IntroductionTop

The avocado (Persea americana Mill.) is cultivated worldwide, but was commercially produced in Europe for the first time in Spain. In the 1970s, commercial avocado orchards were established in southern Spain (provinces of Málaga and Granada) because the microclimate in this area bears similarities to the conditions in different regions of America, such as Mexico, Peru and California, which have a long tradition of growing this crop, with high levels of production (http://faostat3.fao.org/browse/Q/QC/E).

However, avocado production is decreasing all over the world due to branch cankers and fruit stem-end rot. Symptomatic trees exhibit red-brown cankers and dieback on branches associated with a characteristic white exudate (McDonald & Eskalen, 2011). The first stages of infection are often caused by mechanical injuries, which allow the access of pathogens. Avocado dieback has been observed in different countries with tropical and subtropical climate, such as Chile (Auger et al., 2013) and Colombia (Burbano-Figueroa et al., 2018) in South America or Spain (Zea-Bonilla et al., 2007) in Europe, and many fungal agents have been identified, especially those belonging to the Botryosphaeriaceae family.

In Spain, other subtropical crops different to avo-­cado, such as loquat (Eribotrya japonica Lindl.), are affected by species of Botryosphaeriaceae, among them, Diplodia malorum Fuckel, Diplodia olivarum A.J.L. Phillips, Frisullo & Lazzizera, Diplodia seriata De Not., species of complex Diplodia pseudoseriata/Diplo­dia alatafructa, Diplodia sp., Dothiorella sar­mentorum (Fr.) A.J.L. Phillips, Alves & Luque, Neo­fusicoccum mediterraneum Crous, Wingf & A.J.L. Phillips, Neo­fusicoccum parvum (Pennycook & Samuels) Crous, Slippers & A.J.L. Philips, Spencermartinsia plurivora Abdollahz., Javadi & A.J.L. Phillips and Spencermartinsia viticola (A.J.L. Phillips & J. Luque) A.J.L. Phillips, A. Alves & Crous (González-Domínguez et al., 2017). In this crop, other pathogens, namely Alter­naria alternata (Fr.) Keissl., Penicillium expansum Link, Botrytis cinerea Pers., Colletotrichum gloeosporioides (Penz.) Penz. & Sacc., Pestalotiopsis clavispora (G.F. Atk.) Steyaert and D. seriata caused post-harvest diseases (Palou et al., 2016). Other Mediterranean crops such as almond (Prunus dulcis Webb) are also affected by fungi Botryosphaeriaceae (Gramaje et al., 2012) and N. mediterraneum and Botryosphaeria dothidea (Moug. ex Fr.) Ces. & De Not. have been isolated from olive (Olea europaea L.) branches (Moral et al., 2017) and N. parvum from mango (Mangifera indica L.) trees (Arjona-Girona & López-Herrera, 2016).

Colletotrichum gloeosporioides and N. parvum have been described as causal agents of anthracnose and stem end rot in avocado fruit (cv. Hass) in Turkey (Akgül & Awan, 2016). N. parvum and D. seriata caused dieback in grapevine (Vitis vinifera L.) (Spagnolo et al., 2017). Over the last two decades, significant losses have been recorded in citrus production in Portugal from anthracnose symp­toms caused by C. gloeosporioides (Ramos et al., 2016).

Neofusicoccum parvum, Neofusicoccum luteum (Penny­cook & Samuels) Crous, Slippers & A.J.L. Phillips and Neofusicoccum australe (Slippers, Crous & M.J. Wingf.) Crous, Slippers & A.J.L. Phillips were identified in California (McDonald et al., 2009). N. parvum was also found in Mexico (Molina-Gayosso et al., 2012), mainly affecting fruits, and Lasiodiplodia theobromae (Pat.) Griffon & Maubl. in Peru (Alama et al., 2006). Although the most important post-harvest disease in Chile is anthracnose, caused by C. gloeosporioides, a new disease caused by N. australe was also found (Montealegre et al., 2016).

The aims of this work were to study the distribution of avocado trees affected by branch cankers in commercial orchards in southern Spain, identify their fungal agents and establish their virulence groups.

Material and methodsTop

Sampling and fungal isolation

Field surveys were carried out on avocado orchards from May to October 2014 in Málaga (southern Spain). Young twigs and branches of avocado trees from commercial orchards showing dieback symptoms were isolated (Fig. 1A-D).

Figure 1. Disease symptoms on avocado tree branches associated with dieback. A, Dry branches. B, External necrosis in twigs. C, Internal lesions in branches. D, External lesions with exudates.

The collected plant material was disinfested in 10 g L-1 sodium hypochlorite for 3 min and pieces of bark or internal wood showing lesions were plated onto potato dextrose agar (PDA) medium (Difco Laboratoires, Detroit, MI, USA) with lactic acid (0.2%). The cultures were kept on acidified PDA at 25°C in darkness for 3 days, and later, pure cultures were obtained by excising and transferring hyphal tips from the fungal colonies to fresh PDA plates.

Fungal identification

The pure cultures obtained were first identified based on colony morphology and conidial characteristics by comparing them with the previous studies by Phillips (2006).

To confirm the previous macroscopic fungal identifi­cation, DNA extractions from each isolate recovered with a similar colony morphology to Botryosphaeriaceae isolates were performed as described by Choi et al. (1992), and a sequence analysis of the internal transcribed spacer (ITS) nrDNA region using the primers ITS4 and ITS5, an analysis of partial β-tubulin gene BT2a and BT2b (Glass & Donaldsons, 1995) and a translation elongation factor 1-α gene regions (EF1-α) EF1-728F and EF1-986R (Carbone & Kohn, 1999) were performed. PCR products (600 bp for ITS, 340-495 bp for BT2 and 350 bp for EF-1α) were sequenced in the two ways (3´-5´and 5´-3´) by the Proteomic Department of SCAI (University of Córdoba, Spain). The sequences of each isolate were used to search for similar sequences in GenBank using BLAST (v. 2.0, National Center for Biotechnology Information, US Nat Inst of Health, Bethesda, MD, USA).

PDA Petri dishes with pure cultures were incubated at 25 ºC in darkness and pycnidia formation was induced on water agar with pine needles and UV light; the length and width of the conidia were then measured.

Pathogenicity tests

Pathogenicity tests were carried out on stems of eighteen-month-old avocado plants of cv. Topa-Topa, growing on Laura substrate consisting of peat, coconut fiber and perlite at a ratio of 6:1:0.6 v/v/v. One plant per fungal isolate was used and five wounds were made on each stem. Five-mm mycelia plugs from the edge of a fresh fungal colony were placed onto the wounded stem and incubated in a greenhouse at 25 ºC ± 5 ºC. The necrotic lesions were measured after 3, 6 and 9 days of inoculation and the standardised area under disease progress curves (AUDPCs) was calculated. To fulfil Koch’s postulates, pieces of necrotic tissue were taken from infected twigs and plated on PDA Petri dishes to identify the pathogen.

Statistical analysis

A completely randomised experimental design with ‘Statistic 9’ was used to study the virulence of the isolates. The treatment averages were compared using Fisher’s Least Significant Difference (LSD) test to separate the means (p<0.05) (Steel & Torrie, 1985).

ResultsTop

Fungal identification

Amplified sequences from each of the 14 representative isolates were compared with isolates from GenBank (accession numbers: see Table 1). Ba­sed on a BLAST search of Gen-Bank nucleotide database, the closest hits of the isolates with the ITS, BT2 and EF1-α sequences are shown in Table 1. The identity percentages between nucleotides were very high – close to 100% in most cases. The number of insertions and deletions in one sequence relative to another were low because there were not many gaps (0-1%), and therefore it was not necessary to introduce a space into an alignment to compensate for the other.

After the macroscopic fungal identification of the morphological characters of 68 isolates and their subsequent confirmation by DNA sequence analyses, all the isolates were identified as belonging to the genera Neofusicoccum, Colletotrichum and Lasiodiplodia, with six different species in different proportions: N. parvum (50%), C. gloeosporioides (18%), N. australe (13%), N. luteum (16%), N. mediterraneum (1.5%) and L. theobromae (1.5%) (Table 2).

The width and length of the conidia (Fig. 2A-F) from 14 representative fungal isolates: N. parvum (AR1, AR7, AR10, AR17 CA, AR22 and AR33), C. gloeosporioides (AR14 CA and AR28 P), N. australe (AR31 and AR41-2), N. luteum (AR18 and AR46), N. mediterraneum (AR30) and L. theobromae (AR12 G) were then measured (Table 3). L. theobromae (21.99±1.43 μm × 13.18±0.5 μm) showed the greatest conidia size, although the shape of the conidia was not the most elongated. N. parvum (17.97±1.73 μm × 6.5±0.49 μm) and N. mediterraneum (22.66±1.57 μm × 7.03±0.18 μm) were of medium size but had a more elongated shape and the rest, C. gloeosporioides (18.14±1.29 μm × 5.23±0.49 μm), N. luteum (16.62±3.25 μm × 4.30±0.54 μm) and N. australe (16.89±1.3 μm × 4.09±0.39 μm) were the smallest, but had the greatest elongation.

Figure 2. Conidia from representative isolates. A, Neofusicoccum parvum. B, Colletotrichum gloeosporioides. C, Neofusicoccum austral. D, Neofusicoccum luteum. E, Neofusicoccum mediterraneum. F, Lasiodiplodia theobromae. Scale bar= 20 µm

Neofusicoccum parvum is the main pathogen causing branch dieback in avocado plants and it was present in all the locations sampled, with a total incidence of 50%. N. luteum was also common, while C. gloeosporioides and N. australe were located only at the biggest locations (Benamocarra and Vélez Málaga) with rates of incidence of 16.2%, 17.6% and 13.2%, respectively. N. mediterraneum and L. theobromae appeared occasionally, with low rates of incidence of around 1.5% (Table 4).

Pathogenicity test

The avocado plants artificially inoculated with fun­gal isolates showed necrotic stem lesions (Fig. 3) and the pathogenicity of all the genera of the fungal isolates was evaluated and compared (F=24.54; df=2, 342; p<0.05). Neofusicoccum was the most virulent genus, with significant differences with Colletotrichum and Lasiodiplodia, although no differences were observed between these last two genera. We also evaluated the pathogenicity of the different species of the Neofusicoccum genus (F=20.03; df=3, 276; p<0.05), with N. parvum being the most virulent, significantly different from N. australe, N. mediterraneum and N. luteum, with no differences observed among the latter. Finally, the pathogenicity of species from the different genera was also evaluated (F=26.33; df=5, 339; P<0.05) with the following results (in the decreasing order of pathogenicity): N. parvum, N. luteum, N. mediterraneum, N. australe, C. gloeosporioides and L. theobromae; there were also significant differences between N. parvum and the other species, and the differences between N. luteum and C. gloeosporioides were also observed.

Figure 3. Artificial inoculations. A, Avocado plant with points of inoculations (red arrowheads). B, Mycelia plugs (5-mm in diameter) placed onto wounded stem and sealed with parafilm. C, Necrotic lesions on avocado stem.

Significantly different virulence groups were esta­blished within each species (from 68 isolates) with more than one representative isolate, and 10 groups for N. parvum, 6 for N. luteum, 3 for N. australe and 5 for C. gloeosporioides were obtained (Table 5). The viru­lence was homogenized for N. australe because it was defined in only three groups, although with a similar number of isolates for N. luteum and C. gloeosporioides, the virulence was defined in more groups (5 or 6).

DiscussionTop

This study shows the incidence and diversity of fungal isolates associated with avocado dieback in commercial orchards in southern Spain. The list in­cludes different genera (Neofusicoccum, Colletotri­chum and Lasiodiplodia) and species (N. parvum, N. luteum, N. mediterraneum, N. australe, C. gloeos­porioides and L. theobromae) that are also usually involved in dieback in avocado orchards in other countries (McDonald & Eskalen, 2011).

Neofusicoccum parvum is the most abundant spe­cies, with an incidence of 50%, followed by C. gloeosporioides with 18%, N. australe and N. luteum with 13% and 16%, respectively, and finally, N. mediterraneum and L. theobromae, which appear only occasionally with an incidence of 1.5% in each case. Our results agree with the greater incidence of the species N. parvum and N. australe associated with almond orchards in other areas of Spain (Gramaje et al., 2012). N. parvum is consi­de­red the most common species associated with grapevine decline syndrome (Armengol et al., 2001; Aroca et al., 2006) and N. mediterraneum was also isolated from olive fruits in southern Spain, showing symptoms of Dalmatian disease (Moral et al., 2010).

In other countries, N. parvum, N. luteum and N. australe associated with avocado dieback have also been described in California (McDonald et al., 2009). Typical Dothiorella canker symptoms observed inclu­ded darkened and friable bark showing a dry, white, powdery exudate. These studies concluded that the higher incidence of these pathogens is a consequence of the high-density planting frequent in Californian avocado crops and they recommend more suitable management strategies (McDonald & Eskalen, 2011). N. luteum was identified in California as the main cause of stem-end rot in harvested avocado fruit (65%), followed by C. gloeosporioides with an incidence of 35% (Twizeyimana et al., 2013). In the same way, the fungi detected in avocado branch cankers in Spain could also affect the fruit directly, leading to a fall in production. L. theobromae has been also described as the causal agent of avocado dieback in Peru (Alama et al., 2006). The symptoms of cankers and red-brown lesion with white exudates observed after artificial inoculation were similar to natural infection. There are other examples in which species of these genera have been associated to avocado causing anthracnose and stem-end rot in Turkey (Akgül & Awan, 2016), or to other plants such as oak (Quercus robur L.) in Portugal (Barradas et al., 2013) and olive in Tunisia (Triki et al., 2015) causing dieback, or shoot blight and plant decay on pomegranate (Punica granatum L.) in Italy (Riccioni et al., 2017).

In our study, we conclude that N. parvum and N. luteum are widely extended, while C. gloeosporioides and N. australe are located only in the biggest areas, which could be due to the wider dispersion or higher production of conidia of N. parvum and N. luteum. Future experiments should be carried out using spore trapping (Eskalen et al., 2013) to confirm this theory. Our results, showing the greater virulence of isolates of N. parvum and N. luteum when compared with C. gloeosporioides and the greater virulence of N. parvum vs N. australe, coincide with the results of Eskalen et al. (2013), who reported that lesions caused by N. parvum and N. luteum were larger than those caused by N. australe. N. mediterraneum and L. theobromae appeared occasionally, but did not appear to be a great threat.

Although N. parvum has been previously described in avocado (Zea-Bonilla et al., 2007) and in other crops such as blueberry (Vaccinium spp.) (Castillo et al., 2013) and mango (Arjona-Girona & López-Herrera, 2016) on the southern coast of Andalusia, Spain, an increased incidence of dieback on branches in avocado orchards has been observed and this could constitute a serious threat, in a similar way to N. luteum, N. australe and C. gloeosporioides, to the yield of avocado orchards in this area. The fungal inoculum tends to increase due to the poor management of pruning remains, which are often shredded and mixed into the soil instead of being burned, as famers usually do in this area.

This is the first study of avocado cankers on branches in southern Spain which evaluates the causal agents and establishes its pathogenicity groups. N. parvum is the most abundant species observed, and is the most important causal agent of dieback avocado in this area. N. luteum, N. australe and C. gloeosporioides showed the lower incidence as causal agents of the disease.

AcknowledgmentsTop

The authors thank the staff of TROPS of Vélez-Malaga (Spain) for their technical assistance in avocado orchard surveys for this study.