FUNGI OF THE BOTRYOSPHAERIACEAE FAMILY CAUSE DIFFERENT LEVELS OF STEM CANKER ON PECAN TREES (Carya illinoinensis) IN BRAZIL

Pecan tree [Carya illinoinensis (Wangenh.) K. Koch] is a widespread species in the southern region of Brazil and commercially important for nut and wood production. Stem cankers in pecan trees negatively impact the orchard, but the biotic causes of cankers are usually not identifi ed. Symptoms of stem canker were observed in about 60% of pecan trees from orchards in the states of Rio Grande do Sul and Paraná. Therefore, this study aimed to analyze the molecular and morphophysiological characteristics of fungi associated with pecan canker in orchards in southern Brazil and to confi rm their pathogenicity in pecan seedlings. Samples from symptomatic stems were collected in diff erent municipalities and possible causal agents were isolated. Molecular identifi cation was performed by the Polymerase Chain Reaction technique for amplifi cation of the EF-1α region and subsequent sequencing. The sequencing, followed by morphophysiological aspects of mycelial growth and colony pigmentation, allowed the identifi cation of a species complex included in the Botryosphaeriaceae family. The isolates were classifi ed as Lasiodiplodia theobromae, Pseudofusicoccum kimberleyense and Neofusicoccum parvum. Pathogenicity was tested by inoculating the isolates into pecan seedlings to evaluate the symptoms. All isolates caused canker in the stem, but there was variation in the severity. Therefore, diff erent species of the Botryosphaeriaceae family are responsible for stem canker occurrence in pecan trees in the southern region of Brazil, but at diff erent levels of injury. Thus, it is crucial to identify and understand the behavior of the fungal isolates to best control cankers in pecan orchards.


INTRODUCTION
Pecan [Carya illinoinensis (Wangenh.) K. Koch] is an arboreal hickory tree, whose main product is the seed, usually consumed as nuts or sometimes as an ingredient for diff erent food recipes in many American countries (Martins et al., 2017). Besides, pecan wood has high malleability being easy to handle for crafted furniture, increasing the commercial value of the tree (Gatto et al., 2008). Pecan is also considered a feasible option to be intercropped with other crops in agroforestry systems or included in silvopastoral systems for animal husbandry (Bremm, 2018).
In addition to North America, pecan is cultivated in Brazil, Argentina, Uruguay, Australia, South Africa and China (Fronza and Hamann, 2016). In Brazil, the southern region is the major pecan producer. There was a great expansion of the pecan cultivated area in southern Brazil recently, where climate conditions are favorable to its development. However, with the increase in the cultivated area, the occurrence of diseases caused by biotic pathogens in the orchards has become more common. Species of fungal pathogens have been reported damaging pecan trees in Brazil. Some of these pathogens cause signifi cant losses, resulting in leaf spots (Lazarotto et al., 2012;Walker et al., 2016), root rot (Lazarotto et al., 2014), vascular wilt (Rolim et al., 2020a) and stem canker (Poletto et al., 2016;Rolim et al., 2020b).
Botryosphaeriaceae is a fungal family comprising 22 genera (Phillips et al., 2019) with cosmopolitan species that are mostly distributed in tropical and subtropical regions, occurring in a wide range of habitats (Abdollahzadeh et al., 2010). The species in this family have distinct characteristics, can survive as saprophytes on the plant debris, cause plant diseases or inhabit plant tissue as harmless endophytes without causing damage to plants (Abdollahzadeh et al., 2010;Luo et al., 2019). However, endophytic species may change their harmless condition to a plantparasitic habit whenever environmental conditions become unfavorable to the host or the host suff ers some injury (Phillips et al., 2019). Among the pathogenic genera of this family are Lasiodiplodia, Pseudofusicoccum, Neoscytalidium, Neofusicoccum, Diplodia, Dothiorella and Botryosphaeria, which cause diff erent symptoms in their hosts, such as bud blight, stem cankers, fruit rot and gummy disease (Abdollahzadeh et al., 2010). Therefore, the incidence of these pathogens may jeopardize the yield of some crops, including pecan.
Stem canker caused by Lasiodiplodia subglobosa was reported in pecan in Brazil, causing lengthwise cracking in the stem, necrosis of the aff ected tissues and' formation of cankers, which consequently compromised plant growth and nut production (Poletto et al., 2016). From that fi rst report, similar symptoms have been identifi ed in pecan trees in diff erent orchards of southern Brazil. Therefore, this study aimed to identify fungal species from cankers in pecan trees by using molecular and morphophysiological techniques and also confi rm their pathogenicity in pecan trees. In addition, it was verifi ed the diff erences in the level of injury caused by the isolates in the stem.

Sample collection and fungal isolation
Typical stem canker symptoms such as lengthwise cracking in the stem, necrosis of the aff ected tissues and formation of cankers in the trunk were found in commercial orchards in the municipalities of Santa Maria, Segredo, Pantano Grande, state of Rio Grande do Sul, and in Guarapuava, state of Paraná (Table 1). According to Köppen's classifi cation, in the municipalities of Rio Grande do Sul the climate is considered subtropical, with warm summers, while in Guarapuava the climate is mesothermal-humid subtropical, with no defi ned dry season.
The trunks and branches from collected symptomatic trees were deeply chipped, with a sharped sterile knife, into 10-cm fragments and stored in properly identifi ed kraft paper bags. Subsequently, the samples were sent for pathogen isolation to the Phytopathology Laboratory Elocy Minussi, Federal University of Santa Maria, Rio Grande do Sul, Brazil.
Initially, the bark was removed from the sampled fragments to fully expose the infected tissues. Then the tissues were subdivided into samples with dimensions of 4 × 4 cm. Subsequently, the samples were sterilized in a bleach solution (20% v/v) for 2 min, followed by double washing in sterile distilled water. The samples were dried by an aseptic fi lter paper and transferred to Petri dishes containing potato dextrose agar medium (PDA) with streptomycin. The samples were incubated for seven days at 25 °C under 12 h lighting. Subsequently, fungal growth from the fragments of infected tissues was observed under an optical microscope to confi rm the presence of plant pathogens. Each diff erent fungal mycelia were transferred to a Petri dish containing pine needle agar culture medium (agar: 20 g; sterile pine needle ground fragments: 250 g; distilled water: 1,000 ml;), under the same incubation conditions described above, but for 20 days to induce sporulation (Poletto et al., 2016). After the incubation period, the spores from each mycelium were collected and the isolates were purifi ed according to the monosporic culture technique (Fernandes, 1993). The isolates received the identifi cation codes GUA, SEG, SM and PAN (Table 1).

Molecular analysis
For molecular identifi cation by phylogenic test, the isolates were fi rst grown on Potato Dextrose medium (PD). DNA extraction protocol used was the "Mini-prep", according to the protocol adapted from Leslie and Summerell (2006). For DNA extraction, the mycelium was collected and macerated in liquid nitrogen. A 100 µg of mycelium was mixed to 700 µl of CTAB (Cationic hexadecyl trimethyl ammonium bromide) buff er, followed by vortexing for 2-3 s and subsequent heating at 65 °C for 25 min. In sequence, 300 µl of chloroform:isoamyl alcohol (24:1, v/v) was added to each sample, then vortexed and centrifuged for 5 min; 600 µl of the resulting aqueous phase was transferred to new tubes of 1.5 ml, 600 μl of isopropanol was added, followed by manual inversion for mixing the contents. The material rested for 5 min and then was centrifuged for 5 min. The supernatant was discarded and 600 μl of 1 × TE buff er (300 μl of phenol: chloroform: isoamyl alcohol -25: 24: 1) was added, vortexed and the tubes were placed in a centrifuge where they rested for 5 min and centrifuged at 130 RPM. With a pipette, 500 µl of the liquid was transferred to new tubes and 1 µl of RNAse was added. The tubes were vortexed and incubated at 37 ºC for 30 min. To each tube 500 µl of isopropanol (2-propanol) was added and manually inverted and rested for 5 min at room temperature, resulting in the precipitated nucleic acids. The samples were centrifuged at 130 RPM, for 5 min. for the sedimentation of DNA. The aqueous portion was discarded, and the tube was inverted on a paper towel for 5 min. The DNA pellets were washed twice with 1 ml of cold 70% (v/v) ethanol, then the paper drying process was repeated. The pellets were resuspended with 50 µl of TE buff er and stored at 4 ºC for later use. The extracted genomic DNA samples were subjected to the Polymerase Chain Reaction (PCR), for the amplifi cation of the elongation factor 1alpha (EF-1α) region with the primers EF1-728F and EF1-928R (Carbone and Kohn, 1999). The PCR reaction consisted of 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl 2 , 0.1 mM from each dNTP, 0.2 µM from each primer, 1 unit of Taq DNA polymerase (Invitrogen, ThermoFisher, California, USA) and 100 ηg of DNA. The reactions took place in a thermocycler, where the initial denaturation was carried out at 94 °C for 3 min.; followed by 40 cycles at 94 °C for 20 s; pairing step at 60 °C for 30 s; extension at 72 °C for 30 s; and fi nal incubation at 72°C for 10 min. The amplifi cations included control samples, without DNA. The amplifi ed fragments and the control were separated by electrophoresis in 1.5% agarose gel, in 1X TBE buff er (40 mM Tris Base, 40 mM acetic acid and 1 mM EDTA) containing 1 µl of red gel and visualized under ultraviolet light. The PCR products were purifi ed by precipitation with 13% polyethylene glycol 8000 (PEG 8000). Then, the samples were separated again by 1.5% agarose, in 1X TBE buff er (40 mM Tris Base, 40 mM acetic acid and 1 mM EDTA) containing red gel and visualized under ultraviolet light, to confi rm the purifi cation. Afterward, the samples were sequenced in ABI-Prism 3500 GeneticAnalyzer equipment (Applied Biosystems, ThermoFisher, California, USA). The GenBank sequences that showed the highest scores (% of similarity) were selected and aligned together with those acquired in the sequencing by the ClustalW algorithm. The sequences were aligned with the BioEdit software. The phylogenetic analysis was performed using the "Neighbor-joining" method with 1,000 replicates by MEGA software version 4 (Tamura et al., 2007). The phylogram was constructed and the similarity of the nucleotide sequences between the isolates was calculated using the Basic Local Alignment Search Tool -BLAST (Altschul et al., 1990). After molecular identifi cation, the isolates were registered on GenBank, under the numbers MT436263, MT533179, MN233653 and MN258734 for GUA, SEG, SM and PAN, respectively. All the isolates were deposited at the Herbarium of the Federal University of Santa Maria under the numbers 18,363; 18,362; 18,364; 18,365 for GUA, SEG, SM and PAN, respectively.

Morphophysiological aspects of the isolates
For the evaluation of mycelial growth, initially, 5-mm PDA disks containing structures (mycelium) from the fungal monosporic isolates were transferred to the center of 90 mm Petri dishes with PDA. The plates were incubated for fi ve days at 25 ⁰C under 12 h lighting. Mycelial growth was estimated by measuring the colony diameter every 24 h for fi ve days in two perpendicular directions to compare the isolate's growth over time. On the fi fth day, the color of the colonies and aerial mycelium of the isolates was also observed, which were determined according to the Munsell color chart (Munsell Color, 2009). For each isolate, four replicates were used, consisting of one Petri dish per replicate.

Pathogenicity test
Pecan seeds, cultivar Barton, were purchased from a nursery located in the municipality of Anta Gorda -Rio Grande do Sul. The seeds were stratifi ed in sterile sand at 4 ºC for 90 days, after which they were sown in 750 ml disposable cups containing sterilized commercial substrate (Florestal 1, MecPlant, Paraná, Brazil). Thirty days after emergence, the seedlings were transplanted to plastic bags (12 cm x 25 cm) with the sterile substrate, for appropriate growth conditions until fungal inoculation. Subsequently, 90 days-old pecan seedlings were inoculated with each fungal isolate. For this, on the stem of each seedling, 12 cm above the root collar, a circular (3 mm) wound was made, using a sterilized scalpel. Then, a 3 mm diameter PDA disk containing the mycelium of each fungal isolate was placed directly in contact with the exposed tissues inside the wound. Also, control seedlings were inoculated with a sterile PDA disk. A small amount of cotton, moistened with sterile distilled water, was placed over the discs, which were protected with plastic fi lm to prevent moisture loss in the wound. The inoculated seedlings were kept in a greenhouse at 25 °C and watered twice a day. The plants were observed for 90 days after inoculation to evaluate visually canker symptoms. At the end of the period, stem injury was evaluated according to the scale proposed by Borges et al. (2018), with minor modifi cations. This scale ranges from 0 to 4, where: Score 0 = seedlings without visible damage; Score 1 = lesion up to 3 cm long; Score 2 = lesion 3 to 6 cm long; Score 3 = lesion longer than 6 cm, and Score 4 = seedlings with lesion longer than 6 cm, deep lesion, darkening of the vessels, showing leaf fall and death. Then, the isolates were re-isolated from symptomatic tissues to complete Koch's postulates, by comparing the morphophysiological characteristics previously observed in Petri dishes.
Four inoculated seedlings per isolate were used and four non-inoculated seedlings represented the control. The data were analyzed by One-way variance analyses. The score attributed to the canker injury scale was then compared by the posthoc Tukey test (P<0.05) using SISVAR 5.3 (Ferreira, 2011).

Molecular identifi cation of pathogens
Based on the phylogenetic analysis in this study and comparing them with those deposited in GenBank, the isolates obtained belonged to three diff erent clades (Figure 1). GUA and SEG isolates were allocated in the Lasiodiplodia theobromae clade, with a bootstrap value of 58 and 88%, respectively. SM isolate was allocated in another clade with isolates belonging to the species Pseudofusicoccum kimberleyense, while PAN, was in the clade of the species Neofusicoccum parvum, with a bootstrap value of 86 % and 100 %, respectively.

Morphophysiological characteristics of isolates
There were diff erences in the mycelial growth of each isolate. The isolate L. theobromae (SEG) fi lled the entire surface of the Petri dish four days after isolation, while the other isolates required fi ve days to reach 90 mm of colony diameter. The Neofusicoccum parvum isolate (PAN) colony was the only one that did not take the whole dish surface (Figure 2). The aerial mycelium of the isolates varied from light gray (SEG and SM), greenish-gray (PAN) to greenishblack (GUA). The colonies ranged from greenishgray (PAN), very dark greenish gray (SEG and SM) to greenish-black (GUA).

Pathogenic test and stem canker symptoms
The four obtained isolates proved to be pathogenic when inoculated in pecan seedlings. The fi rst symptoms were observed thirty days after inoculation, starting with lengthwise cracking in the stems, followed by tissue necrosis and lumps in the form of cankers. GUA (Lasiodiplodia theobromae) was the fi rst isolate to cause symptoms, thirty days after inoculation. Forty days after inoculation, isolates SEG (L. theobromae) and PAN (N. parvum) also caused longitudinal cracking in the plants, whereas SM (P. kimberleyense) caused symptoms in all replicates only 65 days after inoculation. The control, inoculated only with a PDA disk, did not show any symptoms, confi rming that the observed symptoms were caused by the fungal isolates inoculated in the plants (Table 2). Ninety days after inoculation, the fi nal evaluation of symptoms was made and the lesions were measured to evaluate the injuries caused by the pathogens. Lengthwise cracking was observed in all plants inoculated with each isolate. Then, the inoculated isolates were re-isolated from symptomatic tissues, confi rming Koch's postulates. The reisolated isolates showed the same characteristics as the inoculated isolates in Petri dishes.
There was a signifi cant diff erence in the injuries and growth speed among the diff erent isolates (P<0.05) ( Table 2). GUA isolate, was the most aggressive, as it caused symptoms such as deep lesions, darkening of plant vessels, leaf fall and plant death, thus reaching a severity score of 4. However, GUA did not diff er signifi cantly from the PAN isolate, which was also potentially aggressive compared to the others. In contrast, the isolates SEG, SM, and PAN did not diff er in severity, with lesions up to 6 cm long, presenting severity scores of 2.5, 1.5 and 2.75, respectively. However, all isolates diff ered signifi cantly from the control treatment, with no symptoms (severity score of zero).

DISCUSSION
Stem canker is a signifi cant disease in the cultivation of pecan nuts, aff ecting about 60-65% of trees in the sampled orchards, which compromised the development and productivity of the trees (Table  1). All cankered plants presented signifi cant damage in the trees and frequently the death of young plants. Cankers in pecan stems are usually found in the southern region of Brazil, but few studies reported the biotic nature of such symptoms (Poletto et al., 2016;Luo et al., 2019;Rolim et al., 2020b). Here, the symptoms observed in the fi eld were replicated in the pathogenicity test. However, there was a variation in severity among the isolates. A variety of symptoms is commonly found in the Botryosphaeriaceae family as a function of the region and genetic groups as a phenotypic response (Luo et al., 2019). Furthermore, Botryosphaeriaceae isolates may cause various symptoms in the same host (Phillips et al., 2019;Abdollahzadeh et al., 2010). However, for the correct management of any disease, the fi rst step is the identifi cation of the agent causing the symptoms.
The combination of morphophysiological and molecular information is the best way to assure   (Hyde et al., 2014;Machado et al., 2014;Correia et al., 2016;Netto et al., 2016). Through amplifi cation and sequencing of the elongation factor, three species causing stem canker were identifi ed: L. theobromae (GUA and SEG), P. kimberleyense (SM) and N. parvum (PAN). Lasiodiplodia theobromae is an important cankercausing pathogen and has been reported causing canker in several hosts, such as in Fraxinus americana , Falcataria moluccana (Ji et al., 2017) and Vaccinium spp. (Borrero et al., 2019).
Pseudofusicoccum kimberleyense was associated with the canker symptom in Mangifera indica and Eucalyptus spp. (Sakalidis et al., 2011). Eucalyptus (Eucalyptus spp.) and pecan (C. illinoinensis) have also been reported hosting Neofusicoccum parvum, and the presence of the pathogen in eucalyptus has been described in many countries, such as Australia, Chile, China, Ethiopia, Indonesia, South Africa and Uruguay (Rodas et al., 2009). Furthermore, N. parvarum has been described in ninety other hosts, mainly wooden species (Sakalidis et al., 2013).
The morphophysiological aspects consolidate the distinction of each isolate, supporting the phylogenetic identifi cation. Initially, the colonies were colored with whitish tones, and over the days they acquired a gray tint and fi nally, some had a greenish-black color. The characteristics of mycelial growth and color observed in the present study have already been described for fungi of the Botryosphaeriaceae family (Pavlic et al., 2008;Liang et al., 2019;Phillips et al., 2019). Here, the pathogenic isolates grew rapidly in the culture medium. The fact that isolates grow diff erently may interfere with the management in orchards and even at nursery seedling production, where the strategies should be chosen according to their characteristics. Therefore, despite SEG and GUA grouped in the same clade of L. theobromae, they behaved diff erently in their morphophysiological aspects in in vitro tests.
Botryosphaeriaceae isolates commonly change their behavior according to environmental conditions, which may range from causing no symptoms to highly cankered plants (Cipriano et al., 2015;Slippers and Wingfi eld, 2007). The isolates SEG and GUA were considered L. theobromae, but SEG grew faster in Petri dishes, while GUA started to cause symptoms in seedlings 10 days earlier and also injured more the inoculated plants. Meanwhile, P. kimberleyense (SM) and N. parvum (PAN) caused injuries similar to L. theobromae (SEG). The severity of the isolates may be directly related to the incidence of the disease in the orchards. In this sense, L. theobromae (GUA) isolate, from an orchard with 65% disease incidence, was considered the most aggressive, since it was the fi rst to cause symptoms and serious severity. In contrast, the other isolates, from orchards with the 60% incidence, caused fewer canker symptoms and severity in all inoculated plants. Usually, in the sampled season, Guarapuava region, where GUA isolate was isolated from, has adequate moisture and temperature conditions for the development of Botryosphaeriaceae members than the other sampled regions (Abdollahzadeh et al., 2010;Phillips et al., 2019). Then, the pathogenic nature of L. theobromae (GUA) may be increased in that region. However, in morphophysiological tests in Petri plates, L. theobromae (SEG) grew faster than GUA, supporting the importance of performing multiple studies to identify and understand the isolate's behavior. Furthermore, the phylogenetic analysis provided a bootstrap value lower than 80% for GUA to be allocated in L. theobromae, which demonstrates that more regions than EF-1α should be sequenced to support the identifi cation of GUA.
Despite their diff erences, all the species found in this study have been reported causing canker in diff erent levels in other plant species (Borrero et al., 2019;Chen et al., 2019;Ji et al., 2017;Sakalidis et al., 2011). The isolates caused canker symptoms similar to L. subglobosa cankers previously reported on pecan trees in Brazil (Poletto et al., 2016). Therefore, all this evidence emphasizes the importance of the correct and early identifi cation of the Botryosphaeriaceae isolates before it signifi cantly damages the pecan tree and interferes with orchard yield.

CONCLUSIONS
Morphological and molecular techniques confi rmed the isolates from the pecan orchards as Lasiodiplodia theobromae (GUA and SEG), Neofusicoccum parvum (PAN) and Pseudofusicoccum kimberleyense (SM). All the isolates were pathogenic to pecan with diff erent injury levels, where L. theobromae (GUA) demonstrated more aggressiveness than all others. Therefore, it is crucial to identify and understand the behavior of the isolates from diff erent regions to best control canker in pecan orchards in the early stages.

AUTHOR CONTRIBUTIONS
Jessica Mengue Rolim: Data collection and organization, analysis the data, discussion of results, and text writing. Jéssica Emilia Rabuske: Data collection and text writing.Lucas Graciolli Savian: Data collection and text writing.Clair Walker: Writing and reviewing the text. Janaína Silva Sarzi: Data collection and text writing. Júlio Carlos Pereira da Silva: Writing, reviewing and text editing. Marlove Fátima Brião Muniz: Research supervision, technical review, and text review.