Co-infection by Botryosphaeriaceae and Ilyonectria spp. fungi during propagation causes decline of young grafted grapevines

Authors


Abstract

Decline of newly planted, grafted grapevines is a serious viticultural problem worldwide. In the Riverina (New South Wales, Australia), characteristic symptoms include low fruit yields, very short shoots and severely stunted roots with black, sunken, necrotic lesions. To determine the cause, roots and wood tissue from affected plants in 20 vineyards (Vitis vinifera cv. Chardonnay grafted to V. champini cv. Ramsey rootstock) were assayed for microbial pathogens. Ilyonectria spp. (I. macrodidyma or I. liriodendra, producers of phytotoxin brefeldin A, BFA, and cause of black foot disease of grapevines) and Botryosphaeriaceae spp. (predominantly Diplodia seriata) were isolated from rootstocks of 100 and 95% of the plants, respectively. Togninia minima and Phaeomoniella chlamydospora (cause of grapevine Petri disease) were isolated from 13 and 7% of affected plants, respectively. All Ramsey rootstock stems of grafted plants sampled from a supplier nursery were infected with Ilyonectria spp. and D. seriata. Diplodia seriata, but not Ilyonectria spp., was also isolated from 25% of canes sampled from the rootstock source block. Root inoculation of potted, disease-free Chardonnay plants with Ilyonectria isolates from diseased vineyards caused typical disease symptoms, while co-inoculation with Botryosphaeriaceae spp. increased disease severity. This is the first study to show that a major cause of young grapevine decline can be sequential infection by Botryosphaeriaceae from rootstock cuttings and Ilyonectria spp. from nursery soil. Although the Petri disease fungi were less common in young declining grafted grapevines in the Riverina, they are likely to contribute to the decline of surviving plants as they mature.

Introduction

Since 2000, wine grape growers in Australia's Riverina wine region of southwestern New South Wales have reported the retarded growth and death of many newly planted grapevines (Fig. 1). Characteristic symptoms include low fruit yields, very short shoots and severely stunted roots with few feeder roots and black sunken necrotic lesions. The syndrome occurs in the absence of other common debilitating conditions of young grapevines such as water stress, waterlogging, nutrient deficiency and parasitic nematodes, and is not confined to particular soil types. In 2006, survey questionnaires were sent to 550 Riverina wine grape growers. The replies showed that over 65% had encountered young grapevine decline in their vineyards.

Figure 1.

A typical diseased young grapevine (Vitis vinifera cv. Chardonnay on V. champini cv. Ramsey rootstock) in a Riverina vineyard affected by young vine decline. Note the very short shoots compared to normal vines in the background.

Young grapevine decline is found in many regions worldwide, and is commonly caused by grapevine pathogens that cause black foot disease (Ilyonectria spp.), Petri disease (Phaeomoniella chlamydospora and Togninia spp.) and botryosphaeria (‘Bot’) canker (Botryosphaeriaceae fungi; Mugnai et al., 1999; Gramaje & Armengol, 2011). However, the causal species vary considerably between regions. The objectives of this study were to identify the pathogens that cause young grapevine decline in the Riverina, the sources of infection and the impact on plant growth and productivity.

In early diagnostic samples of declining Riverina grapevines, species of Ilyonectria, and less commonly of Botryosphaeriaceae, were consistently isolated from the roots (M. A. Whitelaw-Weckert, unpublished data). Sweetingham (1983) described a decline disease (black foot) affecting mature Vitis vinifera cv. Cabernet Sauvignon grapevines in Bream Creek, Tasmania, Australia, that was consistently associated with Cylindrocarpon destructans (subsequently renamed Ilyonectria macrodidyma). Furthermore, Whitelaw-Weckert et al. (2007) reported that Cylindrocarpon (now Ilyonectria) liriodendri also caused grapevine disease, with typical black foot symptoms, in the Hunter Valley, NSW, Australia.

In a previous study of declining grapevines in Australia, the Petri disease fungi P. chlamydospora and Togninia minima were consistently isolated from the trunk and branch samples of 124 individual plants sent for diagnosis by grape growers between 1998 and 2002 (Edwards & Pascoe, 2004). These authors isolated Cylindrocarpon and Botryosphaeria spp. from ‘wood’ samples from two and four, respectively, of seven grapevines showing Petri disease symptoms from the Riverina but, notably, the species were not found together. Furthermore, with two exceptions, all seven plants were non-grafted, most were known to be at least 4 years old and the ages of three of them ranged between 10 and 50 years. Although some of the plant symptoms, viz. graft failure, weak growth, shoot dieback and gradual death of the grapevine, were similar to those reported from the young Riverina grapevine study, the characteristic symptom of Petri disease, internal black wood streaking and black tar-like exudation from severed stems and trunks, was not common in the Riverina samples. Moreover, the roots of young declining Riverina grapevines were severely stunted, a feature not examined in the Edwards & Pascoe (2004) diagnostic study.

Accordingly in the current study, a comprehensive search for putative fungal pathogens within both wood and root samples from affected vineyards was conducted. Although a range of scion and rootstock cultivars were reported to be affected, this study focused on V. vinifera cv. Chardonnay grafted to Vitis champini cv. Ramsey, the predominant scion–rootstock combination in the Riverina. Aerial remote sensing was used to reveal the spatial distribution of affected plants to aid epidemiological interpretation. Putative causal fungi were tested in glasshouse pathogenicity studies to fulfil Koch's postulates. Some Cylindrocarpon/Ilyonectria isolates produce the phytotoxin brefeldin A (BFA) in vitro (Sweetingham, 1983; Barbetti, 2005). As BFA (‘cyanein’) inhibits mitosis in root tips (Frank, 1974) and can stunt the growth of grapevines (Sweetingham, 1983), Riverina isolates of Ilyonectria were screened to determine their ability to produce BFA in vitro.

The origins of fungal infection were identified in an epidemiological study of an affected vineyard, the supplier propagation nursery and a rootstock source block. Concurrently, the impact of the condition on grapevine growth and fruiting was determined over three successive seasons in a typically affected Riverina vineyard.

Materials and methods

Survey of Riverina vineyards

Fungal species

A survey for incidence of wood and root fungal pathogens was conducted in 20 Riverina vineyards showing decline symptoms in V. vinifera cv. Chardonnay grapevines grafted on to V. champini cv. Ramsey rootstock. Samples were obtained from young grapevines, ranging from 1 month to 8 years old, that were reported to have been affected by young vine decline symptoms from the time they were planted.

Three diseased and three symptomless grapevines were removed intact from each of the 20 vineyards. Roots were exposed by carefully washing the soil away. Segments of roots, stems and shoots were cut longitudinally and transversely to expose any internal symptoms of vascular staining. Tissue segments, from which exfoliating bark had been removed, were surface-sterilized for 3 min with sodium hypochlorite (1% active chlorine) and rinsed three times with sterile deionized water (SDW). Root segments (3 mm), thin disks cut from whole cross sections and small slivers (1 × 5 mm) cut from woody tissue were placed on dichloran Rose Bengal chloramphenicol agar (DRBC, Oxoid) and incubated at 25°C for 4 weeks in darkness, with daily inspection for fungal colonies. Prior tests had shown that DRBC slows the growth of Botryosphaeriaceae and Ilyonectria spp. sufficiently to enable isolation of P. chlamydospora and Togninia spp. without overgrowth by faster growing fungi. Similarly treated tissue samples were also placed on moist paper towelling in sterile containers and incubated at room temperature for up to 6 months to allow for the slower growth of P. chlamydospora and Togninia spp. The paper towelling was remoistened periodically as necessary. Phaeomoniella chlamydospora and T. minima were isolated from mixtures of fungi and bacteria in xylem exudates from the incubated wood. The exudates were filtered when necessary to remove larger spores (Rooney et al., 2001) and streaked onto DRBC using a microbiological loop.

All developing fungal colonies were transferred to potato dextrose agar (PDA, Oxoid) and incubated in darkness for 7 days. In order to promote sporulation, cultures were incubated in clear plastic boxes on the bench. In addition, Botryosphaeriaceae isolates were incubated with sterilized pine needles on 2% water agar, according to the method of Smith et al. (1996) and were examined weekly for formation of pycnidia and conidia.

The growth rates and colour of all isolates on PDA at 25°C in darkness were measured using colonies generated from 5-mm diameter mycelial plugs obtained from the margins of 5-day-old single-spore PDA colonies. Fungal species were identified by colony and conidial morphology using the following references: Ilyonectria (Halleen et al., 2004, 2006), Botryosphaeriaceae (Alves et al., 2008; Phillips et al., 2012), Phaeoacremonium spp. (Mostert et al., 2006) and T. minima (Crous & Gams, 2000). All fungal isolates were further identified by comparison of the β-tubulin gene (Table 1) and ITS sequences (data not shown) of typical Riverina isolates with sequences from GenBank. DNA extraction, PCR amplification and sequencing were performed on representative Ilyonectria, Phaeomoniella, Togninia and Botryosphaeriaceae isolates using universal primers ITS1 and ITS4 for the 5·8S ITS region of the rDNA and primers T1 and T2 for the partial β-tubulin 2 gene (O'Donnell & Cigelnik, 1997). These two sets of primers are considered suitable for identification of these fungi (Gramaje & Armengol, 2011). blast was used to compare the sequences with those of known fungi archived at GenBank.

Table 1. Australian grapevine fungal isolates sequenced and used in brefeldin A and pathogenicity studies
DAR culture no.a (lab number)Plant organ, location of isolationGenBank accession number (β-tubulin)Closest blast match GenBank accession no., (isolate number), species% identityBrefeldin A producerPutative species
  1. NT, not tested; NA not applicable.

  2. a

    DAR numbers relate to isolates logged with the Plant Disease Herbarium Collection of the Agricultural Scientific Collection Unit at NSW DPI in Orange, New South Wales, Australia.

  3. b

    Sequenced using β-tubulin primers T1/T2.

  4. c

    Sequenced using using universal primers ITS1 and ITS4 for the 5·8S ITS region of the rDNA.

DAR77402 (MW126)

Root,

Riverina,

NSW,

Australia

JN626950

JF735469,

(CBS 112603)

Ilyonectria macrodidyma

100YesI. macrodidyma complexb

DAR81459

(MW604)

Trunk,

Riverina,

Australia

JX901051

JF35469,

(CBS 112603)

I. macrodidyma

100YesI. macrodidyma complexb

DAR81460

(MW725; P42 UTAS)

Root,

Bream

Creek,

Tasmania,

Australia

JX901052

HM214449

(I: Cymbidium roots, southern China)

I. macrodidyma

100YesI. macrodidyma complexc

DAR81461

(MW613)

Trunk,

Riverina,

NSW,

Australia

JN626949

HM214449

(I: Cymbidium roots, southern China)

I. macrodidyma

100YesI. macrodidyma complexb

DAR81690

(MW580)

Root,

Riverina,

NSW,

Australia

JX901053

JF35469,

(CBS 112603)

I. macrodidyma

100YesI. macrodidyma complexb
(MW585)

Root,

Riverina,

NSW,

Australia

JX901054 JF735460 (Cy115) I. macrodidyma100YesI. macrodidyma complexb

DAR81853

(MW1117)

Nursery soil,

Riverina,

NSW,

Australia

JX901056

JN859422, REF202,

I.macrodidyma

99NTI. macrodidyma complexb

DAR81691

(MW587)

Root,

Riverina,

NSW,

Australia

JX901055 JF7355469 (CBS 112603) I. macrodidyma100YesI. macrodidyma complexb

DAR77396

(MW122, MW806)

Root,

Riverina,

NSW,

Australia

JN626951 DQ178171 (CBS 117526) I. liriodendri100Yes I. liriodendri b

DAR81462

(MW630)

Trunk,

Riverina,

NSW,

Australia

JX901050 DQ356355; Diplodia seriata99NA D. seriata c

The isolates used in this study were deposited in the Plant Disease Herbarium Collection of the Agricultural Scientific Collection Unit at the Department of Primary Industries in Orange, NSW (DAR).

Spatial distribution of young declining grapevines

Normalized difference vegetation index (NVDI), a common indicator of photosynthetically active biomass used in remote sensing, was used to map the distribution of young declining grapevines in one of the surveyed vineyards, 4 years after the grapevines were planted, using high-spatial-resolution aerial multispectral images (Hall et al., 2011). Remote sensing imagery was acquired close to véraison (the onset of ripening, when grapevine canopies are normally fully developed) using a custom-built multispectral airborne digital imaging system. The system comprised four identical digital video cameras with 994 × 1014 pixel arrays attached to an image-capturing PC mounted within a Cessna 310. Each camera lens was filtered to produce four spectral bands, centred at 450, 550, 675 and 800 nm (having 25 nm bandwidths for all but the 800 nm filter, which had a 40 nm bandwidth). Images were acquired at an altitude of 760 m above ground level, delivering a pixel resolution of c. 0·25 m. Imagery was collected within an hour of solar noon, minimizing shadow. The geographic coordinates of 18 ground control points located at the edges and corners of the vineyard block were used to register the image to map coordinates using a spatial warping function from image processing software (envi v. 4.7 (Linux 64), ITT Visual Information Solutions). The red and near-infrared (NIR) bands were converted to an NDVI image. NDVI provides a quantitative estimate of photosynthetically active biomass (PAB), commonly referred to as canopy vigour in a viticultural context, on an image cell-by-cell basis. NDVI values range between −1 and 1. Larger NDVI values indicate a greater level of PAB; values below zero are rare in natural scenes. Canopy vigour of the sample plants was estimated by extracting the NDVI values of representative pixels at the locations of the sample plants collected using a differential global positioning system (GPS). Using a geographic information system (GIS; arcmap and arcinfo, v. 10.0, ESRI Inc.), pixels within a radius of half the along-row inter-vine spacing were identified and summed to give a total vigour metric for each sample plant. NDVI values for individual plants were correlated against preassigned proximal, observational ratings of the health status (i.e. affected or not affected) of 30 plants in the vineyard.

Pathogenicity

Once identified, I. macrodidyma, I. liriodendri and Diplodia seriata isolates from Riverina grapevines, plus an I. macrodidyma isolate of known pathogenicity from a Tasmanian grapevine (DAR81460; Sweetingham, 1983), were selected for pathogenicity studies (Table 1).

Ilyonectria spp. isolated from diseased vineyards

The stems of dormant 1-year-old Chardonnay plants were trimmed to 0·1 m long, and the roots were trimmed to 50 mm. The shortened plants were grown in 500-mL pots filled with twice autoclaved coarse river sand: loam: Canadian peat moss (2:2:1) (field capacity 7·4%, pH (CaCl2) 5·61). A 10-mm diameter auger was used to remove three cores of potting mixture aseptically from the pots and 0·3 g sterilized wheat germ was placed into each hole. Two 5-mm discs from 21-day-old PDA colonies of I. liriodendri (DAR77396), I. macrodidyma (DAR81459) and I. macrodidyma (DAR81460) were applied to the wheat germ cores and the holes were back-filled. Four uninoculated pots were treated in the same way with SDW as controls. After an initial watering, each pot was topped with a sealed clear plastic dome to prevent air contamination and maintain high humidity. The 16 pots were arranged in randomized complete blocks in a glasshouse maintained at 15–25°C. After 16 weeks’ growth, the plants were destructively sampled. The roots were scored for root rot, where 0 = no disease symptoms, 1 = 1–25%, 2 = 26–50%, 3 = 51–75% and 4 = 76–100% root rot. Washed roots were placed into a Perspex tray for scanning using an Epson scanner. Images were saved and analysed using WinRHIZO Pro 2007c software. Root and shoot dry weight (d.w.) (dried at 50°C until constant weight) were also determined as previously described. Roots and stem segments were surface sterilized and incubated first on DRBC and then PDA as described previously.

Root subsamples for cryoscanning electron microscopy (cryoSEM) were washed and frozen in liquid N2 on sampling and stored at −80°C. Root sections were mounted on stubs with carbon paste and frozen in liquid N2 slush, and then transferred to the cold stage of a cryoSEM (Cambridge Model S360, Cambridge Instrument Co.), etched for a few minutes at −85°C to show the cell outlines, cooled to –160°C, coated with gold and observed at 18 kV (McCully et al., 2000). In addition, woody tissue was softened in boiling glycerol and water (1:10; v/v), sectioned by hand (c. 250 μm), fixed (formalin–propionic acid–alcohol, FPA) and dehydrated before being dried in a critical point dryer (K850; Quorum Technologies), mounted onto stubs using carbon conducting tape (ProSciTech) and then coated with gold (3 min at 25 mA) using a sputter coater.

Co-infection by I. macrodidyma and D. seriata

The roots of nine 1-year-old Chardonnay plants, each with one leafy shoot (1 m long), were trimmed to 0·15 m length. Immediately afterwards, in bright sunny conditions, the roots and basal trunk of six plants were submerged for 90 min in a suspension of conidia and mycelium of I. macrodidyma DAR81461 (1 × 106 conidia mL−1). Three control plants were submerged in water only. The plants were then pruned to three basal buds and planted in 0·28-m diameter pots (8·1 L capacity) containing steam-pasteurized (65°C for 45 min) coarse river sand: loam: Canadian peat moss (2:2:1). The pots were arranged in randomized complete blocks in a glasshouse maintained at 15–25°C and watered daily to field capacity. After 7 weeks, 10 mL aliquots of D. seriata DAR81462 suspension (1 × 105 conidia mL−1) were applied to the surface of three of the pots previously inoculated with I. macrodidyma and washed in with 0·1 L water. Previous investigations had shown that this was an effective method to infect potted grapevines, probably via uncovered pith at the base of the stem (Whitelaw-Weckert et al., 2006). Equivalent volumes of water were added to the remaining pots. The plants were destructively sampled 16 weeks after the D. seriata inoculation. The roots were scored for root health, where 0 = no roots remaining, 1 = 75–100% blackened, 2 = 50% blackened, 3 = 33% blackened and 4 = 100% healthy roots. Root and shoot d.w. were determined as previously described. Recovery of the inoculum from roots and stems was used to fulfil Koch's postulates. This was tested by surface sterilizing and incubating subsamples of roots, stems and shoots on DRBC and PDA as described previously.

Brefeldin A production

One I. liriodendri and six I. macrodidymum isolates (obtained from plants with symptoms from different vineyards) were tested for production of BFA in vitro following an amended method of Weber et al. (2004). Ethyl acetate extracts from malt extract agar (Oxoid) cultures were analysed by HPLC using a LiChiroCART 250 × 4 mm column with LiChrosphere-100 RP-18 5 μm packing. Operating conditions were: flow rate 1·4 mL min−1; injection volume 100 μL; column temperature 40°C; spectral acquisition range 200–600 nm; solvent A water; solvent B methanol; solvent gradient (step, time, proportion of solvent A) (i) 0 min 100%, (ii) 20 min 30%, (iii) 30 min 0%, (iv) 35 min 0%, (v) 40 min 100%, (vi) 50 min 100%.

Epidemiology

In 2008, 1075 of 6200 dormant grafted grapevines (V. vinifera cv. Pinot Noir on Ramsey rootstock) planted in a Riverina vineyard failed to shoot in the first spring. The most visibly affected plants were replaced immediately but many others grew with typical symptoms of young grapevine decline in the following years. The propagation nursery had grafted dormant scions to dormant rootstock cuttings obtained from a single rootstock source block and field-grown them for one season before dispatch.

A possible connection between the grapevine wood/root pathogens isolated from the vineyard, the nursery and the rootstock source block was investigated in this study. Soil samples from a depth of 0–0·1 m from the vineyard and the nursery were used for a plant baiting experiment. The vineyard soil was collected 2 m away from young declining plants, where it was not colonized by roots of affected plants. Nursery soil was collected from the source field that had been used to propagate the grapevines showing symptoms. In addition, a subsample of the soil from the source field was sterilized by autoclaving (121°C, 20 min, three times over 6 days).

Dormant 1-year-old Chardonnay plants, confirmed to be free of fungal pathogens by plating subsamples as described previously, were planted in 1·5 L pots containing 1·2 kg d.w. soil from each source. The pots were arranged in randomized complete blocks in a glasshouse at 15–25°C and watered daily to field capacity. The plants were destructively sampled 8 months later. Root and shoot d.w. (dried at 50°C until constant weight) were determined. Roots and stem segments were surface sterilized and incubated first on DRBC and then on PDA as described earlier to detect fungal species transmitted from the soil.

Root, rootstock trunk and scion samples were taken from ten 1-year-old, dormant, field-grown, grafted grapevines (V. vinifera cv. Colombard on Ramsey rootstock) in the nursery. Two lignified shoots from each of four Ramsey grapevines in the rootstock source block were also sampled. To detect fungal species, roots and stem segments were surface sterilized and incubated first on DRBC and then on PDA as described previously.

Longer term impacts of young vine decline

The growth and fruiting of 30 randomly selected 5-year-old Chardonnay grapevines on Ramsey rootstock with typical decline symptoms from the time of planting (subsequently confirmed as co-infected by I. macrodidyma and D. seratia) was compared with 30 randomly selected symptomless plants in seasons 5–8 (consecutively) after planting. Total fruit weight and bunch number were determined at harvest. Shoot elongation was determined weekly from soon after budburst and the total weight of pruned canes was determined after leaf fall.

All field and pot experimental data were subjected to analysis of variance (anova) and least significant differences (LSD) by GenStat v. 14.

Results

Pathological survey

Sixty symptomless and 60 grapevines with symptoms were sampled from 20 Riverina vineyards. No pathogens were isolated from the rootstock roots or trunks of symptomless plants. Ilyonectria macrodidyma and Botryosphaeriacae fungi were the predominant pathogens isolated from the rootstocks of plants with symptoms; these were found together in 95% of the vineyard samples with symptoms. Botryosphaeriaceae fungi were isolated from the scion cordons of 18% of symptomless plants. The lack of infection in the rootstock or scion trunks of symptomless plants indicates that these cordons were probably infected by airborne spores through pruning wounds, and not during propagation (data not shown).

Ilyonectria spp.

Ilyonectria macrodidyma was the most common pathogen, present on roots, rootstock, wood and graft unions of young diseased grapevines. It was isolated from the rootstock of 100% of plants with symptoms, where it was found in roots and stained patches of wood from the oldest xylem rings (Table 2; Fig. 2a). blast searches of GenBank (using both β-tubulin and 5·8S ITS sequences) established that the typical I. macrodidyma isolated from young declining grapevines in the Riverina may belong within the new ‘I. macrodidyma complex’ (Cabral et al., 2012b). The β-tubulin sequence for the less commonly isolated I. liriodendri isolates was identical to that of an I. liriodendri grapevine black foot isolate previously isolated from Hunter Valley, NSW, Australia (Whitelaw-Weckert et al., 2007). The phytotoxin BFA was produced by all Riverina Ilyonectria spp. tested (Table 1).

Table 2. Fungal pathogens isolated from grafted grapevines (Vitis vinifera cv. Chardonnay on V. champini cv. Ramsey) with symptoms of young vine declinea
Pathogenic fungiGrapevines from which fungi were isolated (%)
RootstockScion
Any position on rootstockRootstock rootsRootstock trunk below graftGraft unionAny position on scionScion trunkScion cordon
  1. a

    Three samples were taken from each of 20 affected Riverina vineyards. No pathogenic fungi were isolated from the rootstock (roots or trunk) of symptomless plants, although Botryosphaeriaceae spp. were isolated from the scion cordons of 18% of symptomless plants.

Ilyonectria spp.       
Ilyonectria macrodidyma 100792933705
Ilyonectria liriodendri 8008000
Botryosphaeriaceae spp.       
Diplodia seriata 79105542271020
Diplodia mutila 8568750
Lasiodiplodia theobromae 8008000
Petri disease fungi       
Togninia minima 13068201010
Phaeomoniella chlamydospora 700033525
Figure 2.

Plant tissue from vineyards affected by young grapevine decline. (a) Vitis champini cv. Ramsey stem from a young vine decline-affected newly planted vineyard. Diplodia seriata and Ilyonectria macrodidyma were isolated from both roots and rootstock wood. (b) Ramsey rootstock stem with wedge-shaped discoloration. This grapevine was in a newly planted vineyard with a 40% strike rate. Diplodia seriata and I. macrodidyma were isolated from roots (not shown) and D. seriata was isolated from rootstock wood. (c) Transverse section in trunk showing I. macrodidyma hyphae (h) in ray cells (periodic Schiff's reagent, analine blue, fluorescence microscopy). Scale bar = 100 μm.

Botryosphaeriaceae spp.

Diplodia seriata, Diplodia mutila or Lasiodiplodia theobromae were isolated from 79, 8 and 8% of rootstocks from plants with symptoms, respectively. Interestingly, these three Botryosphaeriaceae species were never found together within one vineyard (i.e. in each vineyard, plants were found to be infected with only one of the species). In the few vineyards where D. seriata was not found in the rootstocks of plants with symptoms, either D. mutila or L. theobromae was usually present in its place. Diplodia seriata and D. mutila were present in the roots of 10 and 5% respectively of plants with symptoms, but were found more often in the rootstock trunk and graft union where they were associated with dark, wedge-shaped discoloration (Fig. 2b). Diplodia seriata and D. mutila were also found colonizing the scion in 27 and 7% respectively of plants with symptoms (Table 2). The identities of selected Riverina D. mutila and L. theobromae isolates (data not shown) and a D. seriata (Table 1) isolate were confirmed by blast comparison of β-tubulin sequences.

Petri disease fungi

Togninia minima and P. chlamydospora were not consistently associated with plants with symptoms, but were isolated from 13 and 7% respectively of affected rootstocks and from 20 and 33% of affected scions, respectively (Table 2). The identities of typical T. minima and P. chlamydospora isolates were confirmed by blast comparison of β-tubulin and ITS sequences respectively (data not shown).

Spatial distribution of affected grapevines within a vineyard

The spatial position of individual plants with symptoms within a typical Riverina vineyard could be identified by remote sensing using the NDVI, which indicated that the affected grapevines were randomly distributed over the entire planted area. The mean NDVI score for affected plants was significantly lower than for symptomless plants (0·192 and 0·98, respectively). There was also a large difference between the data sets for affected and symptomless plants in terms of their range and variability. Much greater variability in mean NDVI amongst the affected plants was present: the interquartile range of the affected set was 0·085 compared to 0·055 for the symptomless set.

Pathogenicity

Ilyonectria spp. isolated from young declining grapevines

At 16 weeks after soil inoculation, none of the three isolates tested, I. macrodidyma DAR81459 (from the Ramsey trunk of a young declining Riverina grapevine), I. macrodidyma DAR81460 (from a diseased Tasmanian grapevine root; Sweetingham, 1983) and I. liriodendri DAR77396 (from a Riverina grapevine root), had affected shoot growth (data not shown) but they had all caused significant root rot symptoms in the potted Chardonnay plants (Table 3). The three Ilyonectria isolates significantly decreased the number of lateral roots and the length of very fine roots (0–0·2 mm diameter), although they did not affect the length of roots thicker than 0·2 mm. The I. macrodidyma isolate DAR81459 had the most deleterious effect on root structure. The symptoms and subsequent recovery of the inoculated fungi from the roots of every diseased plant, but not from the control plants, fulfilled Koch's postulates as to the pathogenicity of the Ilyonectria isolates.

Table 3. Root growth of potted Chardonnay grapevines in soil inoculated with Ilyonectria species isolated from young declining grapevines, 16 weeks after inoculation
Ilyonectria spp.Root rot symptomsTotal number root lateralsTotal length rootsLength of roots (mm) of different diameters
0–0·2 mm0·2–0·4 mm0·4–0·6 mm0·6–0·8 mm0·8–1·0 mm1·0–1·2 mm1·2–1·4 mm>1·4 mm
  1. Values within a column followed by the same letter are not significantly different based on LSD (P < 0·05).

Uninoculated control1·0 a3232 a33384·1 a98·849·833·725·111·010·810·1
I. liriodendri, DAR773961·5 a1075 b24034·0 b62·052·926·620·68·57·95·6
I. macrodidyma DAR814603·0 b1381 b23734·8 b68·442·931·325·118·27·87·2
I. macrodidyma, DAR814594·0 b136 c284·7 c6·53·62·51·42·52·13·8
P <0·001<0·0010·061<0·0010·1160·3770·2400·2140·2620·5690·625
LSD1·311051802962·741·926·719·312·69·5313·1

CryoSEM of the infected roots revealed that the Riverina Ilyonectria isolates had degraded the root cortical cells. Their hyphae were present in dead root cells devoid of water and normal cell contents, but were not present in live, undisrupted cells (Fig. 3a). The hyphae were attached predominantly on root hairs (Fig. 3b,e) and at root branch junctions (Fig. 3c), and were commonly aggregated into hyphal strands on the root surface (Fig. 3d). Clusters of macro- and microconidia were also observed on the root surface (Fig. 3d).

Figure 3.

Cryoscanning electron microscopy (cryoSEM) micrographs from pot experiments with Ilyonectria and Botryosphaeriaceae species isolated from Riverina young vine decline affected vineyards. (a–d) Chardonnay roots inoculated with I. macrodidyma alone, (e) Chardonnay root in soil inoculated with I. liriodendri alone. (f–h) cryoSEM from Chardonnay plants with roots co-inoculated with I. macrodidyma and Diplodia seriata. (a) Ilyonectria macrodidyma in decomposed root cortical cells. Hyphae (*) occur only in dead, empty cells and not in living cells with normal cell solutes (×). (b) Ilyonectria macrodidyma hypha (white arrow) on surface of root, attached to root hair (rh). (c) Ilyonectria macrodidyma on root branch junction, showing microconidia (mic) and hyphae (white arrows). (d) Ilyonectria macrodidyma on root surface, showing macroconidia (mac), microconidia (mic) and hyphal strands (hs). (e) Ilyonectria liriodendri hyphae on branch point of a blackened root. Hyphae are attached to root hair (rh). (f) Diplodia seriata hypha (arrow) inside trunk xylem, 10 cm above ground. (g) Diplodia seriata (Bot) and I. macrodidyma (Ilyo) hyphae inside shoot xylem, 4 cm from the trunk/shoot intersection. (h) Ilyonectria macrodidyma microconidium (arrow) near pit membrane (arrowhead), inside xylem of 3-mm diameter root. Scale bars: a, h = 10 μm; c, d, e, f, g = 20 μm; b = 50 μm.

Co-infection with I. macrodidyma and D. seriata

Potted Chardonnay grapevines that had been exposed to root and basal stem inoculation with Riverina I. macrodidyma isolate DAR81461 alone had significantly decreased shoot length (43%), shoot d.w. (50%), total leaf d.w. (61%), d.w. per leaf (33%), number of leaves (42%) and root d.w. (66%) after 16 weeks’ growth (Table 4). Dual root/basal stem inoculation with I. macrodidyma and D. seriata respectively caused a further decrease in total leaf d.w. (63%), total number of leaves (21%), root health rating (29%) and fine root rating (49%) compared to inoculation with I. macrodidyma alone. As mass per leaf was unchanged, the decrease in total leaf mass was due to decreased leaf numbers. The inoculated fungi were reisolated from the roots of every inoculated plant but not from the control plants, thus fulfilling Koch's postulates in regard to the causes of the plant responses. Diplodia seriata was also reisolated from surface-sterilized trunk tissue, indicating acropetal growth.

Table 4. Co-infection of potted Chardonnay by Diplodia seriata and brefeldin A-producing Ilyonectria macrodidyma, 16 weeks after inoculation. The inoculum was applied via wounded grapevine roots, analogous to the situation where contaminated grapevines are planted in non-contaminated (clean) vineyard soil
TreatmentTotal shoot length (cm)Shoot d.w.a (g)Total leaf d.w. (g)d.w. per leaf (mg)Number of leavesRoot d.w. (g)Root health ratingFine root rating
  1. a

    d.w., dry weight.

  2. Values within a column followed by the same letter are not significantly different based on LSD (P = 0·05).

Uninoculated control844 a77·6 a54·0 a45 a120 a116·9 a4·70 a7·29 a
Trimmed roots inoculated with I. macrodidyma DAR81461478 b38·8 b20·8 b30 b70 b39·6 b4·32 a6·62 a
Trimmed roots inoculated with I. macrodidyma DAR81461 plus D. seriata DAR81462337 b19·0 b7·6 c30 b55 c27·1 b3·07 c3·36 c
P <0·0010·001<0·0010·007<0·001<0·0010·002<0·001
LSD17129·712·613·724·925·81·162·19

CryoSEM showed D. seriata and I. macrodidyma hyphae alone and together within trunk and shoot xylem (Fig. 3f,g), and I. macrodidyma spores within the root xylem (Fig. 3h).

Epidemiology

Grapevine fungal isolates from representative diseased vineyard

Six months after planting in the vineyard, I. macrodidyma was isolated from rootstock stem and roots from all sampled plants with symptoms. Diplodia seriata was also isolated from all rootstock stems and from 33% of graft unions of plants with symptoms (Table 5). The rootstock stems and the graft union of every diseased plant exhibited dark discoloured patches and wedges in cross section. Light microscopy of thin sections of the infected roots (data not shown) and trunks showed that the xylem was heavily colonized by fungal hyphae 1·5–2 μm thick, typical of I. macrodidyma hyphae (Fig. 2c). No pathogens were isolated from any scions, or from the roots, stems or graft unions of the healthy plants.

Table 5. Epidemiological case study: fungal pathogens isolated from diseased grafted vines (%)
SpeciesYoung diseased grapevines from the vineyard (6 months after planting)aSupplier nursery rootling grapevines and soilbRootstock source block plantsc
RootsRootstock trunk below graftGraft unionScionThick roots (≥4 mm)Rootstock trunk below graftGraft unionScionSoilCanes
  1. a

    Vitis vinifera cv. Pinot Noir grafted to V. champini cv. Ramsey rootstock in a representative Riverina vineyard.

  2. b

    1-year-old grapevine rootlings (V. vinifera cv. Colombard grafted to V. champini cv. Ramsey rootstock) and soil from the supplier nursery.

  3. c

    V. champini cv. Ramsey rootstock canes from source block plants.

Ilyonectria spp.
 I. macrodidyma 100100001000001000
 I. liriodendri 0000000000
Botryosphaeriaceae
 Diplodia seriata 010033000100025
 Diplodia mutila 0000000000
 Lasiodiplodia theobromae 0000000000
Petri disease fungi
 Togninia minima 000001617000
 Phaeomoniella chlamydospora 0000000000

Fungal isolates from nursery plants and rootstock source block

The rootstock stems of every plant sampled from the nursery had dark discoloured patches, wedges or spots in cross section. Ilyonectria macrodidyma was isolated from the thick (≥4 mm diameter) roots of all sampled rootstocks and soil samples. Diplodia seriata was isolated from 10% of rootstock graft unions, while T. minima was isolated from the stems and the graft unions of 16 and 17% of the plants, respectively. No P. chlamydospora was isolated from the nursery plants. No pathogens were isolated from the scions that appeared macroscopically uninfected in cross section. Although the canes from the rootstock source plants appeared macroscopically uninfected in cross section, D. seriata was isolated from 25% of the canes (Table 5).

Grapevine baiting experiment: pathogenic fungi in the nursery soil

Potted Chardonnay grapevines grown in soil from the source nursery field for 8 months had significantly lower root and shoot mass compared to those planted in the sterilized soil from the same site. This indicated that the soil must have contained a biotic factor limiting grapevine growth. Ilyonectria macrodidyma was isolated from the surface-sterilized roots of the bait Chardonnay plants, indicating that it was present in the soil. Although the root growth of plants in the vineyard soil was not significantly different to that of the plants in source vineyard soil, the shoot growth was greater. The absence of fungal pathogens in the roots of plants grown in the vineyard soil indicates a nursery source of plant infection (Table 6).

Table 6. Root and shoot growth of potted Chardonnay grapevines after 8 months’ growth in vineyard soil and in soil from the source field of the supplier nursery
Source of soilRoot dry weight (g)Shoot dry weight (g)Pathogenic fungi isolated from surface sterilized roots
  1. a

    Soil collected 2 m away from young, declining grapevines, in position not colonized by roots of those plants.

  2. Values within a column followed by the same letter are not significantly different based on LSD (P = 0·05).

Vineyard soila7·1 ab4·1 a0
Nursery soil from source field4·0 b2·3 bIlyonectria macrodidyma DAR81853
Autoclaved nursery soil from source field9·7 a3·5 a0
P 0·0330·05 
LSD5·41·7 

Longer term impacts of co-infection with I. macrodidyma and D. seratia

Three years of growth and yield data recorded for the typical Riverina affected vineyard (Chardonnay on Ramsey rootstock) showed that surviving diseased plants consistently produced 37–69% lower total fruit yields and 33–62% lower bunch numbers per plant. Shoot lengths were 44–45% shorter throughout the season and pruning weights were 38% lower in the affected plants (Table 7).

Table 7. Shoot growth and fruiting impacts of co-infection by Ilyonectria macrodidyma and Diplodia seratia on grapevines in a typical affected Riverina vineyard
Harvest parameterYear
2008–092009–102010–11
  1. Data from 30 diseased plants and 30 healthy plants (Vitis vinifera cv. Chardonnay on V. champini cv. Ramsey rootstock) in seasons 5 to 8 after planting.

  2. For any given season and disease status, yields marked with different letters are significantly different (< 0·001).

Fruit yield (kg per vine)
Healthy plants14·4 a9·8 a13·8 a
Diseased plants4·4 b4·5 b8·7 b
P <0·001<0·001<0·001
LSD1·61·11·7
Bunches per vine
Healthy plants168 a136 a144 a
Diseased plants64 b88 b96 b
P <0·001<0·001<0·001
LSD191317
Pruning weight (kg)
Healthy plants 1·3 a 
Diseased plants 0·8 b 
P  <0·001 
LSD 0·20 
Average shoot length (cm)

Early spring

2009–10

Early summer

2009–10

Harvest

2009–10

Healthy plants495 a729 a961 a
Diseased plants272 b407 b526 b
P <0·001<0·001<0·001
LSD5794148

Discussion

This study identified, by isolation of pathogens from diseased grapevines in the Riverina and fulfilment of Koch's postulates, that co-infection by two different types of fungal pathogens, the root-invading I. macrodidyma and the wound-invading Botryosphaeriaceae – most commonly D. seriata but occasionally D. mutila – can cause a particularly severe form of decline of young grapevines. In vineyards, even without plant death, co-infected plants had greatly diminished growth and fruit yield; in laboratory experiments, the root health and vegetative growth of co-infected plants were more greatly diminished than infection by either pathogen alone.

Although disease of mature grapevines and nursery stock caused by each of the pathogens individually is well known, this study determined that most, if not all, co-infection of young grafted grapevines took place within the propagation system. That determination is based on the following evidence. First, spatial analysis of affected plants, within a short time from planting indicated that causal pathogens were unlikely to have spread substantially from any particular source(s) within the vineyard. Secondly, the co-infected plants had advanced internal stem discoloration of the oldest xylem tissue below the graft union within two months of planting. Thirdly, both Ilyonectria spp. and Botryosphaeriaceae spp. were isolated from grapevines with symptoms in the vineyard and from the supplier nursery, and Botryosphaeriaceae spp. alone were isolated from the rootstock source block.

There have been indications that Botryosphaeriaceae and Ilyonectria species might occur together in young grapevines in nurseries in Portugal (Oliveira et al., 2004). However, to the authors’ knowledge, this is the first examination of the interactive effects of Botryosphaeriaceae and Ilyonectria spp. co-infections and the infection sources during propagation of grafted grapevines.

The present study indicates that the initial Botryosphaeriaceae infection occurred in the rootstock source plant canes during the nursery propagation of grafted grapevines, and that the Ilyonectria root infection occurred in nursery soil in which the grafted plants were grown for a season prior to distribution. The extensive infection of roots, rootstock trunk and graft union of young declining grapevines by D. seriata and relatively low infection of the scion tissue, together with the presence of D. seriata in the young nursery plants and rootstock canes from the source block, indicate that the fungus was present prior to grafting and subsequently grew from the rootstock through the graft union and ultimately into the scion.

Grapevine rootstock source plants are particularly susceptible to Botryosphaeriaceae spp. infection because, unlike grapevine source plants used for ungrafted plants and scions for rootstocks, all the shoots of rootstock source plants are usually pruned at the base, thus directly exposing the perennially retained trunk to wound infection. Botryosphaeriaceae species are well known invaders of grapevine wounds in wet weather when spores are released from pycnidia on diseased wood and pruning debris. It is likely that infection of rootstock plants originates this way and that canes become infected seasonally when, after budburst, hyphae grow within the xylem of the current season's shoots. The isolation of D. seriata from canes of symptomless rootstock grapevines accords with findings of Botryosphaeriaceae fungi in basal internodes of 1-year-old rootstock canes in South Africa (Fourie & Halleen, 2004) and in the basal and central parts of rootstock canes in New Zealand (Billones et al., 2010). Those findings also indicate acropetal infection from the trunk of the plant. Amponsah et al. (2012) reported that hyphae in xylem of grapevine stems inoculated with the Botryosphaeriaceae fungus Neofusicoccum luteum moved faster acropetally than basipetally. The susceptibility of rootstock source plants to infection by Botrysphaeriaceae is likely to be high where, as is common, the trunk is kept short and the shoots are allowed to sprawl on the ground, thus promoting infection from soilborne inoculum sources (Gramaje & Armengol, 2011).

During grapevine propagation, one of the earliest stages of disease spread from infected canes to non-infected canes is during the post-collection hydration/storage process (Waite & Morton, 2007) and it appears likely that the Botryosphaeriaceae spread during that stage. As spore release from Botryosphaeriaceae pycnidia is triggered by water (van Niekerk et al., 2010), hydration of Botryosphaeriaceae-contaminated canes or storage in moist conditions is likely to release conidia which then, in the absence of appropriate fungicides, infect the cut ends of other canes. Later, Botryosphaeriaceae fungi may enter the many other plant wounds created in disbudding the rootstock cuttings and through incompletely healed graft unions. Basipetal hyphal growth from infected rootstock canes seems to account for the unexpected occurrence of Botryosphaeriaceae (not normally considered root pathogens) in the roots of the young diseased Riverina grapevines. Although Botryosphaeriaceae fungi have been found in grapevine rootstock source plants elsewhere, this is the first study to show an Australian grapevine rootstock nursery source block as a source of those fungi.

The high percentage of isolations of pathogenic Ilyonectria spp. from the roots, rootstock trunk and grafting unions of diseased plants in the Riverina vineyards, young nursery plants and in soil from a supplier nursery, despite its absence from rootstock source plants, strongly indicate that the nursery soils harboured pathogenic Ilyonectria spp. and that the Ilyonectria fungi entered the propagation process when young plants became infected from soilborne sources in the field nursery.

Ilyonectria spp. are the most common pathogens associated with young grapevines in Californian field nurseries (Dubrovsky & Fabritius, 2007) and they have also been isolated from young nursery grapevines in South Africa (Halleen et al., 2003), Uruguay (Abreo et al., 2010), Spain (Giménez-Jaime et al., 2006), New Zealand (Bleach et al., 2007) and Portugal (Rego et al., 2009). South African studies have also shown that Ilyonectria infections in nursery stock establish predominantly after callused cuttings are planted (Halleen et al., 2003, 2006). At that stage, undifferentiated callus tissue and especially uncovered pith at the base of the cutting are vulnerable to infection. In addition, the callus and root initials are fragile and can break during planting, resulting in infection-susceptible wounds (Halleen et al., 2003).

The Ilyonectria group (e.g. I. macrodidyma, I. liriodendri) are known root pathogens (Cabral et al., 2012a). Although their grapevine pathology is not well known, Ilyonectria spp. have been isolated from roots and basal ends of grafted cuttings in young vineyards and nurseries in many viticultural regions worldwide (Petit et al., 2011). The inability to isolate I. macrodidyma from canes of Ramsey rootstock plants accords with reports by Halleen et al. (2003) from South Africa that Ilyonectria spp. were rarely isolated from canes of rootstock grapevines. It is known that BFA produced by I. macrodidyma is toxic to grapevines (Sweetingham, 1983). The confirmation of BFA production by both Ilyonectria species isolated from grapevines with symptoms in the Riverina indicates that the stunted root and shoots of those plants may be induced by this phytotoxin. As BFA suppresses the growth of endophytic fungal competitors within the plant, it may also have a role in substrate defence for Ilyonectria spp. (Weber et al., 2004).

Although both Botryosphaeriaceae and Ilyonectria fungi alone can cause the decline and death of young grapevines, this study has shown that in Australia's Riverina region single infections of grafted grapevines by either pathogen are uncommon and that co-infection leads to particularly severe disease symptoms. To the authors’ knowledge, this is the first examination of interactive effects of these fungi and the infection sources during propagation of grafted grapevines. In the Riverina, widespread decline of young grafted grapevines is attributable to initial infection of rootstock cuttings by Botryosphaeriaceae fungi and spread of that infection during propagation, followed by infection by Ilyonectria spp. in nursery soil. Thus the roots and the rootstock stem below the graft and the graft union become internally infected with both fungi. As a result, the Ilyonectria spp. disrupt root function and retard early plant development, while Botryosphaeriaceae fungi gradually invade the xylem, both basipetally and acropetally, and contribute to the decline and eventual death of the plant. The phytotoxin BFA produced by the Ilyonectria spp. probably facilitates that process. In the Riverina, the Petri disease fungi were less common than Ilyonectria and Botryosphaeriaceae spp. in young declining grafted grapevines, but they probably contribute to the decline of surviving plants as they mature.

Acknowledgements

This research was supported by Australia's grape growers and winemakers through their investment body, the Grape and Wine Research Development Corporation, with matching funds from the Australian government. The authors thank the many Riverina wine grape growers and the Wine Grapes Marketing Board, Griffith (particularly Kristy Bartrop and Jason Capello) for their cooperation in this research. Dr Jacqueline Edwards and Dr Ian Pascoe are thanked for their advice at the initiation of the project. Dr Tom Nordblom and Dr Ray Cowley (NSW Trade and Investment) are thanked for their helpful advice regarding the manuscript, and Professor Marlene Jaspers (Lincoln University, NZ), Dr Nicola Wunderlich and Leo Quirk (NWGIC) are acknowledged for helpful discussions during the course of this work. The authors thank Mark Wilson and Lindsay Greer (NWGIC) for their assistance in remote sensing of the Riverina vineyard and with brefeldin A determinations respectively. The authors express their gratitude to Dr John Harper (School of Agriculture and Wine Sciences, Charles Sturt University) for advice with scanning electron microscopy and to Dr Cheng Huang and Professor Martin Canny, Australian National University Electron Microscopy Unit, for their valuable guidance and advice with cryoscanning electron microscopy.

Ancillary