Dealing with Frost: Describing vine response to frost damage and the impact of post-frost management on vine performance McArtney S., Chatterton, D. and Good, M. October 2003 Report to New Zealand Winegrowers HortResearch Client Report No. 12037 Reference S/150276/01

HortResearch Corporate Office 120 Mt Albert Rd, Private Bag 92169 Mt Albert, AUCKLAND, NZ Tel: +64-9-815 4200 Fax: +64-9-815 4201

McArtney S., Chatterton, D. and Good, M. HortResearch Hawke’s Bay Cnr Crosses and St Georges Roads Private Bag 1401, Havelock North, NZ Tel: +64-6-877 8196 Fax: +64-6-877 4761

CONTENTS Page EXECUTIVE SUMMARY........................................................................................................ 1 INTRODUCTION...................................................................................................................... 2 METHODS................................................................................................................................. 4 Vine response to level of frost damage .................................................................................. 4 Post-frost management options .............................................................................................. 4 RESULTS................................................................................................................................... 5 Vine response to level of frost damage .................................................................................. 5 Vegetative development and vine productivity ............................................................... 5 Shoot development .......................................................................................................... 6 Cluster development ........................................................................................................ 8 Fruit quality...................................................................................................................... 9 Post-frost management options ............................................................................................ 10 DISCUSSION .......................................................................................................................... 13 REFERENCES......................................................................................................................... 14 ACKNOWLEDGEMENTS ..................................................................................................... 14

EXECUTIVE SUMMARY Dealing with Frost: Describing vine response to frost damage and the impact of post-frost management on vine performance Report to New Zealand Winegrowers McArtney S., Chatterton, D. and Good, M.

October 2003

The effects of different levels of frost damage (none, half of the shoots frosted, all shoots frosted) on vine growth and productivity, fruit development and maturity were investigated on one commercial vineyard of Merlot and Chardonnay following a series of damaging frosts in spring 2002. In addition, the effects of various post-frost management strategies on vine growth and productivity were evaluated on three commercial vineyards in Hawke’s Bay following a series of damaging frosts in spring 2002. • Veráison occurred 10-12 days later in Merlot (Clone 3) bunches on shoots that developed from secondary buds compared with bunches on shoots from primary buds, and consequently soluble solids concentration was generally 1-1.5% lower when all bunches were harvested on the same day. •

Cluster number and yield (kg) of totally frosted Merlot (Clone 3) vines was reduced by 40% and 20%, respectively, compared with vines that were not injured by the frost.

•

Merlot (Clone 3) clusters on shoots that developed from secondary buds had more berries per cm rachis length, and subsequently had more disease at harvest. This ‘bunch density’ effect was probably due to the positive effect of high temperature and light conditions over the period when these clusters were flowering.

•

When frost-damaged shoots were broken out manually there was a trend for fewer shoots per vine to develop from secondary buds compared to the situation where frostdamaged shoots were pruned out (23-33% fewer shoots depending on the site). This difference was probably due to secondary buds at the base of frosted shoots being damaged when shoots were manually broken out. Interestingly, fruit maturity was delayed by approximately 6 days where frosted shoots were broken out manually.

•

Differences in the time at which frost-damaged shoots were pruned out (up to 2 weeks) did not result in any differences in juice soluble solids concentration at harvest.

Differences in fruit maturity between bunches on shoots from primary buds and those from secondary buds can create problems at harvest, particularly if fruit are maturing during wet autumn conditions. Where possible, frosted vines should be managed to produce a crop with uniform fruit maturity. In some cases, this may mean removal of the more fruitful undamaged shoots arising from primary buds. Further studies should be undertaken to describe variation in potential fruitfulness of secondary buds within a variety or clone in order to identify viticultural practises that could be used, after a frost event, to maximise that potential. For further information contact:

Dr. Steven J. McArtney HortResearch Private Bag 1401, Havelock North, Ph (06) 877 8196 Email [email protected]

2

INTRODUCTION Low grapevine yields in the 2000, 2001 and 2003 vintages occurred at a time of unprecedented international demand for New Zealand wines. Chardonnay and Sauvignon Blanc accounted for 60% of total national production in the 2002 vintage, yet the production of these two varieties was down by 38% in the 2003 vintage compared with the previous vintage. Supply shortfalls such as these create major problems that limit development of export markets. It is generally agreed that spring frost injury was the major contributing factor to the low yields in recent vintages. Growth was initiated earlier than normal in spring 2002 in Hawke’s Bay vineyards as a consequence of mild temperatures during August and September. Unexpected frosts on September 15, September 26 and October 5 (Figure 1) were therefore quite devastating in their impact, particularly on those varieties breaking bud earlier (Chardonnay, Pinot Noir and Merlot). Due to these frost events, production of Chardonnay in 2003 in Hawke’s Bay was down by 81% compared with the previous vintage (approximately 6000 tonnes). Assuming an average price of $1000/tonne this represents $6 million in lost revenue to Chardonnay growers in this region alone.

Figure 1. Hourly screen temperatures during the three radiation frosts in spring 2002 in Hawke’s Bay. Data are from the Longlands Road weather station. 4

Sep-15

Temperature (°C)

2

Sep-26 Oct-05

0 22 23 24

-2

-4

1

2

3

4

5

6

7

8

3 Grapevines are almost entirely resistant to freezing during frosts of less than -3°C (bud tissue temperature) by virtue of their ability to supercool (Fuller and Telli, 1999). Dissolved solutes may lower the freezing point of liquid in bud tissues by 1-2°C through freezing point depression. This means that water may be retained in a liquid state at temperatures less than 0°C (= supercooling). When water freezes within plant cells it ruptures cell membranes, killing the cells and tissues. Ice nucleation normally occurs when the bud temperature falls to -3°C to -3.5°C. If frosts occur at an early stage of development (up to and including DS02: buds swollen but not yet at cotton bud stage, see Figure 2) then injury is slight and non-lethal. Whereas, if frosts occur when buds are at DS03 or later (DS03: emergence of brown down among the scales, i.e. cotton buds) then they can be almost completely killed. The freezing of buds can be correlated with their water content. During the early stages of development (DS0, DS01, DS02, DS03) the bud water content rapidly increases from approximately 40% to 80% (Figure 2). During this transition, buds gradually lose the ability to supercool and the risk of frost damage increases.

Figure 2. Changes in water content of grapevine buds during initial development stages (DS), (adapted from Fuller and Telli, 1999).

Developmental stage Bud water content (%)

40

55

78

80

85

85

Spontaneous ice nucleation always occurs internally within the canes, before it occurs within the buds, during early stages of bud growth. The rate of ice spread in canes is comparable to the rate of spread in pure supercooled water (0.47 cm s-1), suggesting that the ice is travelling in the bulk water contained within xylem vessels in the canes (Hamed et al., 2000). The lack of a fully functional xylem system between canes and buds, that exists during bud burst, is proposed to act as a barrier to the spread of ice from the canes into bud tissues during this time. This report discusses a series of trials that were undertaken on commercial vineyards in Hawke’s Bay following the radiation frosts that occurred during spring in 2002. The objectives of these trials were to assess productivity, fruit development and fruit quality characteristics of Chardonnay and Merlot vines that were either partially or completely damaged during spring frosts, compared with vines that suffered no frost damage on the same vineyard, and to evaluate the effects of different post-frost management strategies on yield, fruit development and fruit quality.

4

METHODS VINE RESPONSE TO LEVEL OF FROST DAMAGE One Chardonnay and one Merlot vineyard were identified in Hawke’s Bay where frost damage was inconsistent across the block: within each block there were areas of undamaged vines (control vines, developing a canopy of shoots from primary buds only), areas of vines that had suffered frost damage to approximately 50% of the developing primary buds (developing a mixed canopy with shoots that developed from both primary and secondary buds), and areas of vines where all of the primary buds had been damaged (developing a canopy with shoots derived from secondary buds only). Ten vines of each canopy type (shoots derived from primary buds only, secondary buds only, or mixed primary and secondary buds) were tagged at each location for monitoring shoot and fruit development. Six sample shoots per vine were tagged on primary and secondary canopies, whereas 12 sample shoots were tagged on the mixed canopies (6 primary shoots and six secondary shoots). The number of clusters per shoot was recorded separately for each shoot type. Shoot development and fruiting characteristics of primary shoots on control vines were compared with development of secondary shoots on damaged vines. One shoot per vine was destructively sampled on 14 January, immediately prior to the first pass with a shoot trimmer. Area per leaf (cm2) of the primary leaves, and the total lateral leaf area at each node were measured with a LICOR LI 3100 leaf area meter. The number of clusters per shoot, number of berries per cluster, and cluster length (distance from the tip of the cluster to the point on the rachis where the shoulder branches) were recorded. Cluster compactness was calculated as the number of berries per cm of cluster length. Progression of veráison within each cluster on tagged shoots (Merlot only) was quantified by rating the proportion of berries within each cluster that had begun to accumulate red colour pigments (0 = 100% green berries; 10 = 100% red berries). This rating system was used to estimate differences in berry development between clusters on the various shoot types (primary shoots, secondary shoots, primary shoots in a mixed canopy, secondary shoots in a mixed canopy), using a veráison rating of 5 as the reference point.

P OST-FROST MANAGEMENT OPTIONS There is little quantitative information available to help viticulturists determine the best management options once frost damage has occurred. When a frost has killed all shoots on the vine then the viticulturists is faced with the choice of either removing them or leaving the dead shoots in the canopy. If they are to be removed then they will need to decide on the best practical method of removal ie. choosing between manually breaking the dead shoots out or pruning them out. Several different post-frost management strategies were adopted on commercial vineyards in Hawke’s Bay in 2002 as described in Table 1. The effects of these different management strategies on bunch development and fruit quality were monitored throughout the season.

5

Table 1.

Description of various post-frost management options studied on three vineyards in Hawke’s Bay in 2002/03.

Site

Post-frost management options investigated

Vineyard A Chardonnay (Clone 15)

Control (frosted shoots left) Frosted shoots cut out 8 October (3 days after last frost, DALF) Frosted shoots broken out manually on 14 October (9 DALF) Frosted shoots cut out 21 October (16 DALF)

Vineyard B Merlot (Clone 3)

Frosted shoots cut back to basal bud on 23 October (18 DALF) Frosted shoots broken out manually on 23 October (18 DALF)

Vineyard C Chardonnay (Mendoza)

Control (frosted shoots left) Frosted shoots cut back to basal 2-3 buds on 18 October (13 DALF) Frosted shoots cut back to basal bud on 18 October (13 DALF)

RESULTS VINE RESPONSE TO LEVEL OF FROST DAMAGE Vegetative development and vine productivity Shoot density of Chardonnay (Clone 15) was similar on control vines and totally frosted vines, indicating that shoots arising from secondary buds replaced those killed when primary buds were frosted. However, shoots arising from secondary buds were less fruitful than those arising from primary buds on control vines, producing on average only 0.7 clusters per shoot compared to 1.0 clusters per shoot (Table 2). Similar trends were observed for shoot density and cluster number on Merlot (Clone 3) vines: shoot density was not affected by frost damage, but the fruitfulness of shoots from secondary buds of Merlot (Clone 3) (0.8 clusters per shoot) was much less that that of shoots from primary buds (1.4 clusters on primary shoots on control vines).

6

Table 2. Productivity of Chardonnay (Clone 15) and Merlot (Clone 3) vines with differing degrees of frost damage. Data in brackets are expressed as a percent of the undamaged control vines. Control vines

Partially frosted vines

Totally frosted vines

Chardonnay (Clone 15) Shoot density (shoots/m) Clusters/vine Clusters/shoot Vine yield (kg)

19.5 35.6 1.0 1.2

25.9 (132) 26.5 (74) 0.6 (60) 1.7 (142)

21.8 (111) 24.8 (69) 0.7 (70) 2.1 (175)

Merlot (Clone 3) Shoot density (shoots/m) Clusters/vine Clusters/shoot Vine yield (kg)

16.3 38.4 1.4 8.1

16.7 (102) 36.4 (95) 1.2 (86) 7.7 (95)

16.9 (103) 23.0 (60) 0.8 (57) 6.5 (80)

Shoot development Shoots that developed from primary buds on control (undamaged) vines were different to shoots that developed from secondary buds on frosted vines. For Chardonnay (Clone 15) vines, there was little difference in the number of main shoot leaves or their individual areas between control and frosted vines, yet the total leaf area of lateral shoots arising from each node position was considerably less on frosted vines (Figure 3). The resumption of shoot growth from secondary buds occurred after growth resumed from primary buds on unfrosted vines. These data indicate that the rate of growth (node formation, expansion of the main axis leaves) of shoots arising from secondary buds of Chardonnay (Clone 15) occurred more rapidly than growth from primary buds on unfrosted vines. However, development of lateral shoots in the axils of the main shoot leaves was suppressed on shoots arising from secondary buds. The net result of this suppression of lateral shoot growth in secondary buds would be a reduction in canopy density, for Chardonnay (Clone 15) at least. Shoots arising from secondary buds on frosted Merlot (Clone 3) vines were less well developed compared with shoots arising from primary buds on undamaged control vines (Figure 3). The number of nodes on shoots from secondary buds was considerably less compared with shoots from primary buds, indicating that development of these shoots was delayed.

7 Figure 3. Comparison of leaf area development on shoots arising from primary and secondary buds of (A) Chardonnay (Clone 15) and (B) Merlot (Clone 3) grapevines. Shoots arising from primary buds are from control (unfrosted) vines; shoots arising from secondary buds are from frosted vines. Areas in black represent main shoot leaf area at each node; areas in white represent total lateral leaf area at each node.

28

Primary leaf area Lateral shoot leaf area

Node number from base

A

shoots arising from secondary buds

25 22 19 16 13 10 7

shoots arising from primary buds

4 1

-400 -300 -200 -100

0

0

100 200 300 400 2

Leaf area (cm ) 28

B

Lateral shoot leaf area 25

Node number from base

Primary leaf area

shoots arising from secondary buds

22 19

shoots arising from primary buds

16 13 10 7 4 1

400 300 200 100

0

0 100 200 300 400 500 2

Leaf area (cm )

8 Cluster development Development of Merlot (Clone 3) clusters on shoots arising from secondary buds was delayed by about ten days compared with clusters on shoots arising from primary buds (Figure 4). Furthermore, the development of clusters on vines with a mixed canopy of shoots from primary and secondary buds was delayed by about 2-3 days. There was no apparent difference in within-cluster variation in fruit maturity on the different shoot types.

Figure 4. Veráison development within Merlot (Clone 3) bunches on shoots derived from primary and secondary buds, either alone or mixed within the canopy. all berries red

10

Control vines - shoots from 1° buds Partially frosted vines - shoots from 1° buds

Veraison score

8

Frosted vines - shoots from 2° buds

Merlot

Partially frosted vines - shoots from 2° buds

6 4 2

all berries green

0

7-Feb 14-Feb 21-Feb 28-Feb 7-Mar Date

Merlot (Clone 3) clusters on shoots from secondary buds were more compact and tended to have a greater severity of bunch rots at harvest compared with clusters on shoots derived from primary buds (Figure 5). The greater compaction of clusters was evident on Merlot (Clone 3) only, and therefore was probably a reflection of higher fruit set in response to more favourable environmental conditions during the period when clusters on shoots derived from secondary buds were flowering/setting berries.

9 Figure 5. Merlot (Clone 3) clusters on shoots arising from primary (left) and secondary (right) buds. Note greater compaction of clusters from frosted vines. Relationship between cluster compactness (berries/cm rachis length) and disease score at harvest (0 = no bunch rot; 5 = severe bunch rot). Solid symbols represent basal clusters. Open symbols represent secondary clusters. Vertical and horizontal bars represent ± SE. 3.5

Frosted vines (2° shoots) Control vines (1° shoots) Disease score

3

2.5

2

1.5

1 4

6

8

10

12

Cluster compactness (berries/cm)

Fruit quality In the case of Chardonnay (Clone 15) the juice from bunches on unfrosted control vines was 1.6 Brix higher than juice from bunches that developed from secondary buds on totally frosted vines (Table 3). Later maturing fruit were less susceptible to disease encouraged by inclement weather during the late March period leading up until harvest. Consequently, the severity of bunch rots was much higher on control vines compared with bunches on frosted vines. Bunch rot in control vines was responsible for low cluster weights and vine yields compared with frosted vines. In this case, vine yield and fruit quality was primarily determined by weather patterns and the extent of fruit susceptibility to disease during the period from veráison to harvest rather than by any direct effect of frost damage. The overall level of disease severity was lower in the case of Merlot (Clone 3), a naturally later maturing variety, so that cluster weights and vine yields more closely reflected the cluster number per vine, and more directly reflected the impacts of frost.

10 Table 3. Fruit quality of Chardonnay (Clone 15) and Merlot (Clone 3) vines with differing degrees of frost damage. Control vines

Partially frosted vines 1°° buds 2°° buds

Totally frosted vines

Chardonnay (Clone 15) Cluster weight (g) Disease score Brix pH TA

34.8 4.9 22.1 3.3 10.0

51.6 3.0 21.4 3.2 9.9

74.7 1.1 20.1 3.1 10.7

85.6 2.0 20.5 3.1 10.1

Merlot (Clone 3) Cluster weight (g) Disease score Brix pH TA

235 1.9 20.2 3.3 8.0

294 2.6 19.7 3.4 8.8

271 1.7 18.5 3.3 9.3

227 1.2 20.3 3.3 8.5

P OST-FROST MANAGEMENT OPTIONS Vineyard A There were no major differences in the number of shoots or clusters per vine following the various post-frost management practises carried out on the Chardonnay (Clone 15) vines in this study (Table 4). The development of clusters on shoots derived from secondary buds of Chardonnay (Clone 15) varied according to how the vines were managed after the frost. Cluster development was delayed by approximately one week on control vines where no frosted shoots were removed compared with vines where frosted shoots were pruned out on 8 October (data not shown). This delay in cluster development was reflected in lower juice Brix levels at harvest on control vines (Figure 6). Interestingly, there was also greater variation in juice Brix levels between control vines compared with vines that had frosted shoots either manually broken or pruned out. Table 4.

Effects of various post-frost management practices on shoot and cluster number on Chardonnay (Clone 15) grapevines.

Treatment

Control Broken out 14 Oct Pruned out 8 Oct Pruned out 21 Oct

Shoots per vine 1° buds 2° buds Total 2.0 3.3 3.3 2.3

12.5 9.0 12.0 10.0

14.5 12.3 15.3 12.3

Clusters Per vine Per shoot 7.3 9.5 8.5 7.3

0.5 0.8 0.6 0.6

11 Figure 6.

Effects of various post-frost management practises on Brix level of Chardonnay (Clone 15) juice at harvest, and variability in Brix level between vines. Data are presented as box plots. Data in black are for clusters on shoots that developed from primary buds whereas data in red are for clusters on shoots that developed from secondary buds.

22

Shoots from primary buds Shoots from secondary buds

21

Brix level

20

19

Broken out 14 Oct Pruned 8 Oct

18

17

Pruned 21 Oct

Control-do nothing

16

Vineyard B The number of clusters per vine was 25% higher where frost-damaged shoots were removed by pruning compared with removal by breaking the shoots out. This increase was due to an increase in the number of fruitful shoots rather than to an increase in the fruitfulness of individual shoots (Table 5). Cluster development on secondary shoots was advanced by approximately 6 days on vines that had frost-damaged shoots removed by pruning compared with manual removal of damaged shoots (Figure 7).

Table 5. Effects of various post-frost management practices on shoot and cluster number on Merlot (Clone 3) grapevines. Treatment

Broken out 23 Oct. Pruned 23 Oct.

Shoots per vine 1° buds 2° buds Total 2.8 1.5

7.5 11.3

10.3 12.8

Clusters Per vine Per shoot 9.5 12.0

0.9 0.9

12 Figure 7. Differences in development of veráison for Merlot (Clone 3) grape clusters on shoots derived from primary buds (data in black) and secondary buds (data in red) after spring frost injury. A comparison is provided between manual removal of frosted shoots (dashed lines) and pruning back frosted shoots to the basal bud (solid lines). All berries 10 red

Veraison score

8

6

4 Frosted shoots pruned, clusters on shoots f rom 1° buds Manual removal, clusters on shoots from 1° buds

2

Frosted shoots pruned, clusters on shoots f rom 2° buds

All berries green

Manual removal, clusters on shoots from 2° buds

0 3-Feb 10-Feb 17-Feb 24-Feb 3-Mar 10-Mar 17-Mar Date

Vineyard C The various post-frost management options investigated in Vineyard C did not have any major effects on shoot or cluster development (data not shown). There was a trend for more secondary shoots to develop when frost-damaged shoots were pruned back to the basal bud compared with pruning these to leave 2-3 basal nodes on the shoot (Table 6).

Table 6. Effects of pruning frost-damaged shoots back to the basal bud or leaving 2-3 basal buds on shoot and cluster number per vine of Chardonnay (Mendoza) grapevines. Treatment

Control Pruned to basal bud Pruned to 2-3 bud spur

Shoots per vine 1° buds 2° buds Total 2.5 2.3 2.8

11.5 14.8 11.8

14.0 17.0 14.5

Clusters Per vine Per shoot 14.5 11.5 11.8

1.0 0.7 0.8

13

DISCUSSION The studies presented in this report have highlighted differences in canopy development and fruit maturity arising from clusters developing on shoots from primary and secondary buds. Severely frost-damaged Merlot (Clone 3) vines developed shoots from secondary buds only, yet cluster number and yield per were reduced by only 40% and 20% respectively, compared to undamaged vines. However, maturity of bunches on shoots from secondary buds was delayed by approximately ten days (resulting in 1-1.5% lower soluble sugars concentration at harvest) compared with bunches on shoots from primary buds, leading to problems of mixed fruit maturity in partially frosted vines. If weather conditions leading up to harvest are unfavourable, as they were in the 2003 vintage, then the problem of mixed fruit maturity will result in a higher incidence of disease in the more mature bunches (on shoots derived from primary buds) and determining the optimum time to harvest becomes a compromise between minimising disease on the most advanced bunches and reaching an acceptable level of ripeness on the more delayed bunches. Vines should be managed, frost or no frost, to minimise variation in fruit maturity within the vine. If frost damage is severe then most shoots within the canopy will develop from secondary buds, and undamaged primary shoots should be removed as soon as practicable after the frost event to stimulate shoot development from the secondary bud at that site. In this instance the higher fruitfulness (i.e. bunch number and weight) of the few remaining primary buds on the vine will be sacrificed for lower yields of more uniform fruit maturity. If the level of frost damage is only minor, say less than 20% of the primary buds damaged, then perhaps the bunches should be removed from shoots derived from secondary buds in order to reduce within-vine variability in fruit maturity. Flower clusters on shoots derived from secondary buds are likely to develop under more ideal environmental conditions (temperature, light) than clusters on shoots from primary buds. Ambient temperatures during the period from bud break to bloom will modify the number of flowers per inflorescence, higher temperatures resulting in more flowers per inflorescence. In addition, high natural light intensities during bloom will favour higher berry set. Data collected from the Merlot (Clone 3) vines in the first part of this study suggested that berry set was increased in bunches sampled from shoots derived from secondary buds, resulting in these bunches having a higher “compactness” rating ie. berry number per centimetre of rachis length was higher. The cluster architecture within “compact” bunches is more favourable for disease development ie. bunch closure occurring earlier, poor penetration of fungicide sprays into the bunch, and longer periods of surface wetness within the bunch due to slower drying times. These responses may explain why Merlot (Clone 3) bunches on shoots derived from secondary buds had a higher disease score at harvest. Frost-damaged shoots need to be removed from the vine. If they are retained then infertile lateral buds may develop from undamaged basal nodes on the shoot, suppressing growth of secondary buds and resulting in an excessively dense canopy. There was no evidence that retention of dead shoot tissue within the canopy resulted in increased disease pressure later in the season in the present studies. Nevertheless, for the reasons stated above, frost damaged tissues should be removed as soon as the risk of frost has passed. Differences in the time of pruning out of frost-damaged shoots of up to 2 weeks did not result in any differences in juice soluble solids concentration at harvest. Premature removal of frost-damaged shoots may result in shoots arising from secondary buds being damaged in subsequent frost events. Frost-

14 damaged shoots can be pruned out or broken out by hand. The cost of removal of frost damaged shoots will be much lower if manually broken out rather than pruned off, but there was a trend for fewer shoots per vine where frost-damaged shoots were broken out. This trend probably reflects a higher incidence of damage to the secondary bud at the base of the shoot when they are broken out, which may also reduce vine yield. Frost damage can severely reduce vine yield and juice quality. Vine yields are reduced because secondary buds of vinifera wine grapes are less fruitful than primary buds. Differences in the fruitfulness of secondary buds exist between vinifera cultivars, but little is known about the effects of viticultural practises on the potential fruitfulness of secondary buds. Practices which encourage post-harvest photosynthetic activity may enhance the potential fruitfulness of secondary buds. Further studies should be undertaken to determine, in the first instance, the extent of variation in potential fruitfulness of secondary buds within a variety or clone. This could be achieved by adopting a survey approach and forcing secondary buds to develop by pruning shoots from primary buds soon after bud break. Where variation in potential fruitfulness exists between regions or vineyards within a region then the sources of this variation should be established. Variation between regions may be related to different climatic conditions between those regions whereas differences within a region may indicate an effect of viticultural practices on the fruitfulness of secondary buds.

REFERENCES Fuller, M.P. and Telli, G. 1999. An investigation of the frost hardiness of grapevine (Vitis vinifera) during bud break. Annals of Applied Biology 135:589-595. Hamed, F., Fuller, M.P. and Telli, G. 2000. The pattern of freezing of grapevine shoots during early bud growth. CryoLetters 21:255-260.

ACKNOWLEDGEMENTS The authors are grateful to Montana Wines Limited, Alistair Noakes and Mel MacLennan for access to their vineyards and continued cooperation throughout the trial.