the incidence of AGYd,RGd and LSLCd in vineyards. From observationsin one vineyard, Bonfiglioli et al. (1997) proposedthat AGYd, RGd and LSLCd diseases were relatedand that LSLCd was followed by the presence ofAGYd in the same grapevines in following years. Itwas also proposed that AGYd often led to thepresence of RGd in subsequent years. Other researchusing an analysis of relative risk on the observationsfrom three vineyards over 3 yr also suggested thatAGYd may be associated with RGd or LSLCd(Constable et al., 2000). Phytoplasmas and diseaseswith a real or putative association withphytoplasmas, including AGYd, RGd and LSLCd,are of considerable concern to the Australianviticulture industry because of reduced grapevine

health and potential yield loss.To further understand the epidemiology of AGYd,

RGd and LSLCd, the yearly and cumulative patternof incidence of each disease was determined in fivevineyards over 3 to 6 yr. The spatial distribution ofeach disease in each year was also determined. Tounderstand disease progression in individualgrapevines, the temporal expression of each diseasein individual grapevines was determined. Thefrequency of association between AGYd and RGdor AGYd and LSLCd was examined to determine ifphytoplasmas might also be a possible cause for RGdand LSLCd, assuming AGYd is a phytoplasmacaused disease.

Materials and Methods

Study sitesTable 1 lists the vineyards used as study sites for

the disease survey. The number and location ofgrapevines with AGYd, RGd and LSLCd wasdetermined at each study site.

Disease descriptorsAustralian grapevine yellows disease (AGYd)

symptoms included one or more shoots showingirregular veinal and interveinal yellowing anddownward rolling of leaves that overlaid one anotherin a shingled appearance. Shoots affected withAGYd may also have displayed abortion of theflowering bunches early in the season or berry shrivellater in the season. Affected shoots often displayedtip death followed by dieback of the shoots, nodeby node. Leaf blades on affected shoots tended tofall early. The petioles remained attached to theshoots longer than the leaf blades but wouldeventually abscise from the shoot. The stem ofaffected shoots often developed a blue, waxyappearance and remained rubbery later in the season.

Restricted growth disease (RGd) symptomsincluded retarded growth resulting in shortenedshoots and smaller leaves. Affected grapevines havean overall appearance of stunting or lack of vigourthroughout the season compared to unaffectedgrapevines. Some grapevines with RGd may havealso displayed uneven or no bud developmentresulting in canes and cordons that were bare inplaces or entirely bare with little or no bunchdevelopment.

Late season leaf curl disease (LSLCd) was definedby the appearance of symptoms from late summerwhere leaves on affected shoots rolled tightlydownward, remained green, and overlaid one anotherin a shingled appearance. The leaves were oftentough, leathery and brittle.

Spatial analysis of disease survey dataThe disease survey data for AGYd, RGd and

207Distribution of Australian grapevine diseases

RGd, AGYd and LSLCd presence or absence werecounted. Log-linear models (McCullagh & Nelder,1989) were used to determine if the presence ofAGYd, RGd and LSLCd changed over time and ifthere was any interaction between the three diseasesand time at each vineyard. The models assume aPoisson distribution with a logarithmic link function.Terms were dropped from the models by usingbackward elimination, which determines theimportance of each term by assessing the change inresidual deviance. Non-significant terms wereomitted from the models.

Results

Vineyard surveys for disease incidenceThe yearly incidence (%) was determined for each

disease from the number of grapevines that wereaffected by a disease in each year. The cumulativeincidence (%) of each disease was the sum of allnew records of diseased grapevines in each year andall records of diseased grapevines in previous years.

The incidence of AGYd in each year was differentat each study site (Fig. 1). The pattern of AGYdincidence from year to year was different betweenstudy sites. The cumulative incidence of AGYd ineach vineyard increased from year to year over thesurvey period (Fig. 1). By 2000/01 the cumulativeincidence of AGYd at each study site was: 77.3% atGol Gol; 71.5% at Karadoc; 53.7% at Wemen (6yr); 55.7% at Wemen (5 yr); 90.8% at Colignan; and9.6% at Paringa.

New occurrences of AGYd were observed ingrapevines that were planted during the survey.AGYd affected 48/151 replacement grapevines atGol Gol, 5/9 replacement grapevines at Karadoc, 3/28 replacement grapevines at Colignan and 1/29replacement grapevines at Paringa. In 1999/00,AGYd was also observed in the new planting ofShiraz grapevines behind the blocks of Shiraz atParinga, but the incidence of affected grapevines wasnot recorded. AGYd affected seven replacementChardonnay grapevines in the same year that they

LSLCd from Gol Gol, Karadoc, Colignan andParinga were analysed using a fixed grid analysisavailable through the �Patchy� spatial analysispackage (Maixner, 1993) to analyse clusteringamongst groups of diseased grapevines. The fixedgrid analysis lays a contiguous grid of subunits overthe plot and all subunits, except those exceeding thetolerance level for missing plants, are used forcalculation. The fixed grid analysis calculates themean number of affected plants in each subunit ofthe grid and the variance amongst subunits acrossthe plot. The size of the subunits tested for clusteringrange from 2 × 2 (minimum size allowed) grapevinesin rows and columns to row-length/2 × column-length/2 (maximum size allowed) grapevines. Eachgrid of a specific subunit size is tested in separatefixed grid analysis.

Spatial autocorrelation using the �join-count� teststatistic was used to test the hypothesis that thespatial distribution of plants with symptoms ofAGYd, RGd or LSLCd was random (Cliff & Ord,1973). The join-count statistic was appropriate forthis study because it tests the spatial autocorrelationof nominal level (presence/absence) spatial data. Inthis study a join-count statistic was calculated forthe number of times diseased grapevines occupiedadjoining locations and for the number of timesdiseased and �healthy� grapevines occupiedadjoining locations in the study sites, either in therow/column direction (Rooks case) or in the diagonaldirection (Bishops case). If a significant result isobtained on both the row/column direction and inthe diagonal direction clustering can be inferred. Asignificant result obtained in either the row/columndirection or the diagonal direction would indicate adirection of the spread of disease.

Temporal analysis of disease survey dataFor a given year, each grapevine at each study site

had the potential to have one of eight combinationsof RGd, AGYd and LSLCd presence or absence.Therefore, for a given year, the number of grapevinesat each study site that had a specific combination of

Table 1. The location, the variety grown, the year the vineyard was established, the size of the survey area and theduration of the survey at each vineyard used in this study

a All areas surveyed at Wemen belong to one vineyard

Location Variety Vineyard establishment year Vineyard survey area Duration of the survey

Gol Gol, Sunraysia,New South Wales

Chardonnay 1992 and 1993 18 × 50 grapevines January 1996 � April 2001

Karadoc, Sunraysia,Victoria

Chardonnay 1992 18 × 50 grapevines January 1996 � April 2001

3 blocks of 10 × 10 grapevines January 1996 � April 2001Wemena, Sunraysia,Victoria

Chardonnay 19934 blocks of 10 × 25 grapevines October 1996 � April 2001

Colignan, Sunraysia,Victoria

Chardonnay 1996 20 × 20 grapevines October 1998 � April 2001

Paringa, Riverland, NSW Shiraz 1998 2 blocks of 20 × 55 grapevines April 1999 � April 2001

208 F E CONSTABLE ET AL.

were planted.The incidence of RGd in each year was different

at each study site (Fig. 1). The pattern of RGdincidence from year to year was different betweenstudy sites. RGd was not observed at Paringa. The

cumulative incidence of RGd in each vineyardincreased over the survey period (Fig. 1). By 2000/01 the cumulative incidence of RGd at each studysite was: 68.9 % at Gol Gol; 73.1% at Karadoc; 76.7at Wemen (6 yr); 54.7% at Wemen (5 yr); 16.8% at

Wemen(5 yr)

Coligan

Gol Gol

Karadoc

Wemen(6 yr)

Paringa

Fig. 1. The yearly and cumulative incidence (%) of grapevines affectedby AGYd, RGd and LSLCd over six seasons at Gol Gol, Karadoc andWemen, 5 yr at Wemen and 3 yr at Colignan and Paringa. Bars = yearly;lines = cumulative.

209Distribution of Australian grapevine diseases

Colignan.New occurrences of RGd were observed in

grapevines that were planted during the survey. RGdaffected 16/151 replacement grapevines at Gol Gol,6/9 replacement grapevines at Karadoc and 1/7replacement grapevines at Wemen.

The incidence of LSLCd in each year was differentat each study site (Fig. 1). The pattern of LSLCdincidence from year to year was different betweenstudy sites. LSLCd was not observed at Paringa. Thecumulative incidence of LSLCd in each vineyardincreased over the survey period (Fig. 1). By 2000/01 the cumulative incidence of LSLCd at each studysite was: 60.9% at Gol Gol; 61.9% at Karadoc;36.7% at Wemen (6 yr); 34.7% at Wemen (5 yr);54% at Colignan

New occurrences of LSLCd were observed ingrapevines that were planted during the survey.LSLCd affected 32/151 replacement grapevines atGol Gol, 6/9 replacement grapevines at Karadoc and1/7 replacement grapevines at Wemen.

Expression of AGYd in Shiraz grapevines atParinga

At Paringa AGYd was not observed in spring orearly summer. AGYd was first observed in Shirazgrapevines at Paringa from late summer andincreased in incidence until leaf fall. The symptomswere similar to those described for white varietiesexcept that reddening was observed instead ofyellowing (Magarey, 1986; Magarey & Wachtel,1983, 1986a). Bunches were not affected.

Disease expression with timeTable 2 shows the proportion of grapevines

affected with AGYd, RGd and LSLCd that wereobserved for the entire survey period at each site.The proportion of grapevines displaying AGYddecreased at Karadoc, Wemen and Colignan as thenumber of consecutive years for which grapevines

were affected by each disease increased. At Gol Golthe proportion of AGYd affected grapevines wassimilar for 1 yr and up to 4 consecutive yr and thenthe proportion decreased when grapevines displayedthe disease for 5 or 6 consecutive yr. At all vineyardsthe proportion of grapevines displaying RGd orLSLCd decreased as the number of consecutive yearsfor which grapevines were affected by each diseaseincreased.

No consistent pattern of AGYd, RGd or LSLCdexpression was observed for grapevines thatdisplayed the diseases for 2 or more yr (data notshown). AGYd, RGd and LSLCd were not alwayspersistently expressed in the same grapevine fromseason to season.

Spatial analysis of AGYd, RGd and LSLCdA fixed grid analysis was used to determine

whether AGYd, RGd or LSLCd affected grapevineswere clustered and the size of clusters. When thenumber of subunits in the grid laid across the plotwas less than 15 the results were not included.

AGYd affected grapevines were significantlyclustered in all years at Gol Gol and Paringa andonly in some years at Karadoc and Colignan (Table3). The cluster sizes of AGYd affected grapevineswere different between most years in each vineyard.

RGd affected grapevines were significantlyclustered in all years at Karadoc and Gol Gol and in1999/00 and 2000/01at Colignan (Table 3). Thecluster sizes of RGd affected grapevines weredifferent between most years in each vineyard.

LSLCd affected grapevines were significantlyclustered in all years at Gol Gol and Karadoc (Table3). The cluster sizes of LSLCd affected grapevineswere different between most years in each vineyard.The distribution of LSLCd affected grapevines atColignan was random.

The results for spatial autocorrelation for AGYd,RGd and LSLCd are given in Tables 4, 5 and 6

Gol Gol Karadoc Wemen (6 yr) Wemen (5 yr) Colignan ParingaYears affected AGY RG LSLC AGY RG LSLC AGY RG LSLC AGY RG LSLC AGY RG LSLC AGY RG LSLC

6 1.8 0.3 0.1 0 0.9 0.3 0 0 0 - - - - - - - - -5 8.3 1.4 1.3 0.3 2.9 2.1 0.7 0 0 0.1 0 0 - - - - - -4 19.8 6.3 2.9 3.8 8.5 5.9 1.7 1 0.3 1.9 0.4 0 - - - - - -3 19.7 17.4 7.6 8.7 17.2 10.9 6.7 8.4 0.3 8 1.2 0.7 4.6 0.3 0.5 0 0 02 19.5 23.5 18.7 24.8 21.4 15.2 17.1 22.1 6.4 15.9 5.1 5.2 38.4 0.5 9.7 0.5 0 01 17.8 30.1 36 38.6 25.1 26 27.9 44.6 29.2 30.5 48.7 31.8 51.9 13.4 46.2 9.1 0 00 13.1 21 33.3 23.7 24 39.4 46 23.8 63.8 43.6 44.6 62.3 5.1 85.8 43.5 90.4 100 100

Total % affected 86.9 79 66.7 76.3 76 60.6 54 76.2 36.2 56.4 55.4 37.7 94.9 14.2 56.5 9.6 0 0

Table 2. The percentage of grapevinesa at Gol Gol, Karadoc, Wemen, Colignan and Paringa that were affectedwith AGYd, RGd or LSLCd for 0-6 yr and the total percentage that had displayed the diseases at sometime during

the survey period

a The total number of grapevines observed at each site for the entire survey period: Gol Gol = 763 grapevines, Karadoc = 860grapevines, Wemen (6 yr) = 298 grapevines, Wemen (5 yr) = 984 grapevines, Colignan (3 yr) = 372 grapevines, Paringa (3 yr) = 2170grapevines

210 F E CONSTABLE ET AL.

respectively. When the total number of joinsassociated with a significant test result was less than5% of the total number of grapevines at the test site,the result was disregarded and is represented by theshaded cells in Tables 4-6.

At Gol Gol there was a significant associationbetween grapevines and neighbouring AGYd plantsin all years (Table 4). At Karadoc, Colignan and someareas at Wemen there was a significant associationbetween grapevines and neighbouring AGYd plantsin specific years. Clustering of AGYd was notassociated with the disease status of the testgrapevine because significant associations often

occurred between both symptomless and AGYdaffected grapevines and neighbouring AGYdaffected plants in the same years in each vineyard.At Wemen and Karadoc significant associationsoccurred more frequently in the column/rowdirection. No directional trends in the distributionof AGYd were observed at Gol Gol or Colignan.

At Karadoc there was a significant associationbetween grapevines and neighbouring RGd plantsin each year (Table 5). At Gol Gol there was asignificant association between grapevines and RGdaffected neighbours from 1996/97 to 1999/00. Insome areas at Wemen there was a significant

Subunit size � the number of grapevines in row × columnSite Disease 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01

AGYd 2 × 15 2 × 14 9 × 5 9 × 5 6 × 8 9 × 6RGd 4 × 12 3 × 8 2 × 22 6 × 7 9 × 6 9 × 6Gol Gol

LSLCd 2 × 15 2 × 15 3 × 12 3 × 12 9 × 6 9 × 6AGYd 6 × 8 3 × 16 2 × 13 7 × 4 3× 12 -RGd 6 × 8 4 × 12 4 × 12 6 × 8 9 × 5 2 × 23Karadoc

LSLCd 3 × 16 2 × 24 6 × 7 2 × 24 6 × 7 3 × 15AGYd - 5 × 5 -

ColignanRGd - 6 × 4 -

Paringa block 1 AGYd 9 × 4 4 × 17 -Paringa block 2 AGYd 6 × 11 2 × 27 -

Table 3. The maximum subunit sizes in which significant (P < 0.05) clustering of diseased grapevines wasdetermined using a fixed grid analysis at each study site in each season

a Rooks = joins between a grapevines and its neighbours in the row and column direction; Bishops = joins between a grapevines andits neighbours in the diagonal directionb HA = tests between unaffected grapevines and AGYd affected neighboursc AA = tests between AGYd affected grapevines and AGYd affected neighboursd + = significant join count statistic at the 5% level with a z-test statistic = 1.96e + in a shaded square = significant join count statistic at the 5% level with a z-test statistic = 1.96 but the number of joins associatedwith the test result was less than 5% of the total number of vines at the test sitef ns = vineyard not surveyed

Table 4. The association between symptomless or AGYd affected grapevines and their AGYd affected neighboursin the row/column direction or the diagonal direction at Gol Gol, Karadoc, the four areas surveyed at Wemen (5

yr) and Colignan in each year

Gol Gol Karadoc Wemen Colignan1 2 3 4 1 2 3 4

Type of joina Year HAb AAc HA AA HA AA HA AA

1995/96 +d + + + e nsf ns ns ns ns ns ns ns ns ns1996/97 + + + + + + ns ns1997/98 + + + ns ns1998/99 + + +1999/00 + + + + + + + +

Rooks

2000/01 + + + + + +1995/96 + + ns ns ns ns ns ns ns ns ns ns1996/97 + + + + ns ns1997/98 + + ns ns1998/99 + + +1999/00 + + + + + +

Bishops

2000/01 + + +

211Distribution of Australian grapevine diseases

association between grapevines and neighbouringRGd plants in some years only. Clustering of RGdwas not associated with the disease status of the testgrapevine because significant associations oftenoccurred between both symptomless and RGdaffected grapevines and neighbouring RGd affectedplants in the same years at Gol Gol, Karadoc andsome areas at Wemen. In areas 1 and 2 at Wemensignificant associations occurred only betweensymptomless and neighbouring RGd affected plants.At Karadoc and Wemen significant associationsoccurred more frequently in the column/rowdirection. No directional trends in the distributionof RGd were observed at Gol Gol.

There was a significant association betweengrapevines and neighbouring LSLCd in each yearexcept 1999/00 at Gol Gol and 1997/98 at Karadoc(Table 6). At Colignan and some areas at Wementhere was a significant association betweengrapevines and neighbouring LSLCd plants in someyears only. Clustering of LSLCd was not associatedwith the disease status of the test grapevine becausesignificant associations often occurred between bothsymptomless and LSLCd affected grapevines andneighbouring LSLCd affected plants in the sameyears in all vineyards. Significant associationsoccurred more frequently in the column/rowdirection in all vineyards.

The association between AGYd and RGd orLSLCd

There was no pattern of association betweenAGYd and RGd only, AGYd and LSLCd only orAGYd and RGd/LSLCd in the same year in anyvineyard (Table 7). RGd did not affect anygrapevines with AGYd at Wemen (5 yr) in 2000/01or at Wemen (6 yr) in 1999/00 and 2000/01. LSLCddid not affect any grapevines with AGYd at Wemen(6 yr) in 1995/96. AGYd affected grapevines werenot affected by both RGd and LSLCd at Wemen (5yr) from 1998/99 to 2000/01 or at Wemen (6 yr) in1999/00 and 2000/01.

There was no pattern of association betweenAGYd and RGd in the following season in eachvineyard (Table 8). There were no grapevines withAGYd that displayed RGd in the following year atColignan in any year or at Wemen (6 yr) in 1995/96and 1999/00 and Wemen (5 yr) in 1999/00.

There was no pattern of association betweenAGYd and LSLCd in the previous season in eachvineyard (Table 9). There were no grapevines withAGYd that displayed LSLCd in the previous yearsat Wemen (6 yr) in 1996/97 and 2000/01.

AGYd affected some grapevines in the absenceof RGd and LSLCd at each vineyard during thesurvey period. AGYd alone affected 78 grapevinesat Gol Gol, 98 grapevines at Karadoc, 119 grapevinesat Wemen (5 yr), 16 grapevines at Wemen (6 yr) and

147 grapevines at Colignan.There was no pattern of association between RGd

and AGYd only or AGYd and LSLCd in the sameyear in any vineyard (Table 10). AGYd did not affectgrapevines with RGd at Wemen (5 yr) in 2000/01 orat Wemen (6 yr) in 1995/96, 1999/00 and 2000/01.

There was no pattern of association between RGdand AGYd in the previous seasons in any vineyard(Table 11). A high proportion of RGd affectedgrapevines that also had AGYd in previous yearswas observed at Colignan compared to othervineyards, particularly in 2000/01, although thenumber of grapevines observed with RGd in 2000/01 was only four. At Wemen (6 yr) in 1995/96 and2000/01, Wemen (5 yr) in 2000/01 and Colignan in1999/00 and 2000/01 grapevines with RGd were notaffected by AGYd in the previous season.

RGd affected some grapevines in the absence ofAGYd and LSLCd at each vineyard during thesurvey period. RGd alone affected 22 grapevines atGol Gol, 61 grapevines at Karadoc, 158 grapevinesat Wemen (5 yr), 80 grapevines at Wemen (6 yr) andeight grapevines at Colignan.

There was no pattern of association betweenLSLCd and AGYd only or AGYd and RGd inaffected grapevines in the same year in any vineyard(Table 12). AGYd did not affect grapevines withLSLCd at Wemen (6 yr) in 1995/96.

There was no pattern of association betweenLSLCd and AGYd in the following year in anyvineyard (Table 13). Grapevines with LSLCd werenot affected by AGYd in the following season atWemen (6 yr) in 1995/96 and Wemen (5 yr) in 1999/00.

LSLCd affected some grapevines in the absenceof AGYd and RGd at each vineyard during thesurvey. LSLCd alone affected 29 grapevines at GolGol, 15 grapevines at Karadoc, 39 grapevines atWemen (5 yr), four grapevines at Wemen (6 yr) and10 grapevines at Colignan.

Temporal analysis of AGYd, RGd and LSLCdusing log-linear models

Log-linear models were used to determine if therewere significant interactions between AGYd, RGdand LSLCd and elapsed time at each vineyard.Significant interactions indicated that the expectednumber of vines for each combination (absence/presence) of the three diseases is significantlydifferent relative to each other and/or significantlydifferent from season to season if year had asignificant interaction with the diseases.

At Wemen the incidence of RGd changedsignificantly with time (P < 0.001).

The incidence of LSLCd affected grapevineschanged significantly with time at Gol Gol (P <0.001), Karadoc (P < 0.041) and Wemen (P < 0.001).

There was a significant interaction between the

212 F E CONSTABLE ET AL.

presence of AGYd and RGd and this interactionchanged with time at Gol Gol (P = 0.035), Karadoc(P = 0.033) and Colignan (P = 0.02). That is, theexpected number of unaffected grapevines,grapevines affected by AGYd and RGd, AGYd aloneand RGd alone were significantly different from eachother in some years at each vineyard.

At Wemen there was an interaction between thepresence of AGYd and RGd (P = 0.007). Thegraphical representation of the data suggested thatthe expected number of unaffected grapevines was

significantly greater than the number of grapevinesaffected by one or both diseases for the entire surveyperiod (data not shown). The incidence of RGd didnot increase with the presence of AGYd.

At Colignan there was an interaction between thepresence of AGYd, RGd and LSLCd (P = 0.052).The graphical representation of the data suggestedthat the expected number of grapevines unaffectedby RGd was significantly greater than the numberof grapevines affected by RGd for the entire surveyperiod, regardless of the presence of AGYd and

Table 5. The association between symptomless or RGd affected grapevines and their RGd affected neighbours inthe row/column direction or the diagonal direction at Gol Gol, Karadoc, the four areas surveyed at Wemen (5 yr)

and Colignan in each year

a HR = tests between unaffected grapevines and RGd affected neighboursb RR = tests between RGd affected grapevines and RGd affected neighboursc na = not appropriate to test because the number of affected grapevines were < 2, therefore spatial autocorrelation was not estimatedd ns = vineyard not surveyed

Gol Gol Karadoc Wemen Colignan1 2 3 4 1 2 3 4

Type of join Year HRa RRb HR RR HR RR HR RR

1995/96 + + + nsd ns ns ns ns ns ns ns ns ns1996/97 + + + + + + + + ns ns1997/98 + + + + + + ns ns1998/99 + + + + nac + na + +1999/00 + + + + na + na +

Rooks

2000/01 + + + + na na na na na na na na +1995/96 + + ns ns ns ns ns ns ns ns ns ns1996/97 + + + ns ns1997/98 + + ns ns1998/99 + + na na +1999/00 + + na na

Bishops

2000/01 na na +

Table 6. The association between symptomless or LSLCd affected grapevines and their LSLCd affectedneighbours in the row/column direction or the diagonal direction at Gol Gol, Karadoc, the four areas surveyed at

Wemen (5 yr) and Colignan in each year

a HL = tests between unaffected grapevines and LSLCd affected neighboursb LL = tests between LSLCd affected grapevines and LSLCd affected neighboursc ns = vineyard not surveyed

Gol Gol Karadoc Wemen Colignan1 2 3 4 1 2 3 4

Type of join Year HLa LLb HL LL HL LL HL LL

1995/96 + + + + nsc ns ns ns ns ns ns ns ns ns1996/97 + + + + + + + + ns ns1997/98 + + + + ns ns1998/99 + + + + + + + +1999/00 + + +

Rooks

2000/01 + + + +1995/96 + + + + ns ns ns ns ns ns ns ns ns ns1996/97 ns ns1997/98 + ns ns1998/99 + +1999/00 + +

Bishops

2000/01 + + + + +

213Distribution of Australian grapevine diseases

grapevines affected by one or both diseases for theentire survey period. The incidence of LSLCd didnot increase with the presence of AGYd.

At Colignan there was an interaction between thepresence of AGYd and LSLCd and this interactionchanged with time (P = 0.009). That is, the expectednumber of unaffected grapevines, grapevinesaffected by AGYd and LSLCd, AGYd alone and

LSLCd. The incidence of LSLCd did not increasewith the presence of AGYd.

There was an interaction between the overallpresence of AGYd and LSLCd at Karadoc (P =0.004) and Wemen (P = 0.011). For both vineyardsthe graphical representation of the data suggestedthat the expected number of unaffected grapevineswas significantly greater than the number of

Vineyard Disease 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01

RGd 10.8 29.4 65.7 48.6 7.9 5.7LSLCd 18.7 13.8 8.0 6.4 11.5 13.1Gol Gol

RGd/LSLCd 1.8 16.7 16.7 10.9 1.0 1.7RGd 11.1 27.0 35.7 20.3 27.4 8.1

LSLCd 11.1 21.6 12.5 18.7 12.9 23.9KaradocRGd/LSLCd 38.9 24.3 33.9 35.7 3.1 6.4

RGd 0 35.9 40.8 30.9 0 0LSLCd 0 7.8 14.5 1.8 10.7 14.0

Wemen(6 yr)

RGd/LSLCd 0 0 10.5 3.6 0 0RGd 44.8 15.3 4.7 5.6 0

LSLCd 8.4 17.8 7.0 4.0 10.6Wemen(5 yr)

RGd/LSLCd 11.7 2.5 0 0 0RGd 14.5 9.2 1.6

LSLCd 5.5 11.6 9.8ColignanRGd/LSLCd 3.6 1.0 0.5

Table 7. The proportion (%) of AGYd affected grapevines that also had RGd, LSLCd or RGd and LSLCd in eachvineyard in each season

Vineyard Disease 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01

AGYd 41.9 40.6 28.7 48.6 39.3 38.3Gol Gol

AGYd /LSLCd 7.0 23.2 7.2 10.9 4.8 11.7AGYd 0.7 2.6 5.5 21.0 56.1 32.7

KaradocAGYd /LSLCd 2.4 2.3 5.2 36.9 6.3 25.7

AGYd 0 19.5 35.6 14.2 0 0Wemen(6 yr) AGYd /LSLCd 0 0 9.2 1.7 0 0

AGYd 22.0 34.1 21.6 36.8 0Wemen(5 yr) AGYd /LSLCd 5.7 5.5 0 0 0

AGYd 32.0 64.3 75.0Colignan

AGYd /LSLCd 8.0 7.1 25.0

Table 10. The proportion (%) of RGd affected grapevines that also had AGYd or AGYd and LSLCd in eachvineyard in each season

Vineyard 1995/96 1996/97 1997/98 1998/99 1999/00

Gol Gol 48.8 53.9 64.9 5.6 5.3Karadoc 66.7 67.6 50.0 20.3 13.2Wemen (6 yr) 0 34.4 38.2 1.8 0Wemen (5 yr) 15.9 4.5 2.3 0Colignan 0 0

Table 8. The proportion (%) of grapevines withAGYd that also displayed RGd in the following

season at each site for each season of the survey

Table 9. The proportion (%) of grapevines withAGYd that also displayed LSLCd in the previousseason at each site for each season of the survey

Vineyard 1996/97 1997/98 1998/99 1999/00 2000/01

Gol Gol 25.5 33.3 24.7 20.8 14.8Karadoc 29.7 73.2 44.0 39.5 20.8Wemen (6 yr) 0 26.3 20.0 3.6 10.0Wemen (5 yr) 18.8 5.3 3.2 0Colignan 47.3 17.6

214 F E CONSTABLE ET AL.

sequential seasons. Very few grapevines displayedAGYd every year. Remission of AGYd has beenreported previously (Magarey & Wachtel, 1986b;Constable et al., 2000). Remission or recovery fromgrapevine yellows symptoms, similar to AGYd, inaffected vines in Europe is also observed (Caudwell,1961, 1964; Osler et al., 1993). The cause ofremission and recurrence of AGYd is not known butone study has shown that phytoplasmas maypersistently infect grapevines and that AGYdexpression may be related to the number and locationof phytoplasmas present in a grapevine (Constableet al., 2003).

In each season there were new records of AGYdin grapevines that were established prior to thesurvey as well as in grapevines replaced during thesurvey. Chardonnay grapevines can display AGYdin the first year after planting but this occurred lessfrequently compared to grapevines that had beenestablished for 2 or more yr. Since AGYd isassociated with phytoplasmas this suggests that newphytoplasma infections may occur in each year.However, as symptomless phytoplasma infectionscan occur (Gibb et al., 1999; Constable et al., 2003),the expression of disease due to previousphytoplasma infection events in grapevinesestablished prior to the survey and infected plantingmaterial cannot be dismissed.

The pattern of incidence of AGYd was differentbetween vineyards for the duration of the surveyindicating that local factors may have had an

LSLCd alone were significantly different from eachother in some years.

There was a significant interaction between thepresence of RGd and LSLCd at Gol Gol (P < 0.001)and Karadoc (P < 0.001). The graphicalrepresentation of the data suggested that the expectednumber of unaffected grapevines was significantlygreater than the number of grapevines affected byone or both diseases for the entire survey period.The incidence of RGd did not increase with thepresence of LSLCd.

Log-linear modeling was not done for Wemen (6yr) due to the small number of grapevines in eachblock. Log linear modeling was not done for Paringabecause the majority of grapevines were not affectedby AGYd in 1998/99 and 1999/2000 and there wasno change in behaviour between the 2 years.

Discussion

The expression of AGYdAGYd affected many grapevines within individual

vineyards. AGYd in Sunraysia vineyards wascharacterised by the expression of disease followedby remission in some grapevines and recurrence ofdisease in others as well as the occurrence of AGYdin previously unaffected grapevines. Somegrapevines expressed AGYd in only one growingseason while some grapevines expressed AGYd intwo or more seasons. However, when AGYd wasrecurrent the disease was not always expressed in

Vineyard 1996/97 1997/98 1998/99 1999/00 2000/01

Gol Gol 25.7 58.6 28.8 26.2 36.7Karadoc 3.1 6.9 15.9 19.6 45.1Wemen (6 yr) 0 25.3 24.2 7.7 0Wemen (5 yr) 41.8 24.3 21.1 0Colignan 0 0

Table 11. The proportion (%) of grapevines withRGd that also displayed AGYd in the previous

season at each site for each season of the survey

Vineyard 1995/96 1996/97 1997/98 1998/99 1999/00

Gol Gol 52.9 23.8 54.5 62.1 70.6Karadoc 6.1 17.8 61.5 63.4 70.4Wemen (6 yr) 0 58.8 26.2 5.9 38.5Wemen (5 yr) 20.3 8.7 8.7 0Colignan 84.1 61.8

Table 13. The proportion (%) of grapevines withLSLCd that also displayed AGYd in the followingseason at each site for each season of the survey

Vineyard Disease 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01

AGYd 14.8 24.6 7.9 17.9 56.5 60.2Gol Gol

RGd/AGYd 1.4 29.9 16.3 30.7 6.7 8.0AGYd 1.1 3.5 5.4 14.2 37.0 50.9

KaradocRGd/AGYd 3.8 3.9 14.6 27.1 8.9 13.6

AGYd 0 14.7 26.2 5.9 23.1 31.8Wemen(6 yr) RGd/AGYd 0 0 19.0 11.8 0 0

AGYd 10.7 35.0 26.1 17.2 44.7Wemen(5 yr) RGd/AGYd 15.0 4.9 0 0 0

AGYd 1.8 7.3 52.3Colignan

RGd/AGYd 1.2 5.5 2.8

Table 12. The proportion (%) of LSLCd affected grapevines that also had AGYd or AGYd and RGd in eachvineyard in each season

215Distribution of Australian grapevine diseases

important influence on disease expression andincidence. Although the incidence of AGYdfluctuated in most vineyards from year to year thelog-linear models indicated that the incidence ofAGYd, on average, was not significantly differentfrom year to year suggesting that AGYd is endemic.The fluctuation in year to year incidence of AGYdwas in contrast to previous research, which showedthat the incidence of AGYd may be temporarily highin one year only in some Chardonnay vineyards(Magarey & Wachtel, 1986b).

The expression of AGYd in Shiraz at Paringa wasdifferent to Chardonnay and Riesling grapevines inother vineyards. In Chardonnay and Riesling AGYdappears from late spring and the incidence of affectedgrapevines increases until harvest (Magarey, 1986;Magarey & Wachtel, 1983, 1986a). AGYd appearedin Shiraz grapevines at Paringa from late summerand the incidence of affected grapevines increaseduntil leaf fall. The symptoms were similar to thosedescribed for white varieties except that reddeningreplaced yellowing (Magarey, 1986; Magarey &Wachtel, 1983, 1986a). Because the AGYd firstappeared in the Shiraz grapevines close to harvest,bunches were not affected. Surveys of Shirazvineyards affected by AGYd in the same and otherregions should be done to determine if the lateexpression of AGYd observed at Paringa is typicalof the variety.

The largest proportion of Shiraz grapevines thatwere affected by AGYd occurred in the first yearafter planting, then the incidence of AGYd declinedand in 2001 no disease was observed. Of the AGYdaffected grapevines that were not removed from thevineyard in the first year of observation, very fewgrapevines displayed AGYd in the subsequentseasons. Consequently, it is unlikely that the removalof grapevines significantly reduced the incidence ofdisease in the following season. In 1999 anotherblock of Shiraz was planted behind the block thatwas surveyed. The planting material used to establishthe block was from a different nursery and hot watertreated to reduce the incidence of various pathogensincluding phytoplasmas. Hot water treatment iseffective for reducing the incidence of phytoplasmatransmission through dormant wood in othercountries (Caudwell et al., 1997, Bianco et al., 2000).AGYd also affected some of the grapevines in thenew block of Shiraz in 1999/00. This observationand the observation of new AGYd incidence in theblocks that were surveyed suggests that, ifphytoplasmas were the cause of AGYd, they mayhave been introduced to the vineyard by aerialtransmission rather than through infected plantingmaterial.

The spatial distribution of AGYdThe fixed grid analysis and spatial autocorrelation

showed that some clustering of AGYd affectedgrapevines occurred at each vineyard in most years.However, the size of the groups in which clusteringAGYd affected grapevines was observed wasdifferent in the same vineyard from year to year andbetween vineyards. Remission, recurrence and newobservations of AGYd would also have contributedto the changes in the spatial distribution of AGYdaffected grapevines in each vineyard. Spatialautocorrelation showed that the presence of AGYdin grapevines was not associated with the diseasestatus of immediate neighbours in the same year.This indicates that phytoplasmas may not spreadfrom a diseased grapevine to neighbouringgrapevines and cause disease in the same year.

Symptomless phytoplasma infections can occur(Gibb et al., 1999; Constable et al., 2003), so thespatial distribution of AGYd cannot be used as anaccurate indicator of the spatial distribution ofphytoplasmas. It would be more useful to determinethe presence or absence of phytoplasmas in eachgrapevine in an area by PCR detection and do spatialanalyses on such data to see how phytoplasmas canspread within a vineyard.

The expression of RGdRGd was present in every Chardonnay vineyard

that was surveyed in Sunraysia and was characterisedby remission and recurrence of the disease as wellas new observations of RGd. In each season therewere new records of RGd both in establishedgrapevines and in new replacement grapevines.However, at Gol Gol and Karadoc a larger numberof grapevines were sometimes observed displayingRGd in 2 or 3 subsequent years, rather thandisplaying RGd over several years with remissionoccurring in between years of expression. Theproportion of grapevines displaying RGd symptomsfor more than 1 yr decreased with the number ofyears of observation. Very few grapevines displayedRGd in every year of the survey.

Remission, recurrence and new RGd incidencewould explain the fluctuation of RGd incidence fromyear to year that occurred in most vineyards. In thefirst 3 to 4 years of the survey, none of the vineyardsthat were surveyed showed the same pattern of RGdincidence from year to year, suggesting that localfactors can affect the expression of RGd in vineyardsover time. However, from 1997/98 to 1998/99 theincidence of RGd declined in most vineyards andthis indicated that local and regional factors mayinfluence the expression of RGd. RGd may havemany causes including pathogens such as virus(Bovey & Martelli, 1992), pests such as mites(Bernard et al., 2001) and environmental factors suchas winter drought and winter chilling (Meiering etal., 1980; Wilson, 1997; Wilson & Hayes, 2000).The contribution of these factors may vary from year

216 F E CONSTABLE ET AL.

to year, influencing the disease status of grapevines.The log-linear model for Wemen indicated that RGdincidence was significantly different with time andthe incidence graph (Fig. 1) shows that RGdincidence was greatest in 1996/97 compared to otheryears. Thus, RGd is endemic in most Chardonnayvineyards used for this study.

The spatial distribution of RGdBoth fixed grid analysis and spatial autocorrelation

indicated that RGd affected grapevines wereclustered in all years at Karadoc and Gol Gol.Clustering of RGd affected grapevines did notalways occur in each year at Colignan. Whenclustering of RGd was observed, the size of thegroups of grapevines in which clustering wasobserved was often different between years in anyvineyard. It is likely that the remission andrecurrence of RGd as well as new occurrences affectthe spatial arrangement of RGd affected grapevinesin each year. The presence of RGd in grapevineswas not associated with the disease status ofimmediate neighbours in the same year in mostvineyards. This indicates that the cause of RGd maynot spread from a diseased grapevine toneighbouring grapevines to cause disease in the sameyear.

The association between AGYd and RGdThe proportion of AGYd affected grapevines that

displayed RGd in the same season, the previousseason or the following season was different betweenvineyards and years. The proportion of grapevineswith RGd that displayed AGYd in the same seasonor the previous season was also variable betweenvineyards and years. Also, AGYd and RGd can occurindependently of one another. Consequently, aphytoplasma aetiology cannot be inferred for RGdin all grapevines.

The log-linear models also indicated that the yearto year pattern of incidence of grapevines with bothAGYd and RGd was different to the year to yearpattern of incidence of AGYd alone and RGd aloneindicating that the two diseases were not alwaysassociated. At Wemen, the log-linear model showedthat the overall incidence of RGd during the vineyardsurvey did not increase when in combination withAGYd indicating that the presence of AGYd andRGd over time were not associated. Other studiessuggested that RGd expression in AGYd affectedgrapevines might be related to phytoplasma titre(Constable et al., 2003). The inconsistency of RGdexpression in AGYd affected grapevines betweenseasons may be a reflection of differences inphytoplasma titre between seasons.

The expression of LSLCdLSLCd was present in every Chardonnay vineyard

that was surveyed in Sunraysia and was alsocharacterised by remission and recurrence of diseaseas well as new observations of disease. In eachseason there were new records of LSLCd both inestablished grapevines and in new replacementgrapevines. Some grapevines expressed LSLCd inone growing season only while others expressedLSLCd in more than one season. However, LSLCddoes not always occur sequentially. Very fewgrapevines displayed LSLCd every year.

The log-linear models for Gol Gol, Karadoc andWemen showed that the incidence with LSLCd wassignificantly different between years. However, theyears in which the incidence of LSLCd increasedand decreased at each vineyard was different. Thisresult indicates that local factors may affectexpression of LSLCd. Remission, recurrence andnew LSLCd incidence would explain the fluctuationof LSLCd incidence from year to year that occurredin most vineyards.

The spatial distribution of LSLCdBoth the fixed grid analysis and spatial

autocorrelation indicated that clustering of LSLCdaffected grapevines occurred in each year at Gol Goland Karadoc only. However the fixed grid analysisindicated that, when clustering was observed, thesize of the groups of grapevines in which clusteringoccurred were different from year to year. It is likelythat the remission and recurrence of LSLCd as wellas new occurrences of LSLCd affect the spatialarrangement of LSLCd affected grapevines in eachyear.

Significant associations between grapevines andneighbouring LSLCd plants occurred morefrequently in the column/row direction in allvineyards. While spatial autocorrelation does notindicate if the direction trend occurs in either therow direction or column direction, grapevines arein closer proximity along trellises, which is thecolumn direction indicated in the analysis. Thus, itis possible that the cause of this disease in somevineyard spreads in this direction from vine to vine.A statistical analysis that determines gradients ofdisease in the column and row direction could havebeen used to determine a directional trend forLSLCd. Significant associations were observedbetween both symptomless and diseased grapevinesand neighbouring LSLCd plants. This resultindicates that the disease status of a grapevine maynot influence the disease status of neighbours in thatyear.

The association between AGYd and LSLCdThe results presented here suggest that AGYd and

LSLCd can occur independently of one another. Thissuggests that AGY phytoplasma may not always beassociated with LSLCd or indeed any phytoplasma.

217Distribution of Australian grapevine diseases

However the TBB phytoplasma has been detectedin grapevines with LSLCd (Gibb et al., 1999)therefore it is possible that some diseased vines werephytoplasma positive. The proportion of LSLCdaffected grapevines displaying AGYd in the sameyear or following years was different from year toyear and between vineyards. The proportion ofAGYd affected grapevines displaying LSLCd in thesame years or previous year was also different fromyear to year and between vineyards. These resultssupport the conclusions from other studies(Constable et al., 2003) and indicate that if LSLCdis associated with phytoplasmas the disease is notalways the first expression of phytoplasma infection.The Log-linear models indicated that the year to yearpattern of incidence of grapevines with both AGYdand LSLCd at Colignan was different to the year toyear pattern of incidence of AGYd alone and LSLCdalone indicating that the two diseases are not alwaysassociated. This suggests that the overall incidenceof LSLCd during the vineyard surveys did notincrease when in combination with AGYd alsoindicating that the two diseases are not alwaysassociated.

Summary

In summary, AGYd, RGd and LSLCd in SunraysiaChardonnay vineyards was characterised by theexpression of disease followed by remission in somegrapevines and recurrence of disease in others aswell as the occurrence of disease in previouslyunaffected grapevines. The pattern of incidence ofeach disease was different between vineyards forthe duration of the survey indicating that local factorsmay have had an important influence on diseaseexpression and incidence. RGd and LSLCd did notaffect the Shiraz grapevines at Paringa. AGYdexpression in Shiraz grapevines at Paringa was laterin the season compared to Chardonnay and very littlerecurrence was observed indicating that the varietyShiraz responds differently to phytoplasmainfections. The results presented here provide noinsight as to the actual cause of RGd or LSLCd.Some grapevines exhibited a combination of AGYdand RGd or AGYd and LSLCd, however, somegrapevines that exhibited RGd or LSLCd were neveraffected by AGYd. Additionally, the log-linearmodels showed that RGd and LSLCd were notalways associated with AGYd. Thus, it is possiblethat phytoplasmas are not the cause of RGd and theirassociation is coincidental. It is also possible thatLSLCd is not associated with phytoplasmas but theTBB phytoplasma has been detected in diseasedgrapevines and may be a factor in the aetiology ofthis disease.

Acknowledgements

We gratefully acknowledge the financial supportof the Australian Research Council, SouthcorpWines, Chalmers Nurseries and Adelaide University.Survey data, from 1995/96 to 1997/98, fromSunraysia were collected during a 3 yr projectentitled �National program for the management ofphytoplasmas in Australian grapevines�. That projectwas funded by the Grape and Wine Research andDevelopment Corporation, the Dried Fruits Researchand Development Council and the Department ofNatural Resources and Environment in Victoria,through the Collaborative Research Centre forViticulture. Many thanks to Tim Griffin (Institutefor Horticultural Development) and Linda Pollock(Sunraysia Horticultural Centre) for assistanceduring fieldwork. We also thank Prof. John Randles(University of Adelaide) for many useful discussionsand for reviewing the manuscript. Thank you to DrMichael Maixner for providing the program Patchyand also for useful discussions.

References

Bernard M, Horne P A, Hoffman A A. 2001. Preventingrestricted spring growth (RSDG) in grapevines by successfulrust mite control � spray application, timing and eliminatingsprays harmful to rust mite harmful to predators are critical.The Australian Grapegrower and Winemaker 452:16-22.

Bianco P A, Fortusini A, Scattini G, Casati P, Carraro S,Torresin G C. 2000. Prove di risanamento di materialeviticolo affetto da Flavescenza dorata mediante termoterapia.Informatore Fitopatologico 50:43-49.

Bonfiglioli R G, Guerrini S, Symons R H. 1996. CooperativeResearch Centre for Viticulture: Sampling program forgrapevine yellows diseases. The Australian Grapegrower andWinemaker 394:22-24.

Bonfiglioli R G, Magarey P A, Symons R H. 1995. PCRconfirms an expanded symptomatology for AustralianGrapevine Yellows. Australian Journal of Grape and WineResearch 1:71-75.

Bonfiglioli R G, Carey C T, Schleifert L F, Kinnear A J,Symons R H. 1997. Description and progression of symptomsassociated with grapevine yellows disease in youngChardonnay vines in the Sunraysia district. The AustralianGrapegrower and Winemaker 400:11-15.

Bovey R, Martelli G P. 1992. Directory of Major Virus andVirus-Like Diseases of Grapevines. Mediterranean FruitCrop Improvement Council (MFCIC), pp. 41-51.

Caudwell A. 1961. Les phénomènes de rétablissement ches laFlavescence dorée de la vigne. Annals de Epiphyties 12:347-354.

Caudwell A. 1964. Identification d�une nouvelle maladie à virusde la vigne, la flavescence dorée, étude des phénomènes delocalisation des symptômes et de rétablissement. Annales desEpiphyties 15:1-193.

Caudwell A, Larrue J, Boudon-Padieu E, McLean G D. 1997.Flavescence doree elimination from dormant wood ofgrapevines by hot-water treatment. Australian Journal ofGrape and Wine Research 3:21-25.

Cliff A D, Ord J K. 1973. Spatial Autocorrelation. London:Pion Limited.

Constable F E, Gibb K S, Moran J R, Wilson Y M. 1998.Incidence of phytoplasma associated with yellows, restricted

218 F E CONSTABLE ET AL.

spring growth and late season leaf curl symptoms ingrapevines. The Australian Grapegrower and Winemaker409:19-20.

Constable F E, Gibb K S, Symons R H. 2003. The seasonaldistribution of phytoplasmas in Australian grapevines. PlantPathology 52:267-276.

Constable F E, Gibb K S, Moran J R, Wilson Y M. 2000.The relative risk of an association between Australiangrapevine yellows, restricted growth and late season leafcurl symptoms and persistence of each disorder in affectedgrapevines. In National Program for the Management ofPhytoplasmas in Australian Grapevines, GWRDC FinalReport, Chapter 1.2. Project Number: CRCV 95/2.

Constable F E, Whiting J R, Jones J, Gibb K S, Symons RH. 2002. A new grapevine yellows phytoplasma from theBuckland Valley of Victoria, Australia. Vitis 41:147-154.

Gibb K S, Constable F E, Moran J R, Padovan A C. 1999.Phytoplasmas in Australian grapevines - Detection,differentiation and associated diseases. Vitis 38:107-114.

Magarey P A. 1986. Grapevine yellows � aetiology,epidemiology and diagnosis. South African Journal ofEnology and Viticulture 7:90-100.

Magarey P A, Wachtel M F. 1983. Australian Vine Yellows, I.Inference of a mycoplasma-like aetiology for a new diseaseof Australian grapevine. 4th International Congress of PlantPathology, Melbourne, Australia, pp.118.

Magarey P A, Wachtel M F. 1986a. Grapevine Yellows - awidespread, apparently new disease in Australia. PlantDisease 70:694.

Magarey P A, Wachtel M F. 1986b. Australian GrapevineYellows. International Journal of Tropical Plant Disease 4:1-14.

Magarey P A, Plavsic B, Wachtel M F. 1988. MLO associatedwith Australian grapevine yellows diseased phloem cells.International Journal of Tropical Plant Diseases 6:175-179.

Maixner M. 1993. Patchy - Ein programm zur analyseraumlicher verteilungsmuster von rebkrankhetien.Nachrichtenblatt des Deutschen Pflanzenschutzdienstes45:157-164.

McCullagh P, Nelder J A. 1989. Generalized Linear Models.New York: Chapman and Hall.

Meiering A G, Paroschy J H, Peterson R L, Hostetter G,Neff A. 1980. Mechanical freezing injury in grapevine trunks.American Journal of Enology and Viticulture 31:81-89.

Osler R, Carraro L, Loi N, Refatti E. 1993. Symptomexpression and disease occurrence of a yellows disease ofgrapevine in northeastern Italy. Plant Disease 77:496-498.

Padovan A C, Gibb K S, Bertaccini A, Vibio M, BonfiglioliR E, Magarey P A, Sears B B. 1995. Molecular detection ofthe Australian grapevine yellows phytoplasma andcomparison with Grapevine yellows phytoplasmas from Italy.Australian Journal of Grape and Wine Research 1:25-31.

Wilson Y, Hayes R. 2000. Understanding and management ofrestricted spring growth (RSG). In National Program for theManagement of Phytoplasmas in Australian Grapevines,GWRDC Final Report, Chapter 1.4. Project Number: CRCV95/2.

Wilson Y M. 1997. Winter chilling injury. The AustralianGrapegrower and Winemaker 402a:157-158.