The Biology and Management of Chestnut Rot in Southeastern Australia

Abstract

Chestnut rot of Castanea sativa Mill. (European chestnut) and Castanea crenata Siebold and Zucc. (Japanese chestnut) x Castanea sativa hybrids is a significant problem facing the Australian chestnut industry. It affects the chestnut kernel, manifesting as pale, medium and dark brown lesions occurring on the endosperm and embryo. Previous surveys of Melbourne Markets showed losses to chestnut rot up to 40% (Anderson 1993)*. This equates to losses of $5.2M in 2010; using projected production figures (HAL 2007). This research project was undertaken to create a better understanding of the scope and distribution of the chestnut rot problem in south-eastern Australia; clarify the confusion surrounding the taxonomy of the chestnut rot pathogen; elucidate the infection process and disease cycle; investigate the effectiveness of flotation disease grading as a post-harvest method of removing rotten chestnuts; and provide recommendations to growers on how to reduce the incidence of chestnut rot in their orchards. Twenty-two orchards were surveyed in 2008 and 21 in 2009, across New South Wales (NSW) and Victoria (VIC) (Chapter 2). The highest incidence of chestnut rot at individual orchards was 72%. Incidence varied widely between and within orchards between the two years sampled. Chestnut rot was present in all of the sampled 2 orchards. The important commercial varieties Decoppi Marone (DM), Purton’s Pride (PP), Red Spanish (RS) all displayed examples of both high incidence (>1%), and acceptable incidence (0-1%). This indicates these varieties are susceptible under the right conditions. There was a positive correlation between incidence and December rainfall of the previous year, indicating environmental factors as key to the infection process. In 2008 and 2009, surveys of Sydney Markets showed incidence >1% (2008: varieties DM, PP; 2009: varieties RS, PP), indicating that these varieties were capable of being affected by chestnut rot. Chestnut rot has recently been reported as caused by two fungal species, minorly in New Zealand by Diaporthe castaneti Nitschke. and majorly in Australia and New Zealand by Gnomonia pascoe prov. nom. (Smith and Ogilvy 2008). The current study only observed one causal agent of chestnut rot in Australia, the novel taxon, Gnomoniopsis smithogilvyi sp. nov. Isolates of G. smithogilvyi were obtained from tissues including rotten chestnuts collected in surveys of NSW and VIC, as ascospores from dead burrs from NSW, and as endophytes from asymptomatic female and male flowers, leaves, and stems from NSW. Morphology and phylogenetics were used to elucidate the taxonomy of the fungus. Morphological examination of G. smithogilvyi included the teleomorph from burrs (perithecia, asci, ascospore characters), and the anamorph in culture (colony, conidiomata and conidia characters). The RNA polymerase II (rpb2), internal transcribed spacer regions 1 and 3 2 encompassing the 5.8S rDNA (ITS), translation elongation factor 1-alpha (tef1-α), and beta-tubulin (β-tubulin) gene loci were sequenced and analysed in the context of the Diaporthales Nannf., Gnomoniaceae Winter. and Gnomoniopsis Berl. All of the chestnut rot isolates, ascospore isolates, and endophyte isolates on Castanea sativa, and Castanea crenata x C. sativa hybrids in Australia (NSW and VIC) were identified as G. smithogilvyi. An ITS phylogeny analysing the G. smithogilvyi isolates from the current Australian study with isolates of Gnomoniopsis on C. sativa from India, C. sativa from Italy, and C. crenata, C. sativa, and Castanea sp. from New Zealand (Chapter 3) grouped the Australian isolates, the Indian isolates, 17 of the 19 Italian isolates, and 3 of the 4 New Zealand isolates in the same lineage with 100% maximum parsimony (MP) bootstrap support, and 1.0 Bayesian posterior probability (BP). This suggests all these isolates belong to the genus Gnomoniopsis, and are highly likely to be G. smithogilvyi. A multi-gene phylogeny needs to be completed with all of these isolates to unequivocally determine if they are G. smithogilvyi. One of the 4 New Zealand isolates grouped with Gnomoniopsis paraclavulata in this analysis indicating that there is likely to be more than one species of Gnomoniopsis on Castanea spp. in New Zealand. Subsequent to the publication of G. smithogilvyi (Shuttleworth et al. 2012a), Gnomonia pascoe prov. nom. and a recently published taxon reported as the casual agent of nut rot of Castanea sativa in Italy, Gnomoniopsis castanea were all found to be synonyms of G. smithogilvyi based on 4 morphology and a two gene phylogeny (ITS, tef1-α) (Chapter 3). Chapters 3, 4 and 5 isolated the G. smithogilvyi in its anamorph form from rotten chestnuts, in its teleomorph form as a saprobe on dead burrs, and as an endophyte isolated from asymptomatic floral and vegetative chestnut tissues. Historically, there has been significant movement of chestnuts and budwood from Europe to Australia. It is therefore possible that the G. smithogilvyi was imported to Australia from Europe. The fungus could also potentially have been introduced from Japan, China, or the USA as Castanea from these countries have all been transported to Australia. There is also a possibility that the fungus has an endemic Australian origin. Further work with native plant species needs to be completed to determine if this is the case. The fungus could also have been transported between orchards in Australia and New Zealand by exchange of chestnuts and budwood between the two countries. G. smithogilvyi was isolated as an endophyte from various vegetative and floral tissues of Castanea in December 2008, and February, April, August, and December 2009 from an orchard in Mullion Creek, NSW (Chapter 4). The ranking of highest to lowest isolation frequency in chestnut tissues was female flowers (December 2008), mature burr equators, mature pedicels, living male flowers, dead male flowers, terminal leaf margins (April 2009), dead styles, dormant terminal buds, immature burr equators, pedicels (February 2009), leaf mid-veins, current-year stems (August 2009, 5 February 2009), and mature shell equators (April 2009). All other tissue types had ≤20% isolation frequency including current-year stems (December 2008, April 2009), 2 year-old stems, petioles, mature kernels, female flowers (December 2009), immature shell equators, living male flowers (December 2009) and 3 and 4 year-old bark. The endophyte was not isolated from 3 and 4 year-old xylem. There was a decreasing trend of isolation with increasing age of chestnut tissues in four of the five months. There was also a 72% reduction in isolation frequency from female flowers between 2008 (82%) and 2009 (10%), indicating a dynamic distribution of the fungus in chestnut flowers that changes over time. It also suggests a seasonal infection of female chestnut flowers. All tested varieties (DM, PP, RS) had the chestnut rot endophyte isolated from their tissues, indicating that they have the potential to be affected by chestnut rot. The observation of chestnut rot perithecia on burrs is central to the hypothesis of a floral infection by ascospores. This study observed G. smithogilvyi on dead burrs and branches in Mullion Creek, NSW (Chapters 3, 4). This observation of perithecia and ascospores on burrs supports the hypothesis of a floral infection. Ascospore infection of chestnut flowers has previously been found to be the primary stage of infection leading to chestnut rot. In this study ascospores were captured on PDA plates in a closed chamber laboratory experiment with chestnut burrs containing overwintered perithecia and ascospores of the G. smithogilvyi (Chapter 5). Three 6 colonies of the G. smithogilvyi anamorph grew in the second week of incubation. The incubation temperature was stable for the duration of the experiment at 23oC, suggesting fluctuations in temperature are not required for ascospore release, with moisture and humidity likely to be more important. A isolate that was grown from an ascospore was identified using morphological and molecular techniques. A segment of the ITS region of rDNA was sequenced and analysed. The captured ascospore isolate was morphologically identical to G. smithogilvyi, and it grouped next to G. smithogilvyi in the maximum parsimony (MP) ITS phylogenetic tree indicating the isolate is G. smithogilvyi. This experiment indicates that ascospores are released from the dead burrs into the air where they can potentially infect chestnut flowers, again supporting the floral infection hypothesis. Ascospores were found to be the primary source of inoculum in the infection of chestnut flowers, leaves and stems in December, leading to chestnut rot symptoms the following year. Chestnut rot ascospores were captured using a Burkard Volumetric Spore-Trap in an orchard in Mullion Creek, NSW (Chapter 5). The instrument was also used to determine daily patterns in ascospore capture from the orchard atmosphere. The highest mean hourly frequency of ascospore capture was 165 ascospores per m3 of air at 10pm. The time period of peak ascospore capture was between 8-11 pm and between 7-9 am. These times of ascospore capture correspond to sunset and the hours following sunset, and the hours following sunrise. No rain fell during the sampling 7 period, indicating ascospores are released even in the absence of rain. Flotation disease grading is a post-harvest method used to separate rotten chestnuts from healthy ones. Chestnuts that float are considered rotten, those that sink considered healthy. An experiment was carried out to investigate the effectiveness of flotation disease grading as a post-harvest method of removing chestnuts affected by chestnut rot (Chapter 6). Hot water treatment of chestnuts has also been found to be effective against fungal growth on chestnut shells and therefore a desirable treatment method used in combination with flotation disease grading. The temperatures tested were 4oC, 30oC, 50oC, 60oC, and 70oC. Both floating and sinking chestnuts were affected by chestnut rot. The method was most discriminating with water at 70oC, although 22 out of 80 of the chestnuts that sank were rotten in this treatment. The method was observed to work well on chestnuts that are highly desiccated, but less effectively on chestnuts with minor chestnut rot symptoms. However, there are many more rotten non-desiccated chestnuts than desiccated ones. This is a problem because non-desiccated rotten chestnuts increase in chestnut rot with increasing time in storage, especially after 60 days (Anderson 1993). Flotation disease grading needs to be used with caution as the method can potentially reduce grower profits by identifying healthy chestnuts as rotten and mis-identifying rotten chestnuts as healthy. Potential losses from mis-identified chestnuts in this experiment was calculated as 160-260 kg of chestnuts per metric tonne (t), valued at $800-$1300 per t.