Overwintering grapevine debris as an important source of Botrytis cinerea inoculum

Authors


E-mail: marlene.jaspers@lincoln.ac.nz

Abstract

Production of Botrytis cinerea conidia from infected grapevine debris (trash) left on the ground and in the canopy in the season following harvest was studied in vineyards in Marlborough, New Zealand. When subsamples were incubated under high relative humidity in the laboratory, rachides had the greatest sporulation potential (P < 0·05), followed by tendrils, cane lengths and petioles. Trash remaining on the ground under the canopy had higher rates of sporulation (P < 0·05) than that in the inter-row. The sporulation potential of rachides at different times during the growing season was assessed by placing them in vine canopies or on the inter-row soil in three vineyards in late spring. Subsamples were removed on five occasions between flowering (capfall) and harvest, and incubated under high relative humidity in the laboratory. Mean numbers of conidia produced from the canopy rachides diminished from 3·5 × 105 per rachis at capfall to 2·6 × 104 at harvest, and from 3·9 × 105 to 2·7 × 103, respectively, from the ground rachides. The greater loss in sporulation capacity of ground rachides was considered to be associated with their earlier spontaneous sporulation and greater degradation in the moist inter-row sward, where they lost 29% of their weight (P < 0·001) and 23% of their pedicels (P < 0·001), compared to the canopy rachides which lost 0% of their weight and 3% of their pedicels from capfall to harvest. This study has shown that necrotic, overwintering grapevine debris can produce B. cinerea conidia throughout the following growing season, so may contribute to the subsequent risk of bunch rot.

Introduction

Botrytis cinerea is a major cause of grape bunch rot, causing lower yields and reductions in fruit and wine quality. Infection of bunches can occur at anthesis, when flower parts are susceptible to conidial infections of styles, ovules, stamens or petals. The pathogen frequently becomes quiescent after infection of flower parts, only resuming growth as the berry ripens, when it rots the fruit. The significant correlations reported between rates of floral infection in spring and subsequent bunch rot in summer show the significance of this infection pathway (Elmer & Michailides, 2004). The pathogen can also develop on abscising flower parts that may remain trapped in the developing bunch, from which it can infect adjacent berries (Elmer & Michailides, 2004). Holz et al. (2003) also reported that flower tissues colonized with B. cinerea could produce continuous crops of conidia during moist conditions, resulting in contamination of the ripening berries which become more susceptible as they ripen. Infection and sporulation of B. cinerea are increased by warm (15–25°C), wet conditions associated with rainfall (Braun & Sutton, 1988).

The saprophytic ability and nutritional versatility of B. cinerea has given the pathogen a survival advantage. Holz et al. (2004) concluded that the pathogen was able to survive in dead, infected host structures from the previous season as dormant or metabolically inactive fungal structures such as sclerotia and chlamydospores, or as saprophytically active mycelium that was also able to produce inoculum. In vineyards, colonized necrotic grapevine tissues are acknowledged as an important inoculum source for the following growing season (Nair et al., 1995; Elmer & Michailides, 2004). Sclerotia were identified as the most important inoculum source of vineyards in the Hunter Valley, Australia (Nair & Nadtotchei, 1987) but they were less common in Marlborough, New Zealand where the climate is relatively dry (Elmer & Michailides, 2004). In New Zealand, winter pruning, summer trimming and senescing leaves may all produce a large quantity of potential saprophytic substrate. In addition, many vineyards are harvested by machine, which leaves rachides attached to the canes. Rachides, mummified berries and pruned canes have all been identified as potential sources of B. cinerea inoculum for the following season (Thomas et al., 1983; Nair & Parker, 1985; Nair et al., 1995) and therefore staff are usually instructed to remove them during winter pruning.

The relative importance of the different inoculum sources depends on the volume of the source and the length of time spores can be produced from the colonized tissue. The pathogen must survive long enough to either produce primary inoculum at a time when the host is susceptible, or when other substrates are available for colonization, from which it will subsequently infect the crop. The host tissues and environmental conditions must also be conducive for pathogen germination, infection, colonization and sporulation.

In cool climate viticulture, management of this disease involves an integration of cultural methods and fungicide applications, mainly at flowering, pre-bunch closure and during fruit ripening. However, to prevent residues in wines, the use of fungicides near to harvest is limited to a few products that have short pre-harvest intervals. The availability of such products is also limited by the development of fungicide resistance (Beever & Brien, 1983). Cultural strategies believed to be effective in reducing the density of the canopy include techniques such as shoot thinning, leaf plucking and leaf trimming (Stapleton et al., 1990) or by using vigour reducing rootstocks, inter-row planting and/or regulated deficit irrigation (Sicher et al., 1995; Valdes-Gomez et al., 2008). These methods, which increase aeration and exposure of bunches to light, have been shown to reduce development of bunch rot (English et al., 1989). However, inoculum sources within the crop may also influence the level of infection. Abundant conidia of B. cinerea were found to be produced from leaves and flowers in French vineyards (Martinez et al., 2005). A quantitative relationship between numbers of airborne conidia of B. cinerea and infection of strawberry flowers was demonstrated by Xu et al. (2000), while Spotts & Cervantes (2008) showed a similar relationship for botrytis rot of pear fruit. Leaves were considered to be the main source of primary inoculum of B. cinerea on strawberries in Ontario, Canada, whereas in Scotland mummified fruits, weeds and barley straw used for mulching were also important primary sources (Braun & Sutton, 1988). However, comprehensive studies have not been conducted on the types of debris present in vineyards and their comparative roles as sources of inoculum.

This study investigated the inoculum potential of a range of overwintering debris types, comprising senescing tissues naturally deposited under the vines and pruned tissues (trash), in vineyards in Marlborough, New Zealand. The effect of ground position for the trash, and of canopy and ground position for rachides was also investigated to determine their effects on sporulation from the saprophytic B. cinerea through the following growing season.

Materials and methods

Survey of vineyards for ground debris

The amount of ground trash and its sporulation potential was investigated in 10 commercial Sauvignon blanc vineyards in Marlborough, New Zealand. In each vineyard, 15 quadrats were randomly selected in early summer (December 1998). At each sampling point, quadrat frames of 0·25 m2 were placed on the ground, directly under the canopy and in the adjacent inter-row area (30 sampling points per vineyard). In Vineyard 2, the entire ground surface was covered in a mown grassy sward. In all other vineyards, the under-canopy area was relatively free of vegetation because it was managed by 3–4 ‘knock-down’ herbicide applications per annum and the inter-rows were covered in mown vegetation, largely grasses.

The necrotic grapevine tissues in each quadrat were removed and sorted into six trash types. The rachides, petioles, tendrils and cane lengths (≥5 cm in length) were counted, whereas the small cane fragments (<5 cm in length, <1·5 cm in diameter) and large cane fragments (<5 cm in length, ≥1·5 cm in diameter) which varied in size, were weighed. From each group a subsample was taken, comprising 20 pieces or 10 g (for the two types of cane fragments). The approximate mean surface areas of these pieces were determined by measuring them and using the appropriate geometric equations (i.e. for cylinders and spheres). This allowed for comparison of the sporulation potentials of the different trash types by surface area. The tissue specific sporulation ability (TSSA) was then determined per piece from each trash group by incubating the tissue pieces in each subsample on moist paper towels in individual humid chambers set on the benches in an air-conditioned laboratory (20 ± 2·5°C) under natural light. After incubation for 5 days, each subsample of sporulating debris was carefully placed into a jar with 40 mL of sterile water and Tween 20 (2 drops per L; polyoxyethylenesorbitan monolaurate, BDH); jars were agitated on a rotary shaker at 170 rpm for 3 min. The solutions were centrifuged at 805 g for 5 min, the supernatants decanted off and the pellets resuspended in 10 mL of sterile water. Counts of conidia that appeared typical of B. cinerea (hyaline, thin-walled and ovoid, 10–12 × 8–10 μm; Bulit & Dubos, 1988) were made from three separate subsamples of each suspension using a haemocytometer. The mean number of conidia per piece was calculated to give the sporulation potential per square millimetre of each trash type and all trash per quadrat. On petioles, sclerotia were also observed and so they were counted on these tissues.

To assess the effect of trash site, sampling position and vineyard on sporulation potential, the sporulation data were log-transformed and then analysed by analysis of variance (anova) with the package systat v. 9.0. Fisher’s protected least significant difference (LSD) test was used to examine differences between trash types in each vineyard, while differences between the inter-row and under-canopy positions, at each sampling point, were determined using paired t-tests.

Survey of vineyards for rachis trash in canopies

In eight commercial Marlborough Sauvignon blanc vineyards, the vines within 10 randomly selected rows of equivalent length were inspected in spring (October 1999) for presence of rachides in the canopy. A sample of 30 rachides was randomly selected from each vineyard, the TSSA assessed by moist incubation and the data analysed as above.

Sporulation potential of rachides over time

The ability of rachides to produce B. cinerea conidia during the growing season was investigated in three commercial Marlborough Sauvignon blanc vineyards that had different mesoclimates. The vineyards were within 2–4 km of each other, having similar temperature ranges but different levels of natural rainfall, with Vineyards 2, 5 and 7 having approximately 1000, 900 and 700 mm per annum, respectively (Rae, 1987). The depth and density of ground and canopy vegetation was estimated near the end of the sampling period, at 14 days before harvest. The mean height and percentage vegetation cover on the ground beneath the vines was estimated within five 0·25 m2 quadrat frames that were placed at random in each vineyard. Current ground cover management practices were also recorded for the under-canopy and inter-row areas. At the same time, the canopy density was measured using the point quadrat method described by Smart & Robinson (1991). The density measurements, comprising leaf layer number and the number of gaps in the canopy, were taken at 25 points selected at random along the fruiting wire in each experimental plot.

The 1500 rachides needed for the experiment had been collected from the ground directly beneath Sauvignon blanc vines from a single vineyard in late spring (November 1999) and stored at 4°C while the sporulation potential of a subsample (25 rachides) was assessed as before. As soon as they were confirmed as being 100% infested with B. cinerea, the rachides were randomly allocated to the three experimental vineyards.

In each vineyard, there were four randomly selected sites, each comprising a 10-vine plot (five vines by two rows). In each plot, 25 rachides were randomly allocated and secured in each of two positions: (i) hung from the fruiting wire in the canopy, and (ii) on the vineyard floor directly beneath the canopy. Assessment of rachis characteristics was conducted at 80% capfall (9 December), pre-bunch closure (PBC; 13 January), post-véraison (15 February), pre-harvest (12 March) and harvest (13 April). On each occasion, five rachides were randomly selected from the ground and canopy positions in each experimental bay. They were air-dried for 5 days and their dimensions (length, number of pedicels and weight) were recorded. The rachides were then placed on moist paper towels in individual humid chambers for 5 days, followed by spore counting as described for ground debris in the previous experiment.

The rachis data, comprising pedicel number and weight as well as log-transformed TSSA data, were first analysed for each sampling time and position using anova, then analysed using general linear model (GLM) with repeated measures of the five sampling times, to assess the effect of trash site. To determine if positive linear relationships existed between rachis length, weight or pedicel number and sporulation potential of rachides, the data were analysed using GLM. Comparisons between consecutive sampling times were made using t-tests for sporulation data, pedicel number, length and weight of the rachides.

Sporulation from rachides in the field

A further 200 of the rachides previously collected and replaced in three vineyards were set out into two of the three vineyards above, within 2 weeks of collection. In each vineyard, 10 plots were randomly selected (five consecutive vines each), with 10 rachides per plot. The rachides were randomly allocated in each plot: (i) five hung from the fruiting wire in the canopy, and (ii) five pegged onto the vineyard floor directly beneath the canopy.

The rachides were examined in situ using a hand lens (×20 magnification), every 3 days until harvest, to determine presence of B. cinerea sporulation. Additionally, assessments were repeated whenever canopy wetness events occurred (rainfall that caused more than 50% of the canopy to remain wet for more than 6 h), at first light or within 4 h of the event, and daily for the next 2 days. The numbers of conidiophores were estimated using the numerical grades described by Fowler et al. (1999).

Results

Survey of vineyards for ground debris

The numbers and weights of trash pieces differed between vineyards (P < 0·05) for all trash types. In all vineyards, the under-canopy quadrats had more trash than the inter-row quadrats (P < 0·05), except for Vineyard 2 and Vineyard 9, where they were similar (Table 1). Overall, there were lower numbers of rachides, petioles and cane lengths (P < 0·001) per quadrat in the inter-row than under-canopy; however, small cane fragments were more abundant (P < 0·05) in the inter-row than the under-canopy, and abundance of large cane fragments was similar in both (P > 0·05; Fig. 1, Table 1).

Table 1. Mean total surface area (SA; mm2) of trash types per quadrat (0·25 m2 of vineyard floor), from under the canopy (UC) and in the inter-row (IR), of 10 Sauvignon blanc vineyards in Marlborough, New Zealand, in early summer (December). Total Botrytis cinerea conidia (× 103) per quadrat in each vineyard, and means for each trash type
VineyardSurface area of trash typesConidia (×102) produced per quadrat
RachidesTendrilsPetiolesCane lengthsaSmall cane fragmentsaLarge cane fragmentsaTotal SA (mm2)
UCIRUCIRUCIRUCIRUCIRUCIRUCIRUCIR
  1. aCane lengths (≥5 cm in length); small cane fragments (<5 cm in length, <1·5 cm in diameter); large cane fragments (<5 cm in length, ≥1·5 cm in diameter).

1113·128·0178·433·7156·77·8399·396·1182·093·80·04·51029·5263·9640·3167·0
227·128·081·3150·670·52·0151·6184·839·7144·733·571·9403·7582·1201·1265·1
365·69·396·584·6232·53·9140·554·260·466·317·40·0612·9218·3366·3107·0
439·36·081·385·2238·949·0314·2107·243·3110·60·025·6717·0383·6328·8121·7
544·31·3169·188·6218·10·0315·554·299·8116·120·86·4867·6266·6393·977·4
67·44·057·539·6374·235·359·273·970·651·40·022·8568·9227·0156·472·0
763·917·3140·1158·6821·548·3305·6177·5102·3145·89·730·81443·1578·3563·4232·7
860·78·0112·385·9141·024·8145·4125·774·970·67·711·5542·0326·5342·3130·1
93·30·011·944·990·10·024·619·725·936·20·05·3155·8106·139·233·6
1049·220·0103·1109·7662·13·9187·3155·394·5123·647·017·51143·2430·0422·2198·7
Means
 Trash SA (mm2) per quadrat43·512·2101·688·1284·417·5205·4104·977·395·915·219·6727·5338·2  
 Conidia (×103) per mm2   3·40  4·01   0·60  0·59   0·22 0·21  0·32  0·32  0·020·020·010·010·00·0  
 Conidia (×103) per quadrat154·948·960·651·862·53·665·634·01·62·00·10·20·00·0345·4140·5
Figure 1.

 Mean amounts of each trash type collected from the ground under the canopy and in the inter-row in 10 Sauvignon blanc vineyards in Marlborough, New Zealand, in early summer: (a) mean number of rachides, tendrils, petioles and cane lengths (≥5 cm in length) in 0·25 m2 quadrats, and (b) mean weight of small cane fragments (<5 cm in length, <1·5 cm in diameter) and large cane fragments (<5 cm in length, ≥1·5 cm in diameter) in 0·25 m2 quadrats. Error bars indicate standard errors of the means.

The sporulation potential of trash per quadrat differed significantly between vineyards (P < 0·05). It was also significantly greater for the under-canopy than the inter-row (P < 0·05), a pattern that was apparent in all vineyards except Vineyards 2 and 9 (Table 1). Rachides had the highest (P < 0·05) TSSA, with means of 3·40 × 103 and 4·01 × 103 conidia per mm2 for under-canopy and inter-row rachides, respectively. The remaining tissues had relatively low TSSAs that did not differ significantly (P > 0·05) between the under-canopy and inter-row (Table 1). The sclerotia on petioles were observed to provide most of the petiole sporulation. The TSSAs of the different tissues did not differ (P > 0·05) between vineyards (Fig. 1, Table 1).

The mean number of sclerotia on petioles differed significantly between vineyards, for petioles collected from under the canopy (2·5–9·5 per quadrat; P <0·001), but not within the inter-row (1·7–4·0 per quadrat; = 0·081). The mean number of petioles found with sclerotia was significantly greater (P < 0·001) from under the canopy than from the inter-row, with mean numbers per quadrat being 5·3 and 3·0, respectively.

Survey of vineyards for rachis trash in canopies

The number of rachides left in the canopy after pruning differed between vineyards (P < 0·001), ranging from 0·035 to 0·38 per vine. The mean number of rachides per vine was 0·20, with a TSSA per rachis of 7·5 ×103 conidia, which was similar between vineyards.

Sporulation potential of rachides over time

The initial subsample of rachides incubated under high RH had 100% incidence of sporulation and had a mean sporulation potential rate of 4·82 × 105 conidia per rachis. After placement in the vineyards, incidence of sporulation did not differ between vineyards (P > 0·05), but incidence decreased significantly (P < 0·05) between all sampling times except between post-véraison and pre-harvest. Incidence differed between sampling positions (P < 0·01), being greater for rachides on the ground than in the canopy at pre-bunch closure, but less on the ground than in the canopy for the three following sampling times (Fig. 2, Table 2).

Figure 2.

 Mean number of rachides showing Botrytis cinerea sporulation from rachides placed on the ground and in the canopy in three Sauvignon blanc vineyards in Marlborough, New Zealand, with five sampling times through the growing season. Error bars indicate standard errors of the means.

Table 2. Changes in the mean sporulation potential (Botrytis cinerea conidia × 103) per rachis in the canopy and on the ground between consecutive sampling times, for three Sauvignon blanc vineyards in Marlborough, New Zealand, during the growing season. P-values indicate significance of the changes
 Capfall to PBCPBC to post-véraisonPost-véraison to pre-harvestPre-harvest to harvest
  1. PBC: pre-bunch closure.

Ground
 Vineyard 2301·3–167·2,
P < 0·01
167·2–41·3,
P < 0·001
41·3–13·8,
P < 0·01
13·8–1·9,
P < 0·01
 Vineyard 5449·4–193·8,
P < 0·01
193·8–44·5,
P < 0·001
44·5–17·2,
P < 0·01
17·2–4·4,
P < 0·05
 Vineyard 7412·5–251·1,
P < 0·05
251·1–38·8,
P < 0·001
38·8–14·4,
P < 0·01
14·4–1·9,
P < 0·05
 Vineyards combined387·7–204·0,
P < 0·001
204·0–41·5,
P < 0·001
41·5–15·1,
P < 0·001
15·1–2·7,
P < 0·001
Canopy
 Vineyard 2246·9–129·2,
P < 0·01
129·2–51·4,
P < 0·05
51·4–51·3,
P = 0·991
51·3–22·0,
P = 0·068
 Vineyard 5369·4–104·8,
P < 0·001
104·8–65·0,
P < 0·05
65·0–45·5,
P = 0·082
45·5–33·1,
P = 0·191
 Vineyard 7421·9–212·0,
P < 0·001
212·0–58·9,
P < 0·01
58·9–40·9,
P = 0·102
40·9–23·8,
P = 0·114
 Vineyards combined346·0–148·7,
P < 0·001
148·7–58·4,
P < 0·001
58·4–45·9,
P = 0·065
45·9–26·3,
P < 0·01

The mean sporulation potentials from rachides also reduced over time. The three vineyards did not differ with respect to sampling time from ground rachides (= 0·112), but did for canopy rachides (P < 0·01). For rachides on the ground, sporulation potentials decreased between all consecutive sampling times (P < 0·05–P < 0·001 for the three vineyards; Table 2), and the conidial numbers per rachis were very low (1·9–4·4 × 103) by harvest. For rachides in the canopy, sporulation potentials declined a little more gradually, differences being significant over the first two consecutive sampling intervals (P < 0·01), but not significant from post-véraison to harvest (= 0·068–= 0·991; Table 2), with higher mean conidial numbers per rachis by harvest (22·0–33·1 × 103) than those on the ground. Table 3 shows that Vineyards 5 and 7 had an under-vine herbicide treatment, with reduced height and density of under-canopy vegetation compared to Vineyard 2, where no under-canopy herbicide treatment was applied. Vineyard 2 had lower canopy density than the other two vineyards.

Table 3. Characteristics of the three Sauvignon blanc vineyards in Marlborough, New Zealand, used to assess the sporulation potential of Botrytis cinerea on rachides in the field over the growing season. Measurements of ground cover and canopy density were taken 2 weeks prior to harvest
CharacteristicVineyard 2Vineyard 5Vineyard 7
  1. aMeasurements taken according to point quadrat protocol outlined by Smart & Robinson (1991).

Inter-row vegetation height (cm)2215<5
Under vine managementUnder-vine mowingHerbicide treatedHerbicide treated
Under vine vegetation (% cover)41·99·18·1
Trellis typeScott HenryVertical shoot postitionVertical shoot postition
Canopy density (% gaps)a805
Number of leaf layersa1·22·41·9

The dimensions of the rachides placed in the vineyards decreased over the growing season. For rachides on the ground, the mean numbers of pedicels decreased significantly over the season (from 56 to 43; P < 0·001), as did the mean mass (from 0·89 to 0·63 g; P < 0·001) and the mean length (from 9·5 to 9·0 cm; P < 0·05; data not shown). For rachides hung in the canopy, the mean number of pedicels decreased significantly over the season (from 60 to 54; P < 0·001) as did the mean weight (from 0·86 to 0·72 g; P < 0·001), but not the length. The reducing dimensions of the rachides over time were related to the reducing rates of sporulation at some sampling times only. Significant relationships (P ≤ 0·05) existed between mean numbers of conidia produced from the ground rachides and the mean pedicel numbers per rachis, rachis weight and length at the second, second and first sampling times, respectively. Similarly, there was a significant relationship (P ≤ 0·05) between conidial numbers from canopy rachides and the mean pedicel numbers per rachis, rachis weight and length at the third, second and first sampling times, respectively.

Sporulation from rachides placed into vineyards

Botrytis cinerea sporulation was observed on only four occasions in both Vineyards 5 and 7, although there were 60 and 39 canopy wetness events, respectively. The number of observed sporulation events and mean number of conidiophores per rachis were the same for ground rachides and canopy rachides and similar for both vineyards. The number of conidiophores produced after each wetness event decreased over the season, with 250–350 per rachis in late November (capfall), decreasing to 10–20 per rachis by late February (post-véraison). Sporulation was not observed on rachides during March and April, even after wetness events.

Discussion

Grapevine trash with the potential for B. cinerea inoculum production was found in all of the vineyards surveyed, although the quantities of the trash types varied between vineyards (P < 0·05) and the sporulation potentials differed between trash types (P < 0·001). Rachides had greater sporulation per unit surface area than canes, petioles and tendrils, which were present in greater quantities. Vineyards with high levels of rachides, particularly in the under-vine area, generally had the greater sporulation potentials.

Reports of overwintering inoculum in many grapevine growing regions have generally concluded that sclerotia provided the main mechanism for overwintering of B. cinerea (Bulit & Dubos, 1988; Holz et al., 2004). In a study by Nair & Nadtotchei (1987) in the Hunter Valley, Australia, sclerotia were identified as an important source of primary inoculum, as they were abundant on canes and were shown to lead to infection of flowers and berries. In the laboratory, these authors found that sclerotia could continue to resporulate, if the fresh conidia were repeatedly removed from their surfaces, for up to 12 weeks after the first sporulation incident. In France, Martinez et al. (2005) reported differences for cultural characteristics of isolates with respect to mycelial growth and sclerotium development that was related to the organs of origin. The sclerotia were commonly found on canes and those isolates recovered in winter always produced high numbers of sclerotia in culture, in contrast to isolates from necrotic tissues recovered in warmer seasons that produced fewer sclerotia in culture. In Hawke’s Bay, New Zealand, P. Wood (Plant and Food Research, New Zealand, unpublished data) also found that most petioles had sclerotia, which were a common source of overwintering inoculum. In contrast, the current study found few sclerotia on petioles and none on canes, which may have been affected by the relatively dry summer climate of Marlborough, an observation that was also made by R. Balasubramaniam (Plant and Food Research, New Zealand, personal communication). However, sclerotia are small (2–4 × 1–3 mm) dark, rough, discoid and attached to plant tissues (Bulit & Dubos, 1988), and would not show up clearly on the dark rough surfaces of rachides, so may in fact have been present and caused some of the profuse sporulation seen in this study.

The role of necrotic tissues in overwinter survival of B. cinerea has also been reported by Nair et al. (1995) who confirmed the results of the current study by demonstrating sporulation potential of overwintering rachides and their attached mummified berries. A regression analysis with subsequent infection events showed that the rachides and mummified berries accounted for up to 62% of the flower infections in the Hunter Valley, and these infections accounted for 21% of the berry infection later in the summer. Holz et al. (2004) also confirmed the overwinter survival of B. cinerea in necrotic tissues, which they suggested could be by mycelium, microsclerotia or chlamydospores. However, the types of structures involved or the effects of them on pathogen longevity has not been investigated. The longevity of isolates in a vineyard may also be affected by their genetic features and the diversity of the populations, as shown by Martinez et al. (2005). They found that the relative frequencies of the three main genetic types of B. cinerea (Group I, Group II vacuma and Group II transposa) during 1998–2000 in three Bordeaux vineyards differed significantly with respect to the different grapevine phenological stages and organs.

The relative importance of sclerotia and necrotic trash over longer periods than one growing season has not been fully investigated. However, B. cinerea seem to be viable in buried sclerotia for longer than the sporulation ability shown for rachides in this study. Thomas et al. (1983) showed that soilborne sclerotia of B. cinerea could survive in the field for up to 15 months. Work done by Hsiang & Chastagner (1992) indicated a longer period of survival, with B. cinerea sclerotia buried at 10 and 20 cm in field soil having an average viability of 77% after 18 months. However, Raposo et al. (2000) reported that when placed on soil surfaces in summer (June) in Madrid, Spain, sclerotia had significantly reduced longevity which appeared to be related to temperature. Those sited outside the greenhouse had 91·3 and 38·5% survival and those inside the greenhouse had 32·5 and 5·1% survival after 100 days in 1995 and 1997, respectively.

In this study, the overall mean differences in sporulation capacity of under-canopy and inter-row trash were due to the differences in amount of trash rather than any important differences in viability of saprophytic B. cinerea within it. However, these trash samples were collected in early summer and so the effect of mulching by mowing the inter-row and the expected degradation by competing saprophytes within the inter-row sward was not well-advanced. It is likely that the integrity of ground trash and viability of saprophytes within it could be reduced by the activities of soil flora and fauna, which are likely to be much greater in the moist shaded inter-row than the dry, exposed under-canopy area. Swift (1984) concluded that the activities of many saprophytic organisms were limited by the physical size of the substrate, and that optimum degradation first required comminution. The importance of comminution in degradation of the trash is recognized by Marlborough viticulturists who instruct pruning staff to thrown the pruned trash into the inter-row where it is broken up by mowing.

The risk presented by each trash type depended on the volume present and its TSSA. Canes were present in large quantities, but their sporulation potential was low in the laboratory incubation test. Rachides contributed 38% of the total sporulation potential of the trash, although they comprised only 5% of the total surface area of trash pieces, while tendrils and petioles comprised 18 and 29% of the surface areas of trash pieces and provided 21 and 12% of the sporulation potential, respectively. The differences in trash amounts and composition between vineyards were probably due to differences in vegetative and reproductive growth of the vines, as well as sanitation and harvesting practices. However, the sanitation practices used in New Zealand vineyards, which usually involves mulching the debris by mowing the inter-row, are likely to be more effective for large trash pieces such as canes than for small pieces such as rachides and tendrils, which may pass between the blades of the mowers.

The few rachides found within the canopies of all vineyards surveyed indicated a high rate of compliance by pruners who are instructed to remove them from the canes being tied down for the next year’s growth. The average of 0·2 rachides found per vine represented a reduction from the 30–60 originally left in the canopy by machine harvesters, which remove the berries from each bunch but leave the rachis attached to the vine. Harvesting grapes by machine is still the most common harvesting method in New Zealand, and was used in all vineyards surveyed, thus rachides were found in the trash of the vineyards.

The experiment that placed overwintering rachides in the canopy and on the ground during the growing season found that they were capable of producing B. cinerea conidia through to harvest, although the mean conidium numbers per rachis did decrease over time, from 3·9 × 105 and 3·5 × 105 at capfall, to 2·6 × 104 and 2·7 × 103 at harvest, for canopy and ground positions, respectively. The effect of the ground environment was clear and was probably due to the presence of moisture and soil microflora or microfauna which could reduce the viability of saprophytic B. cinerea mycelium and the rachis substrate (Swift, 1984). The rachides used to set up this experiment were collected from the under-canopy area within one vineyard. Before collection, they were in the canopy for 2–3 months and on the ground, where they were thrown by pruners, for a further 2–3 months. It is possible that the survival of the saprophytic B. cinerea within them might have been affected by the winter environment in these positions. Further research could investigate the B. cinerea sporulation over time from rachides that remained in the canopy for the entire winter period.

The differences over time recorded in sporulation potential of rachides placed in different positions within each vineyard, indicated the potential effects of the external environment on sporulation by B. cinerea. Sosa-Alvarez et al. (1995) reported that B. cinerea sporulated more frequently with rising temperatures in spring and Rotem et al. (1978) suggested that cooler temperatures preserve sporulation potential of some saprophytic plant pathogens in the longer term, because the nutrients within the mycelium were not depleted by the sporulation events. In this study the drier canopy environment was shown to preserve the sporulation capacity of the rachides, possibly due to fewer sporulation events. Exhaustion of resources by repeated sporulation of B. cinerea from strawberry mummies was the reason given by Braun & Sutton (1988) during a time course of sampling mummies in the field. Thomas et al. (1983) also concluded that B. cinerea survival on pruned grapevine wood was lower at 15°C than the other temperatures tested because this temperature was most favourable for sporulation. In this study, it seems likely that although the rachis still remained relatively intact, the viability of the pathogen within it may have been reduced by sporulation.

When rachides from the same source were maintained in vineyard positions they exhibited less sporulation than when placed under moist conditions in the laboratory, which indicates that they probably have lower sporulation potential in the field. Sosa-Alvarez et al. (1995) also concluded that longer wetness periods were required for B. cinerea sporulation from trash in the field than in the laboratory. They reported that in Ohio, B. cinerea sporulated sporadically from infested strawberry leaf trash in the field, and only after long periods of wetness. In this study, the rachides were likely to be too dry to respond well to the limited rainfall events recorded. However, the longevity of rachides as an inoculum source may be very important for disease development because even limited sporulation from rachides could initiate colonization of other necrotic tissues such as leaves cut off by summer trimming, whenever rainy weather occurs. The role of conducive weather conditions in initiating ongoing cycles of B. cinerea sporulation and infection has been well-documented (Elmer & Michailides, 2004; Holz et al., 2004). The prolonged survival of the pathogen in grapevine debris could therefore ensure availability of airborne B. cinerea spores throughout the growing season.

The physical state of the rachides was expected to be an indicator of their value as a substrate and to reflect the sporulation potential. Results showed some evidence of degradation by the end of the season, with rachides on the ground losing approximately 29 and 23% in weight and numbers of pedicels, respectively, while rachides placed in the canopy lost only 3 and 0%, respectively. Length of rachides did not change appreciably over the season, with some later measurements even indicating a minor gain, an effect that was probably due to problems with methodology rather than the effects of the treatments. The role of pedicels as sites for fungal survival was also reported by Daykin & Milholland (1984) who found that Colletotrichum gloeosporioides-infested pedicels produced conidia until véraison and mummies infected with the fungus produced conidia through to harvest, although the number of conidia produced gradually decreased.

The results from this study show that trash colonized by B. cinerea, especially rachides, can produce conidia through much of the growing season. The longevity of trash as an inoculum source may be very important for disease development because B. cinerea is likely to initiate secondary cycles of inoculum production on other necrotic tissues, such as trimmed leaves. This could ensure sufficient inoculum is available by the time of fruit ripening to initiate an epidemic of bunch rot if conducive weather conditions prevail during that period.

Acknowledgements

The authors wish to thank the Foundation for Technology and Research (NZ) for funding this PhD study and the many grape growers who allowed this study to be conducted within their vineyards.

Ancillary