Pathogenicity of Marssonina betulae and other fungi on birch

Authors


*E-mail: sarah.green@forestry.gsi.gov.uk

Abstract

Marssonina betulae, Discula betulina, Melanconium bicolor and Fusarium avenaceum were inoculated onto shoots of 1- and 2-year-old seedlings of Betula pendula and B. pubescens and symptom development monitored over several seasons. Marssonina betulae caused disease on B. pendula, but not on B. pubescens. On B. pendula symptoms included discrete lesions, which often girdled, causing dieback of inoculated leading shoots, and the development of secondary sunken cankers on the main stems, which were usually centred around a dead sideshoot. Cankers on the main stems expanded during subsequent growing and dormant seasons, and often coalesced, girdling stems and causing the death of some seedlings. All isolates of M. betulae caused disease on B. pendula and conidia were able to infect young shoots in early flush without requiring a wound. Discula betulina caused lesions and dieback on B. pendula and B. pubescens within 3 months of inoculation, but disease did not progress thereafter and all inoculated seedlings recovered. Melanconium bicolor and F. avenaceum caused very little disease on either birch species. This study showed that M. betulae is an aggressive pathogen on B. pendula, causing sunken stem cankers and progressive crown dieback, which are symptoms commonly observed on young, planted birch at field sites across Scotland.

Introduction

Birch (Betula spp.) is a major component of lowland and upland native woodland throughout Scotland. Betula pendula (silver birch) and B. pubescens (downy birch) are valued increasingly for landscape, biodiversity and amenity purposes, although there has also been some recent interest in the potential of B. pendula as a timber species in the UK (Malcolm & Worrell, 2001). Betula pendula and B. pubescens are two of the more important broadleaved species in native woodland afforestation schemes, which have been ongoing in Scotland since the late 1980s. Since the late 1990s, widespread dieback of young, planted birch has been reported, with more than 20 native woodland planting schemes in Scotland, varying in size from 10 to 436 ha, known to be affected to date (Green, 2004). Affected trees grow well initially but, approximately 5–10 years after planting, begin to develop shoot dieback from the lower crown upwards and from the outer crown inwards. Specific symptoms displayed by such trees include multiple, sunken cankers on the stem and branches, often associated with a dead sideshoot (Fig. 1a), and discrete lesions and tip dieback on young shoots.

Figure 1.

Sunken cankers on the branches of Betula pendula at a woodland grant scheme planting in Scotland (a), secondary cankers on the main stems of B. pendula 21 months after inoculation with Marssonina betulae showing multiple main stem cankers (b), and canker centred around a dead sideshoot (c).

A survey of five affected planting schemes was conducted in Scotland in 2002 as the first stage in a research programme to determine the potential roles of fungal pathogens in causing birch dieback (Green, 2004). A broad range of fungi were isolated from birch shoots with and without symptoms, and the most frequently isolated fungi were inoculated onto birch seedlings in pathogenicity tests (Green, 2004). Subsequent observations of disease over a single growing season indicated varying degrees of pathogenicity in Marssonina betulae, Discula betulina, Melanconium bicolor and Fusarium avenaceum (Green, 2004). Both M. betulae and D. betulina are common foliar pathogens on birch, causing characteristic leaf spots (Peace, 1962; Bennell & Millar, 1984), although neither fungus has been considered previously to be a causal agent of shoot dieback. In pathogenicity tests, both M. betulae and D. betulina caused lesions on birch shoots which were wounded and inoculated, and M. betulae also caused shoot dieback without requiring prior wounding, which indicated that this fungus is a more aggressive pathogen than suggested in the current literature (Green, 2004). Melanconium bicolor is known as a frequent colonizer of damaged or declining birch shoots (Peace, 1962; Bennell & Millar, 1984) and tests have shown that this fungus may act as a weak or opportunistic pathogen on shoots or stems predisposed by wounding (Green, 2004). Fusarium avenaceum causes stem lesions on birch seedlings (Lilja et al., 1996, 1997; Green, 2004) and has been isolated frequently from diseased buds and shoot tips of birch (S. Green, unpublished data).

To date, pathogenicity tests carried out with these fungi on birch have been limited, employing a single isolate of each fungal species inoculated onto young, actively growing shoots and with disease monitored over a single growing season only (Green, 2004). It is not known whether any of these fungi can cause disease to develop on birch over subsequent growing seasons, producing disease symptoms commonly seen at affected field sites, nor if disease severity is increased by inoculation during host dormancy. These questions were addressed in the present study.

Materials and methods

Inoculum production

Single-conidial isolates of M. betulae, D. betulina, M. bicolor and F. avenaceum were collected from leaves or shoots of B. pendula and B. pubescens from various locations in Scotland during 2001–04 (Table 1). Isolate stock cultures were maintained as conidia and mycelium frozen in 5% skimmed milk and 20% glycerol at −80°C. Isolates were grown on 2% malt agar (MA, Oxoid) and incubated at 20/15°C day/night temperatures with a 12-h photoperiod consisting of cool white fluorescent and near-ultraviolet (nuv) light. Colonies were subcultured no more than three times after isolation from the host before tests. To obtain conidial inoculum, sporulating colonies (grown on 2% MA in 9-cm-diameter Petri dishes) were flooded with 10 mL of sterile distilled water and the colony surface was rubbed gently with a sterile bent glass rod to dislodge conidia. The conidial suspension was filtered through two layers of cheesecloth and adjusted to approximately 106 conidia mL−l determined using a haemocytometer. For mycelial inoculum, 3-mm- (2003 experiments) or 5-mm- (November 2004 experiment) diameter mycelial plugs were cut using a sterile cork borer from the growing margin of colonies which were still expanding. Excess agar was sliced off the base of each plug before inoculation.

Table 1.  Origin of isolates of Discula betulina, Fusarium avenaceum, Marssonina betulae and Melanconium bicolor used for inoculation studies
IsolateSpeciesLocation of origin in ScotlandHost materialDate of isolation
2005D. betulinaPentland Hills, MidlothianB. pendula shootJuly 2001
2013D. betulinaCornharrow, Dumfries-shireB. pubescens shootAugust 2001
2041D. betulinaCaplawhead, ClackmannanshireB. pendula leafJune 2003
2044D. betulinaBlair Atholl, PerthshireB. pendula leafJune 2003
2030F. avenaceumGladsmuir, East LothianB. pubescens shootMarch 2002
2033F. avenaceumStrontian, Inverness-shireB. pendula shootOctober 2002
2014M. betulaeCornharrow, Dumfries-shireB. pubescens shootAugust 2001
2020M. betulaeGlen Artney, PerthshireB. pendula shootSeptember 2001
2039M. betulaeBlair Atholl, PerthshireB. pendula leafJune 2003
2042M. betulaeGlen Rinnes, BanffshireB. pubescens shootJuly 2003
2045M. betulaeGlentress, Peebles-shireB. pendula shootJuly 2003
2054M. betulaeLoch Glascarnoch, Ross & CromartyB. pubescens leafAugust 2004
2060M. betulaeUlzieside, Dumfries-shireB. pendula leafSeptember 2004
2003M. bicolorStrontian, Inverness-shireB. pendula shootJuly 2001
2016M. bicolorGrantown on Spey, Inverness-shireB. pendula shootAugust 2001
2018M. bicolorGlen Artney, PerthshireB. pendula shootSeptember 2001

To check the viability of conidial inoculum and sterility of control treatments, conidial suspensions and sterile distilled water used to inoculate control seedlings were brushed onto 2% MA in two Petri dishes which were placed in the dark at room temperature. After 24 h, percentage conidial germination was assessed for 200 conidia per dish. To check the viability of mycelial inoculum and the sterility of the control treatment, eight mycelial plugs of each isolate and four plugs of sterile 2% MA were assessed for growth on 2% MA after 1–2 weeks.

Plant preparation and inoculation

2003 Experiments

Scottish-provenance seedlings of B. pendula (Bonskeid provenance) and B. pubescens (Huntly provenance) were used in these experiments. In early March 2003, the 1-year-old seedlings, which were cell-grown and had been maintained outside in an area away from diseased trees, were placed in a darkened Conviron growth chamber at 10°C to inhibit flushing. Starting from the end of April 2003, batches of seedlings were removed from the growth chamber at 3, 6 or 9 weeks before inoculation, transplanted into 1·5-L plastic growth bags containing a mix of 70% peat, 15% bark and 15% perlite with 70 g P and 120 g K m−2 at pH 5, and placed outside to stimulate flushing.

Seedlings were inoculated with M. betulae isolate 2020 on 24 June, D. betulina isolate 2013 on 1 July, M. bicolor isolate 2018 on 8 July, and F. avenaceum isolate 2030 on 15 July (Table 1). Experiments were designed to test the pathogenicity of each fungal species with the following main factors: (i) birch species (B. pendula or B. pubescens), (ii) inoculum type (conidia or mycelium), (iii) wounding or non-wounding of the inoculation site, and (iv) age of leading shoot (3, 6 or 9 weeks post-flushing). Consequently, each fungal species had 24 treatment combinations, with six replicate inoculated seedlings and two control seedlings per treatment combination.

Seedlings were inoculated on the leading shoot, 2–3 cm above the base of the current year's shoot extension, after treatment with 50% v/v ethanol to surface-sterilize the shoot area before inoculation. The wounding treatment was made by scraping back a 10 × 2-mm strip of epidermis with a sterile scalpel, and the inoculum was then placed directly into the wound. For the non-wounding treatment, inoculum was placed directly onto the intact epidermis.

Conidial suspensions were applied using a paintbrush which was surface-sterilized by dipping in 50% v/v ethanol, then dipped once into the well-shaken suspension and applied to the inoculation site with three short brush strokes. Sterile distilled water was applied to control seedlings in the same manner. The paintbrush was dipped in 50% ethanol and rinsed in sterile distilled water after inoculating each seedling. Inoculated seedlings were placed immediately into a growth chamber at 20/15°C day/night temperatures with a 16-h photoperiod and watered daily directly into the potting mix. The growth chamber contained a misting system providing fine misting for 60 s every 5 min during the daytime and 30 s every 10 min during the night. This was sufficient to ensure that seedlings had continuous foliage wetness without runoff. After 6 days, inoculated seedlings were placed outside for the remainder of the experimental period with regular supplementary watering carried out during dry periods in the growing season.

For mycelial inoculation, plugs were placed mycelial side down onto the inoculation site. Control seedlings were inoculated with sterile plugs of 2% MA. A droplet of sterile distilled water was placed onto each plug, and the plug sealed in place with parafilm. Inoculated seedlings were placed immediately into a growth chamber with conditions as described for the conidial inoculation treatment, except that the growth chamber did not contain a misting system. After 6 days, the seedlings were also placed outside with supplementary watering as above. The parafilm was removed from seedlings 2 weeks after inoculation.

July 2004 experiment

This experiment was set up to further investigate the pathogenicity of M. betulae on B. pendula using five isolates of the fungus. Scottish-provenance seedlings of B. pendula (Wigtownshire provenance) were potted up into 1·5-L growth bags containing the peat/bark/perlite mix described above and maintained outside. The seedlings were in their second growing season and had 8–19 nodes’ extension on the leading shoots when inoculated on 21 July.

Each seedling was inoculated with a conidial suspension of one of five isolates of M. betulae onto the leading shoot, which was wounded or non-wounded before inoculation. Wounding and conidial inoculations were carried out as described for the 2003 experiments. Because of differences in sporulation among the five isolates tested, the conidial concentration of the suspensions used for inoculation varied as follows (conidia mL−1): isolate 2014 (9·7 × 104), isolate 2020 (1·4 × 106), isolate 2039 (1·2 × 105), isolate 2042 (7·2 × 105) and isolate 2045 (1·98 × 106) (Table 1). Eight replicate seedlings were inoculated with each isolate and eight seedlings inoculated with sterile water as controls. Inoculated and control seedlings were placed in a growth chamber containing a misting system, as described for the 2003 experiments. After 6 days, seedlings were transferred outside for the remainder of the experiment with supplementary watering as described for the 2003 experiments.

November 2004 experiment

This experiment was conducted to assess whether the host's state of dormancy influenced the infection success and rate of development of fungal isolates on B. pendula. Scottish-provenance seedlings of B. pendula (Wigtownshire provenance) were potted up into 1·5-L growth bags containing the peat/bark/perlite mix described for the 2003 experiments and maintained outside.

Two-year-old seedlings were inoculated on 15 November 2004 with four isolates each of M. betulae (2020, 2045, 2054 and 2060) and D. betulina (2005, 2013, 2041 and 2044), and two isolates of F. avenaceum (2030 and 2033) (Table 1). On 22 November 2004, another batch of seedlings was inoculated with three isolates of M. bicolor (2003, 2016 and 2018) (Table 1). Seedlings were wounded and inoculated with mycelial plugs at a point located 20 cm below the tip of the leading shoot using the inoculation technique described for the 2003 experiments. Six replicate seedlings were inoculated with each isolate and six seedlings inoculated with sterile 2% MA plugs as controls. Inoculated and control seedlings were placed outside and the parafilm removed after 2 weeks. Seedlings were maintained outside for the remainder of the experiment, with supplementary watering as described for the 2003 experiments.

Experimental design, disease assessments and analysis

All experiments were laid out as randomized complete-blocks, with inoculated seedlings blocked by replicate number according to spatial arrangement at all stages of each experiment.

2003 experiments

Lesions which developed at the inoculation sites were examined under a dissecting microscope for the presence of fruiting structures in August 2003. Re-isolations were attempted from six to eight randomly selected diseased seedlings and from six randomly selected control seedlings per fungal species in September or October 2003. For re-isolations, 20- to 25-mm-long shoot sections containing a lesion or healthy tissue (controls) were surface-sterilized in 70% v/v ethanol for 1 min, 1·25% sodium hypochlorite for 5 mins and 70% v/v ethanol for 30 s. They were then rinsed twice in sterile distilled water and blotted dry on sterile filter paper. Pieces of tissue approximately 2–5 mm long were removed from either a lesion margin or from healthy tissue (controls), plated onto 2% MA in 9-cm-diameter Petri dishes and incubated in the dark at room temperature. Developing colonies were examined after 3–4 weeks and identified by morphological characteristics. Sampled seedlings were removed from the experiment.

For each fungal species, the numbers of seedlings having no disease (scored as 0), a discrete, non-girdling lesion at the inoculation site (scored as 1), or a girdling lesion causing dieback of the leading shoot (scored as 2) were recorded 12 weeks after inoculation (September/October 2003) and 12 months after inoculation (June/July 2004). Data recorded 12 months after inoculation were analysed using an ordinal logistic regression (logistic procedure, sas software version 9·1) to determine the significance of the main factors and their interactions. All seedlings were retained for longer-term monitoring. Seedlings of B. pendula inoculated with M. betulae continued to develop symptoms and the following five parameters were recorded in June and September 2004, April and September 2005 and April 2006: (i) length (mm) of the lesion at the inoculation site, (ii) dieback of the leading shoot, (iii) distance (mm) from the lowest point of dieback on the leading shoot to the base of the main stem, (iv) length (mm) of secondary cankers on the main stem and their location (distance in mm from the lowest point of the lesion to the base of main stem), and (v) seedling death. Lesion and canker lengths and distance to the base of main stems were measured in order to determine when disease development occurred and to locate and keep track of secondary stem cankers. In April 2006, re-isolations were carried out as described above from inoculation lesions and secondary cankers on the main stems of four randomly selected seedlings showing these disease symptoms.

July 2004 experiment

Disease assessments were carried out 12 weeks after inoculation (October 2004), in January, April and October 2005 and in April 2006, by recording the same five parameters described for B. pendula inoculated with M. betulae in 2003. In April 2006, re-isolations were carried out from inoculation lesions and secondary cankers on the main stems of four inoculated seedlings as described for the 2003 experiments.

November 2004 experiment

Disease assessments were carried out 12 weeks after inoculation (February 2005), in April and October 2005 and in April 2006, by recording the same five disease parameters as described above. For seedlings inoculated with M. betulae, the presence of characteristic M. betulae lesions on leaves and 2005 sideshoots was also noted in October 2005. In April 2006, re-isolations were carried out from secondary cankers on the main stem of one inoculated seedling as described above.

Results

2003 Experiments

Mean conidial germination on test plates was 65% for D. betulina and between 94 and 100% for M. betulae, M. bicolor and F. avenaceum. All mycelial test plugs developed into healthy looking colonies, and all control plugs were sterile. No control seedlings developed disease and so were excluded from the analyses.

Assessments of seedlings [n (sample size) = 72 for each birch species] in August 2003 for fungal fruiting structures on inoculation lesions found M. betulae on 13 seedlings of B. pendula and none on B. pubescens; D. betulina on 26 seedlings (14 B. pendula and 12 B. pubescens); M. bicolor on three seedlings (two B. pendula and one B. pubescens); and F. avenaceum on two seedlings (one of each birch species). In August 2003, conidiomata of M. betulae were also observed on leaf lesions on 10 inoculated B. pendula seedlings. None of the control seedlings developed leaf symptoms. For isolations carried out in September/October 2003, M. betulae was not re-isolated from any of the eight sampled seedlings, D. betulina was re-isolated from seven out of eight sampled seedlings (four B. pendula and three B. pubescens), M. bicolor was re-isolated from five out of six sampled seedlings (two B. pendula and three B. pubescens) and F. avenaceum was re-isolated from three out of eight sampled seedlings (two B. pendula and one B. pubescens). None of these fungi grew from the re-isolated controls.

Marssonina betulae

Disease did not develop on B. pubescens, so this birch species was excluded from the analysis. Many seedlings of B. pendula inoculated with M. betulae developed disease, which continued to progress throughout the monitoring period from September 2003 until April 2006. Lesions developed on seedlings at the inoculation site and also on the main stem below the original inoculation site, often centred at the base of a dead sideshoot (Fig. 1b,c). For descriptive purposes, discrete lesions at the inoculation site are referred to as ‘lesions’, whereas lesions developing on the main stem below the original inoculation site are termed ‘secondary cankers’. Lesions and secondary cankers developed as dark, sunken areas of dead tissue (Fig. 1b,c), which expanded longitudinally and transversely during subsequent dormant and growing seasons, sometimes coalescing if multiple cankers occurred. Lesions and secondary cankers often girdled stems, and in some cases caused extended dieback of the leading shoot resulting in the death of nine seedlings. The length of discrete lesions caused by M. betulae on B. pendula in September 2004 ranged from 5 to 49 mm (mean 18·5 mm) and lesion expansion since June 2004 ranged from 4 to 15 mm (mean 8·6 mm).

Three months after inoculation with M. betulae, 44 B. pendula seedlings showed a discrete lesion at the inoculation site, three seedlings had dieback of the leading shoot (Fig. 2a) and 25 seedlings were symptomless. By June 2004 (12 months after inoculation), 13 more seedlings exhibited dieback of the leading shoot, 22 seedlings had developed one or more secondary cankers on the main stem and one seedling had died (Fig. 2a). Analysis of the number of seedlings without disease, a lesion at the inoculation site or dieback of the leading shoot 12 months after inoculation found that mycelium caused greater incidence of disease than conidia (P < 0·01) (Table 2), but there was no significant effect of wounding (P = 0·11) (Table 2). There was a significant effect, however, of flushing stage on disease (P < 0·05) with seedlings inoculated at the 3-week flushing stage having a greater incidence of dieback than seedlings inoculated at the 6- or 9-week flushing stages (Table 2). There were no significant interactions among the main effects. During the experimental period (June 2003 to April 2006), a total of 52 B. pendula seedlings inoculated with M. betulae developed disease and 27 of these diseased seedlings developed one or more secondary cankers on the main stem. The majority of these seedlings developed cankers within 2 years of inoculation (Fig. 2a) and all seedlings which developed secondary cankers had either a lesion at the inoculation site or dieback of the leading shoot when assessed in June 2004. All of the nine seedlings which died during the course of the experiment had developed multiple secondary main-stem cankers. Table 3 shows the number of seedlings within each treatment combination which developed secondary cankers. Of the 27 seedlings that developed secondary cankers, 16 had been inoculated with conidia, with nine non-wounded and seven wounded seedlings within this treatment (Table 3). Also, 16 of the 27 seedlings had been inoculated via 3-week-old shoots, whereas seven and four of these seedlings had 6- and 9-week-old shoots, inoculated respectively (Table 3).

Figure 2.

Cumulative number of Betula pendula seedlings inoculated with Marssonina betulae with a lesion at the inoculation site, secondary cankers, dieback of the leading shoot and which died up to (a) 29 months after inoculation in June 2003 (n = 72); (b) 21 months after inoculation in July 2004 (n = 80); and (c) 17 months after inoculation in November 2004 (n = 24).

Table 2.  Main effects of birch species, inoculum type, wounding treatment and shoot age on the number of seedlings of Betula pendula and B. pubescens with no disease, a lesion at the inoculation site, or dieback of the leading shoot in June 2004, 12 months after inoculation with Marssonina betulae, Discula betulina or Melanconium bicolor
 Number of seedlings at each disease level
Birch speciesInoculum typeWounding treatmentShoot age (weeks)
B. pendulaB. pubescensMyceliumConidiaWoundingNon-wounding369
  • a

    Data presented are for B. pendula only and a total of 65 seedlings were included in the analysis.

  • b

    A total of 142 seedlings were included in the analysis.

  • c

    Data are for mycelial inoculations only and a total of 68 seedlings were included in the analysis.

Marssonina betulaea
No disease   513 612 4 8 6
Lesion at inoculation site  16131316 8 912
Dieback  13 511 710 6 2
Discula betulinab
No disease264410602644222622
Lesion at inoculation site 9 711 5 511 0 610
Dieback362049 74115261614
Melanconium bicolorc
No disease2625  1734221415
Lesion at inoculation site 6 7  11 2 1 7 5
Dieback 2 2   4 0 1 2 1
Table 3.  Number of Betula pendula seedlings with secondary cankers following inoculation either with mycelium or conidia of Marssonina betulae onto wounded or non-wounded 3-, 6- or 9-week-old shoots in June 2003 (n = 72)
Shoot age (weeks)Number of seedlings with secondary cankers
MyceliumConidiaTotal
WoundedNon-woundedWoundedNon-wounded
3543416
61033 7
90112 4
Total657927

In June and September 2004 and April 2006, all lesions and cankers were examined for fruiting structures. In June 2004, M. betulae was identified in lesions of four seedlings, and on secondary cankers of three seedlings. In September 2004, M. betulae was only found fruiting on a lesion on one 2004 sideshoot. In April 2006, M. betulae was re-isolated from the original inoculation lesion on two of the four sampled seedlings.

Discula betulina

Birch species, inoculum type and wounding treatment all had highly significant effects on disease in June 2004, 12 months after inoculation with D. betulina (P < 0·0001 for all three main effects). Betula pendula was more susceptible, with a greater incidence of dieback than B. pubescens (Table 2). Mycelial inoculations caused a greater incidence of lesions and dieback on seedlings than conidial inoculations, and wounding before inoculation resulted in more seedlings with dieback than non-wounding (Table 2). Shoot age also had a significant effect (P < 0·05), with the greatest incidence of dieback occurring on seedlings on which 3-week-old leading shoots had been inoculated (Table 2). There were no significant interactions among the main effects. All infected seedlings subsequently recovered, with lesion sites subsequently occluded by callus growth and dead leaders being replaced by a newly dominant upper sideshoot.

Melanconium bicolor

Conidial inoculation with M. bicolor did not cause disease so this treatment was removed from the analysis. Mycelial inoculations caused a low incidence of disease, with discrete lesions at the inoculation site and dieback of the leading shoot developing on 13 and four seedlings, respectively, out of a total of 72 inoculated seedlings. There was no effect of birch species, with similar levels of disease on B. pendula and B. pubescens (Table 2). Wounding before mycelial inoculation resulted in a greater (P < 0·001) incidence of lesions and dieback on seedlings than non-wounding, and seedlings inoculated via 3-week-old shoots had less disease (P = 0·055) than those inoculated via 6- and 9-week-old shoots (Table 2). There were no significant interactions among the main effects. All seedlings subsequently recovered from infection.

Fusarium avenaceum

No disease developed on seedlings inoculated with conidia. Wounding and inoculation with mycelium caused discrete lesions at the inoculation site on only three seedlings (two B. pubescens and one B. pendula) and dieback of one seedling (B. pendula). All affected seedlings had been inoculated via 6- or 9-week-old shoots. All seedlings subsequently recovered from infection.

July 2004 experiment

Mean conidial germination on test plates ranged from 85 to 99% for all M. betulae isolates. No control seedlings developed disease. Disease on inoculated seedlings progressed similarly to that reported for B. pendula inoculated with M. betulae in 2003. At 3 months after inoculation (October 2004), 24 seedlings had developed lesions at the inoculation site (Fig. 2b). Between October 2004 and April 2005 (3–9 months after inoculation), 23 seedlings developed secondary cankers on the main stems (Fig. 2b) and 34 seedlings had inoculation lesions or secondary cankers which had expanded. Secondary cankers only developed on seedlings which had a lesion at the inoculation site. No new seedlings developed secondary cankers after April 2005, but between April 2005 and October 2005 (9–15 months after inoculation), leading shoots died back on eight seedlings (Fig. 2b) and 17 seedlings had extension of dieback and/or canker expansion. Only four seedlings had further increases in disease between October 2005 and April 2006. Of the 34 seedlings which developed disease during the course of the experiment, 27 had been wounded and seven non-wounded before inoculation with the conidial suspensions. Although the aggressiveness of isolates could not be compared directly because of differences in inoculum concentrations, the number of seedlings which became diseased with each isolate ranged from two to 11 out of a total of 16. In April 2006, M. betulae was re-isolated from a secondary canker on the main stem of one of the four sampled seedlings.

November 2004 experiment

All isolates of M. betulae, D. betulina and F. avenaceum, and isolates 2003 and 2018 of M. bicolor, were 100% viable, as indicated by outgrowth of mycelium from all of the test plugs. For isolate 2016 of M. bicolor only two of the four test plugs developed into colonies of the isolate. All of the control test plugs were sterile. No control seedlings developed disease.

Seedlings inoculated with M. betulae developed symptoms similar to those reported for the 2003 and July 2004 experiments. Inoculation lesions developed on 11 seedlings by February 2005 (3 months after inoculation), on a further three seedlings by April 2005 (5 months after inoculation) and on four more seedlings by October 2005 (11 months after inoculation) (Fig. 2c). The leading shoots died back on 13 seedlings between April and October 2005 (5–11 months after inoculation) (Fig. 2c). In October 2005, leaf lesions characteristic of M. betulae were observed on 21 inoculated seedlings and 12 seedlings had diseased sideshoots. Secondary cankers developed on the main stems of eight seedlings between October 2005 and April 2006 (11–17 months after inoculation). By the end of the experimental period (November 2004 to April 2006), a total of 19 out of 24 inoculated seedlings had developed disease. The number of seedlings which became diseased with each isolate ranged from two to six out of a total of six. Marssonina betulae was re-isolated from a secondary canker on the main stem of one seedling in April 2006. Seedlings inoculated with D. betulina, M. bicolor or F. avenaceum only developed small lesions at the inoculation site which did not expand and which were occluded by subsequent growth.

Discussion

This study demonstrated that M. betulae can be an aggressive pathogen when inoculated onto seedlings of B. pendula, causing sunken stem cankers and progressive shoot dieback, symptoms consistent with those commonly seen on young birch trees at planting sites throughout Scotland. These results are contrary to reports in the current literature which consider M. betulae to be a weak pathogen, occurring predominantly on leaves (Peace, 1962; Bennell & Millar, 1984; Bäucker & Eisenhauer, 2001). All isolates of M. betulae tested in this study were able to cause disease on B. pendula. Although the experiments were not designed to compare the aggressiveness of isolates, some differences among isolates were observed. In the experiments reported here, M. betulae did not cause disease on B. pubescens, although it has been observed on B. pubescens at several planted sites in Scotland (H. De Silva & S. Green, unpublished data). It is likely that B. pubescens is generally less susceptible to infection by M. betulae, but genetic variability within the birch species may influence disease expression. The other fungi examined in this study, D. betulina, M. bicolor and F. avenaceum, were not found to be serious pathogens on shoots of B. pendula or B. pubescens.

The pattern of disease caused by M. betulae on B. pendula was similar for all experiments, with discrete lesions developing at the inoculation site on leading shoots, followed by the development of one or more secondary cankers on the main stem below the inoculation lesion. The fact that M. betulae caused severe cankering and dieback on 2- to 3-year-old main stem tissues appears to be unusual for the genus. Marssonina spp. cause leaf spots and blights on a number of other tree hosts, including Populus, for which they are regarded as important forest diseases (Sinclair & Lyon 2005). In general, lesions caused by Marssonina spp. develop predominantly on leaves and very young shoots, leading to premature defoliation during disease epidemics (Phillips & Burdekin, 1982; Sinclair & Lyon, 2005). However, M. salicicola is also able to cause damaging cankers on 1- and 2-year-old shoots of Salix spp., which can enlarge and coalesce (Peace, 1962; Butin, 1995) and severe outbreaks of M. brunnea on Populus spp. can cause the death of buds and scattered young branches (Butin, 1995).

From the experiments described here it could not be determined exactly how M. betulae spread from the inoculation site to cause the secondary cankers, which usually occurred at a dead sideshoot, although in some cases the associated sideshoot was living. Marssonina betulae was able to form acervuli on inoculation lesions, with the potential to release conidia which are probably spread via rainsplash, the primary dispersal mechanism for these fungi (Sinclair & Lyon, 2005). Conidia could therefore be washed down the main stem during rainfall to form infection loci at the junction with sideshoots or buds, causing the sunken cankers and death of the associated bud or sideshoot. Conidia may have been washed down the main stem at the time of inoculation, although the likelihood of this occurring was reduced by ensuring that the seedlings were not misted to the point of runoff and by watering the seedlings directly into the potting mix. Typically, conidia of Marssonina spp. are able to cause repeated cycles of infections during wet summers (Sinclair & Lyon, 2005). Therefore, disease observed on leaves and young sideshoots during the growing season was most likely to be caused by conidia spreading from inoculation lesions during rainfall. The likelihood of these secondary infections being caused by natural inoculum of M. betulae is low since control seedlings did not become diseased. Young sideshoots infected with M. betulae had often died back to the main stem by the end of the growing season. This may have provided a conduit for the fungus to grow into the main stem at this point, thus acting as another potential route for the formation of secondary main-stem cankers.

Lesions and secondary cankers continued to expand throughout the course of each experiment, girdling stems to cause dieback and occasionally seedling death, with no evidence for increased disease development during dormancy compared with the growing season. Marssonina betulae was re-isolated only infrequently from inoculation-induced lesions and secondary cankers. The fungus grows very slowly in vitro and if this is the case in planta it may have been rapidly overgrown by the faster-growing fungi which tended to dominate isolation plates. However, M. betulae was isolated from expanded lesions and cankers on a few seedlings at the end of each experimental period. This supports the other experimental evidence presented here that M. betulae is the causal agent of these sunken cankers.

In this study, conidia of M. betulae caused disease on B. pendula when inoculated onto non-wounded shoots. The 2003 experiment indicated that young shoots in early flush were quite susceptible to infection by conidia of M. betulae, whereas in 2004 most seedlings inoculated with conidia in late July, after cessation of shoot extension, required wounding for disease to become established. Primary infections by Marssonina spp. tend to occur in spring shortly after the leaves emerge on the host, and are initiated by conidia from acervuli overwintering in lesions on shoots and fallen leaves (Sinclair & Lyon 2005). Study of the infection biology of M. betulae is needed to determine which tissues are most susceptible to conidial infection, and to define more accurately the stage of flushing at which young shoots are most likely to become infected in the spring.

The other fungal species studied here were less pathogenic than M. betulae on shoots of B. pendula and B. pubescens. Mycelial inoculations with D. betulina caused lesions and dieback on shoots of both birch species within 3 months of inoculation, which suggested initially that this fungus was more aggressive than M. betulae. However, all inoculated seedlings subsequently recovered from infection. Conidial inoculations with D. betulina caused very little disease, but since this fungus is a common endophyte in both diseased and healthy shoots of birch in Scotland (Green, 2004), infection may initially be symptomless, causing shoots to die back only when combined with other stress factors initiated by climatic damage or unsuitable site conditions. Discula betulina caused more disease on B. pendula than on B. pubescens in these experiments, which is at odds with the results of an earlier study (Green, 2004). However, this may be related to the use of different provenances in each study rather than to differences in susceptibility between the two species. Discula betulina did not cause progressive disease on birch in these experiments, and therefore appears unlikely to have contributed significantly to the cases of birch dieback encountered to date in Scotland. However, D. betulina can cause a damaging leaf disease on birch, and may contribute further to the problem of dieback as severe leaf infections can cause premature defoliation (Phillips & Burdekin, 1982).

Melanconium bicolor is another common endophyte of birch in Scotland and F. avenaceum has been frequently isolated from diseased birch shoots (Green, 2004). Neither fungus caused significant disease when inoculated onto birch in this study, which strongly suggests that these fungi are more likely to be secondary colonizers of diseased and dying shoots in the field.

The main finding of this study was that M. betulae is an aggressive, primary pathogen on shoots and stems of birch, causing cankers and dieback when inoculated onto B. pendula, thereby producing symptoms equivalent to those observed on young birch in affected planting schemes across Scotland. Other factors, such as unsuitable provenance and site selection, poor silvicultural management and climatic damage, may also cause birch dieback or predispose trees to disease. Further work is currently underway to assess the incidence and severity of M. betulae foliar and canker disease on planted birch at a range of sites across Scotland to determine whether severity of disease correlates with severity of dieback and to confirm the high degree of pathogenicity of this fungus.

Acknowledgements

The authors wish to thank Mrs Heather Steele for technical assistance during this study, Mr Andrew Peace for advice on statistics and Dr Steven Hendry for his critical review of the manuscript.

Ancillary