Fungi associated with necrotic galls of Dryocosmus kuriphilus (Hymenoptera: Cynipidae) in northern Spain

The Asian chestnut gall wasp (ACGW), Dryocosmus kuriphilus Yasumatsu (Hymenoptera: Cyn-ipidae) is one of the most important pests in Castanea species worldwide. In 2012, it was found for the first time in Catalonia (Spain) and a year later, in the north of Spain (Cantabria). Today, it is present in 14 Spanish provinces. In search of biological control against the ACGW, several authors have previously found the relationship between the presence of some Fusarium Link species in necrotic galls and wasp mortality due to the production of different types of wall-degrading enzymes and entomopathogenic mycotoxins. The objective of this study was to investigate the mycobiota associated with necrotic galls to find interesting perspectives for biological control of the ACGW. For this purpose, in 2014, 119 necrotic galls of Castanea sativa Miller were plated to isolate and identify the associated fungi. The fungal isolates were identified by the morphology of the fruiting bodies and DNA analyses. From necrotic galls, 7 species of fungi were identified. Of these, we highlight three species of Fusarium Link as well as the presence of Gnomoniopsis smithogilvyi Shuttlew, Liew & Guest due to its toxic capacity. Further studies are required to verify the effectiveness of these fungal species as biocontrol agents against the ACGW.

Dryocosmus kuriphilus is a gall-inducing insect, native from China, which was introduced for the first time in Japan, later in North America and more recently, in Europe (2002).After the first record in Italy, it was detected in other 14 neighboring countries.Currently, this pest is the most important on Castanea species worldwide with losses of nuts that reach 80% when serious infestations occur (EFSA 2010) and, in addition, the mortality of branches and trees in very severe infestations of young plantlets or weak plants (Cooper and Rieske 2007).
In Europe, it is also the most impactful alien pest for Castanea sativa (Sartor et al. 2015) and the level of damage depends on the cultivar (Botta et al. 2006;Panzavolta et al. 2012).The ACGW can reduce the yield in chestnuts by preventing the formation of the female flower when galls are formed in the apical buds of the shoots, stopping the growth and producing floral abortion.Moreover, leaf area, photosynthesis and tree biomass are also reduced (Kato and Hijii 1997;Battisti et al. 2013;Gehring et al. 2017).
Additionally, the gall wasp could also be related to the increase of the chestnut blight disease Chryphonectria parasitica (Murrill) Barr as Meyer et al. (2015) pointed out in their work because the abandoned galls of Dryocosmus kuriphilus could act as a point of entry and a source of pathogen inoculum.Other authors such as Rigling and Prospero (2017) have also mentioned this possible relationship.
Throughout the world, there have been several attempts at control methods against the ACGW.It has been shown that traditional treatments (pruning methods or protection of immature twigs with nets) are impractical solutions for large-scale use (Maltoni et al. 2012;Payne et al. 1975;Zhang et al. 2009).
Other biological tools such as the breeding of resistant chestnuts varieties were successfully carried out in Japan but only for twenty years, since the resistance of the plants was overcome by a new aggressive biotype of D. kuriphilus (Murakami 1981).In addition, Cooper and Rieske (2007) have already seen the inefficiency of the use of chemical pesticides against the immature stages due to the protection of plant tissues (galls).The biological control agents (BCAs) have been recognized as an alternative to the use of chemical products, which are very restricted by the European Union regulations (Directive 2009/128/EC).
With respect to the BCAs, it has been shown that the use of natural enemies such as the native parasitoid Torymus sinensis Kamijo is an effective control agent, but on a medium or long time scale (Moriya et al. 1989;Quacchia et al. 2008;Colombari and Battisti 2016;Matosevic et al. 2017).However, it should be taken into account as several authors have already mentioned (Yara et al. 2010;Gibbs et al. 2011;Cooper and Rieske 2011;Ferracini et al. 2017) that there may be a negative impacts on local fauna parasitoids due to the risk of hybridizations, hyperparasitations or displacements.
The recruitment of native parasitoids needs at least two or three years (Matosevic and Melika 2013;Palmeri et al. 2014) to be effective in reducing the gall wasp population.It should be also taken into account that the parasitization rate can be very variable depending on the different geographical locations.This rate has been classified as medium or low by authors such as Santi and Maini (2011), García (2013) or Quacchia et al. (2013).However, some species of Torymidae such as Torymus flavipes (Walker) could be a potential and effective tool in the control of the ACGW in the future (Panzavolta et al. 2013).Other authors such as Iskender et al. (2017) have even proposed other biological control techniques based on the use of the properties and possibilities of the bacteria associated with the ACGW.
Another potential alternative tool against the ACGW could be the use of entomopathogenic fungi.This was already investigated by authors such as Cooper andRieske (2007, 2010) a few years ago in the USA.These authors found within the galls, an unknown endophyte causing up to 14% of the mortality of the pupae.
Several authors have previously found the relationship between the presence of some species of Fusarium in the necrotic galls and the mortality of the ACGW.This mortality is due to the production of different types of wall-degrading enzymes (e.g.cellulases, glucanases or glucosidases) as well as entomopathogenic mycotoxins such as beauvericin, moniloformin or fumonisins (Blaney et al. 1985;Lorgrieco et al. 1998;Monzón 2001;Mirete et al. 2003;Addario and Turchetti 2011;Summerell and Leslie 2011;Tosi et al. 2014;Stepien et al. 2016).These compounds cause the death of larvae, pupae and adults.Moreover, other secondary metabolites such as abscisic acid ((+)-ABA) can also act as phytotoxins (Vinale et al. 2014).In addition, in previous studies (Bezos et al. 2013) Fusarium spp. was detected in 15% of the necrotic galls of Dryocosmus kuriphilus in Northern Spain.The objective of this work was to identify the fungal species associated with the necrotic galls of the ACGW as a first step for the use of entomopathogenic tools for the future control of the pest.

Material and methods
In 2014, two C. sativa trees (C4 and C5) affected by the ACGW were selected in Vejorís (Cantabria, Northern Spain) to carry out a biweekly sampling from July to September.ACGW galls were collected and classified in necrotic or green galls.Subsequently, data regarding the position of the gall in the branch (petiole or leave) was recorded.Moreover, before plating the necrotic galls, the following data were recorded: number of chambers made by the wasp, presence/absence of the wasp inside the chambers, the development stage of the insect (larvae, pupae or adult) as well as the state of the insect (dead or alive).
One hundred nineteen necrotic galls were plated on Potato Dextrose Agar (PDA, previously autoclaved for 20 min at 121 °C) after surface sterilization (1 min soaked in tap water, 1 min in ethanol 70%, 1 min in sodium hypochlorite 20% and finally another min in destilled sterilized water) with the aim of isolating the associated fungi species.Cultures were incubated in darkness at 23 °C.After 4 days, all outgrowing fungi were transferred by taking a ca. 9 mm 2 piece of agar from the edge of each colony to fresh medium, and during 1 month, a weekly check was carried out in order to find new colonies.Fungal isolates were counted and stored at 4 °C.Finally, assemblages were grouped according to colony morphology on PDA.Fungal isolates were classified into "colonial morphotypes" (CMSs) attending to macromorphological features based on colony color, size, texture, and presence of aerial hyphae (Wang et al. 2005).
One isolate from each CMS was selected for DNA extraction following the protocol described by Vainio et al. (1998).Once the DNA was extracted, the polymerase chain reaction (PCR) was run to amplify the Internal Transcribed Spacer region (ITS) of the rDNA with primers ITS1F (5′-TTG-GTCATTTAGAGGAAGTAA-3′) and ITS 4 (5′-TCCTCCGCTTATTGATATGC-3′) (Gardes and Bruns 1993).For amplification, the thermal cycling program was: 10 min denaturation at 95 °C followed by 13 cycles of 35 sec at 95 °C, 55 sec at 55 °C and 45 sec at 72 °C; 13 cycles of 35 sec at 95 °C, 55 sec at 55 °C and 2 min at 72 °C; 9 cycles of 35 sec at 95 °C, 55 sec at 55 °C and 3 min at 72 °C; and a final elongation 7 min at 72 °C.PCR product was sent to sequencing (Macrogen Europe) after purification (Nucleo Spin Gel and PCR Clean up, Macherey Nagel).The ITS region sequences were corrected with Genious Pro 6.1.5software package for proper search with Blast in the Gen Bank data base.

Results
During the months of July, August and September 2014, 477 galls from the selected chestnuts trees were collected.45% of the galls were located on the petiole and 55% on the leaves of the branches.119 galls showed necrosis, which represents 25% of the total galls collected (Table 1).
In relation to the gall occupancy by the insect (Table 1), 2% of galls were found with 3 chambers and 9% with 2 chambers, with most of the galls (81%) presenting a single chamber.The adult wasps began to leave the galls before the month of July.In that month and in both chestnuts, more than half of the galls were already empty.This percentage increased in the following months, reaching its maximum in August in C4 and at the end of September in C5, respectively.As for the presence of the insect within the gall, the larvae and pupae were observed until the end of July and, as of this date, only adults were detected.
Only 16% of the wasps died and more than half were adults.It was not possible to establish any direct relationship between the fungi isolated from these necrotic galls and the dead insects inside them.We were able to isolate Fusarium oxysporum and Epiccocum nigrum Link from 3 galls with dead adult insects.From a total of 125 cultures, seven fungal taxonomic units were identified from the 11 selected CMSs isolated from the necrotic galls: Pestalotiopsis Steyaert, Epicoccum nigrum, Penicillium ramulosum Visagie & K. Jacobs, Fusarium oxysporum, Gnomoniopsis smithogilvyi, Fusarium avenaceum (Fr.)Sacc., and Fusarium sp.(Table 2).The sequences from the 11 isolates identified were uploaded to the Genbank database with accession numbers from 095868 to 095878.The frequency data for each morphotype are shown in Table 3. Gnomoniopsis smithogilvyi and Fusarium avenaceum were the fungi more frequent with 25 and 24 isolates correspondingly.Epicoccum nigrum and Pestalotiopsis spp.were quite rare with only one isolate each.

Discussion
Of the seven species of fungi found in this study, we highlight the three species of Fusarium, as well as the presence of Gnomoniopsis smithogilvyi the most isolated fungus species in our study.Pestalotiopsis sp. is a ubiquitous genus that acts as an endophyte, saprotroph or pathogen in different hosts throughout the world (Jeewon et al. 2004).In addition, Lv et al. (2011) have also considered it as a useful entomopathogen in Pinus halepensis Mill.affected by the pine needle hemiberlesian scale.
Epicoccum nigrum the less isolated fungus in our study, together with the previous one, are well known for the production of secondary metabolites that frequently act as antibiotics and could be good candidates for the biological control of some pathogenic fungi such as Monilinia spp.Honey (Larena et al. 2005;Pascual et al. 1996).
Penicilium ramulosum was also isolated but like other species of the genus, it could be a saprotroph in the tree because it rarely appears as an endophyte in healthy tissues (Zamora et al. 2008).On the other hand, Chaoyang et al. (2015) found this fungus in decaying wood.
Species of Gnomoniopsis on Castanea spp.are documented as endophytes and are associated with rotten chestnuts in Italy, being in some cases, the most abundant fungus in the old galls (Gentile et al. 2009;Magro et al. 2010;Vettraino et al. 2011;Visentin et al. 2012;Vannini et al. 2012Vannini et al. , 2017;;Ugolini et al. 2014;Lione et al. 2016).This fungus was also found in New Zealand (Sogonov et al. 2008) in chestnut blight cankers in India (Dar and Rai 2013).Gnomoniopsis smithogilvyi appears on dead burrs and branches of Castanea sativa (Sogonov et al. 2008) and was also isolated as an endophyte from symptomatic flowers, leaves and stems (Shuttleworth 2012).Maresi et al. (2013) and Vinale et al. (2014) found Gnomoniopsis castanea in necrotic galls and explained its entomophatogenic capacity due to the toxicity of abcisic acid production.More recently, Vanini et al. (2017) described a severe impact of G. castanea on the vitality of D. kuriphilus, mainly affecting the adult stage within the gall.
The potential entomophatogenic capacity of Fusarium spp. as well as the production of dangerous mycotoxins for humans and animals is well known and has already been mentioned by several authors such as Teetor- Barsch and Roberts (1983), Bottalico and Perrone (2002) and Logrieco et al. (2002).Lorgrieco et al. (1998) as well as Blaney et al. (1985) also noted the ability to produce insecticidal toxins by F. incarnatum (Roberge) Sacc.and F. equiseti (Corda) Sacc.In addition, Addario and Turchetti (2011) demonstrated the effectiveness of these two fungal species to cause the death of individuals of the ACGW with a time of action of seven days penetrating directly into the tissues of the gall.The potential entomopathogenicity of Fusarium proliferatum against Dryocosmus kuriphilus has already been demonstrated by Tosi et al. (2014) in Italy reaching mortality rates of 33 to 97% in laboratory tests.Moreover, it was also seen in Argentina, that other species of the genus, Fusarium verticilloides is an effective entomopathogen against grasshoppers (Pelizza et al. 2011).
In this work, three Fusarium taxa were isolated: Fusarium oxysporum, F. avenaceum and Fusarium sp.Prakash et al. (2010) already observed that Fusarium oxysporum was an entomopathogenic fungus against Diptera larvae and that the production of (+)-ABA was responsible for the toxicity of the fungus.Other authors such as Bustillo et al. (2002) already mentioned it as a natural enemy of the coffee berry borer (Coleoptera: Scolytinae) while Asensio et al. (2007) showed its entomopathogenic capacity against the red scale (Hemiptera: Pseudoccocidae).
Fusarium avenaceum is a cosmopolitan plant pathogen with a wide and diverse range of host and is responsible for diseases in more than eighty genera of plants (Leach and Hobbs 2013) due to its capacity to produce toxins such as enniatins, moniliformin and beauvericin (Morrison et al. 2002;Kokkonen et al. 2010).Batta (2012) demonstrated for the first time its entomopathogenic capacity against the rice weevil (Coleoptera: Curculionidae) but other authors such as Strongman et al. (1987) had previously found this toxicity in the spruce budworm larvae (Lepidoptera: Tortricidae) due to the production of enniatins.In addition, Wenda-Piesik et al. (2006) found it as a colonizer in cadavers of sawflies (Hymenoptera: Symphita) larvae.
The detection of these potentially entomopathogenic species in the necrotic galls of D. kuriphilus opens the way for its use as BCAs.However, deeper studies are required to verify the applicability of this potential tool, testing it on fresh galls to see the ability of these fungi to penetrate the galls, leaves and twigs and cause the death of the insect at its different stages of development (larvae, pupae and adults).

Table 1 .
Number of collected galls, number of necrotic galls and number of insects died as larval stage, pupa or adult within the necrotic galls and percentage of empty chambers from each collecting date in both chestnuts (C4 and C5).

Table 2 .
Fungal species isolated from the morphotypes of the necrotic galls.

Table 3 .
Relative abundance for each morphotype among the isolated cultures.