Identification and Characterization of Diaporthe spp. Associated with Twig Cankers and Shoot Blight of Almonds in Spain

Two hundred and twenty-five Diaporthe isolates were collected from 2005 to 2019 in almond orchards showing twig cankers and shoot blight symptoms in five different regions across Spain. Multilocus DNA sequence analysis with five loci (ITS, tub, tef-1α, cal and his), allowed the identification of four known Diaporthe species, namely: D. amygdali, D. eres, D. foeniculina and D. phaseolorum. Moreover, a novel phylogenetic species, D. mediterranea, was described. Diaporthe amygdali was the most prevalent species, due to the largest number of isolates (85.3%) obtained from all sampled regions. The second most frequent species was D. foeniculina (10.2%), followed by D. mediterranea (3.6%), D. eres and D. phaseolorum, each with only one isolate. Pathogenicity tests were performed using one-year-old almond twigs cv. Vayro and representative isolates of the different species. Except for D. foeniculina and D. phaseolorum, all Diaporthe species were able to cause lesions significantly different from those developed on the uninoculated controls. Diaporthe mediterranea caused the most severe symptoms. These results confirm D. amygdali as a key pathogen of almonds in Spain. Moreover, the new species, D. mediterranea, should also be considered as a potential important causal agent of twig cankers and shoot blight on this crop. Agronomy 2020, 10, 1062; doi:10.3390/agronomy10081062 www.mdpi.com/journal/agronomy Agronomy 2020, 10, 1062 2 of 23


Introduction
The worldwide cultivated area for almond (Prunus dulcis (Mill.) D.A. Webb) is over 2,000,000 ha. Spain, with 657,768 ha, is the country with the largest area for almond production in the world, followed by the United States, with 441,107 ha [1]. Almond is the second largest tree crop in Spain, after olive, and it is widely distributed in all regions of the country [2]. Nevertheless, Spain only contributes approximately 10% to world almond production, because the trees have been traditionally grown under rain-fed conditions and planted in marginal areas with poor soils, low rainfall and a high incidence of frost [3], thus presenting low average yields (5154 kg ha −1 ) [1].
In recent years, almond production in Spain has been experiencing a highly favorable period, in which crop intensification, with the introduction of drip irrigation and the use of new highly productive cultivars, has increased the yield in new plantations [4]. However, the incidence of almond-associated fungal diseases, such as twig cankers and shoot blight caused by Diaporthe spp., is increasing and compromises crop productivity, especially in coastal areas with higher humidity and milder temperatures [5,6].
Diaporthe amygdali (Delacr.) Udayanga, Crous and K.D. Hyde is considered the causal agent of twig canker and shoot blight of almond and peach (Prunus persica (L.) Batsch) [7,8]. Symptoms of this disease are characterized by the quick desiccation of buds, flowers and leaves in late winter or early spring. Brown lesions (1 to 5 cm diameter), initially formed around buds on green shoots, further develop into annual sunken cankers, sometimes with a gummy exudate, as well as withering of twigs. As a result, leaves wilt and, when the disease is severe, defoliation can occur. In summer, pycnidia develop just under the dry canker epidermis [7,9,10].
The species D. amygdali was first described as Fusicoccum amygdali Delacr., associated with almond cankers in France [11]. Tuset and Portilla [9] re-examined the type specimen of F. amygdali and, based on morphology and symptomatology, they re-classified this fungus into Phomopsis as P. amygdali (Delacr.) J.J Tuset and M.T. Portilla. Additionally, they also considered P. amygdalina Canonaco to be a synonym of P. amygdali. Diogo et al. [8] used morphological, molecular and pathogenicity data to clarify the identity of a collection of Phomopsis isolates obtained from almond in Portugal. In this research, as no cultures of P. amygdali were linked unequivocally to any existing type, the authors proposed the fungus in voucher CBS-H 20420 (from Portugal) as the epitype for this species (isolate CBS 126679). Udayanga et al. [12] re-evaluated the phylogenetic species recognition in the genus Diaporthe using a multi-locus phylogeny based on the internal transcribed spacer (ITS) region of the nuclear rDNA, and partial sequences from translation elongation factor 1-α (tef-1α), β-tubulin (tub) and calmodulin (cal) genes. In this study, P. amygdali was transferred into Diaporthe as D. amygdali based on multi-locus DNA sequence data.
In recent years, the taxonomy of the genus Diaporthe has been deeply revised. The generic names Diaporthe and Phomopsis are no longer used to distinguish different morphs of this genus, as Rossman et al. [13] proposed that the genus name Diaporthe should be retained over Phomopsis because: (i) it was introduced before Phomopsis and (ii) Diaporthe represents the majority of species described, and therefore it has priority over Phomopsis. Diaporthe was historically considered as monophyletic based on its typical sexual morph and Phomopsis's asexual morph [14]. However, Gao et al. [15] revealed its paraphyletic nature. Recent studies have demonstrated that morphological characters are inadequate to define species in this genus [16], due to their variability under changing environmental conditions [14]. Therefore, genealogical concordance methods based on multi-gene DNA sequence data provide a better approach to resolving the taxonomy for Diaporthe [17].
Literature about recent characterization studies of collections of Diaporthe isolates, obtained exclusively from almonds or including them together with isolates from other fruit or nut crops, is very scarce. Diogo et al. [8] examined Diaporthe isolates from almond and other Prunus species in Portugal through combining morphology, pathogenicity data and a phylogenetic study based only on ITS sequences. These authors concluded that D. amygdali was the main species on almond, reported D. neotheicola for the first time on this host and a third species represented by a single isolate, which could not be unequivocally identified. Later, Lawrence et al. [18] characterized morphologically different Diaporthe isolates associated with wood cankers of fruit and nut crops in northern California, including three almond isolates, which were assigned to the species D. australafricana and D. novem, based on multi-gene, ITS, tef-1α and cal sequence analyses.
In Spain, the studies of Tuset and Portilla [9] and Tuset et al. [10] described almond diseases and their associated pathogens, including D. amygdali. These studies were based solely on the morphological characterization of the isolates. Additional studies using molecular tools to ascertain the identity of representative sets of Diaporthe isolates from almond in this country are lacking. Gramaje et al. [19] reported only one isolate of D. amygdali, which was collected from a survey of wood-associated fungal trunk pathogens of almond trees on the island of Mallorca. Thus, the objectives of the present study were: (i) to characterize a wide collection of Diaporthe isolates collected from almond trees in Spain by means of phenotypical characterization (fungal morphology and temperature growth) and DNA sequence analyses and (ii) to evaluate the pathogenicity of these Diaporthe isolates to almond twigs. The final goal was to obtain updated and more complete information about the Diaporthe species causing twig cankers and shoot blight of almonds in Spain.

Sampling and Isolation
A total of 225 Diaporthe isolates were collected from 2005 to 2019 in almond orchards showing twig cankers and shoot blight symptoms (Figure 1) in five different regions across Spain (Andalucía (n = 56), Islas Baleares (n = 39), Cataluña (n = 43), Comunidad Valenciana (n = 76) and La Rioja (n = 11)). Fo isolation, wood segments with cankers were cut from the affected branches, washed under running tap water, surface disinfected for 1 min in a 1.5% sodium hypochlorite solution and rinsed twice in sterile distilled water. Small pieces of affected tissues taken from the margin of the lesions were plated on potato dextrose agar (PDA; Biokar-Diagnostics, Zac de Ther, France) supplemented with 0.5 g/L of streptomycin sulphate (Sigma-Aldrich, St. Louis, MO, USA) (PDAS). Plates were incubated at 25 • C in the dark for 7 to 10 d, and all colonies were transferred to PDA. All isolates were hyphal-tipped and maintained in 15% glycerol solution at −80 • C in 1.5 mL cryovials in the fungal collection of the Instituto Agroforestal Mediterráneo-Universitat Politècnica de València (IAM-UPV) (Spain) ( Table 1).
Agronomy 2020, 10, x FOR PEER REVIEW  3 of 26 reported D. neotheicola for the first time on this host and a third species represented by a single isolate, which could not be unequivocally identified. Later, Lawrence et al. [18] characterized morphologically different Diaporthe isolates associated with wood cankers of fruit and nut crops in northern California, including three almond isolates, which were assigned to the species D. australafricana and D. novem, based on multi-gene, ITS, tef-1α and cal sequence analyses.
In Spain, the studies of Tuset and Portilla [9] and Tuset et al. [10] described almond diseases and their associated pathogens, including D. amygdali. These studies were based solely on the morphological characterization of the isolates. Additional studies using molecular tools to ascertain the identity of representative sets of Diaporthe isolates from almond in this country are lacking. Gramaje et al. [19] reported only one isolate of D. amygdali, which was collected from a survey of wood-associated fungal trunk pathogens of almond trees on the island of Mallorca. Thus, the objectives of the present study were: (i) to characterize a wide collection of Diaporthe isolates collected from almond trees in Spain by means of phenotypical characterization (fungal morphology and temperature growth) and DNA sequence analyses and (ii) to evaluate the pathogenicity of these Diaporthe isolates to almond twigs. The final goal was to obtain updated and more complete information about the Diaporthe species causing twig cankers and shoot blight of almonds in Spain.

Sampling and Isolation
A total of 225 Diaporthe isolates were collected from 2005 to 2019 in almond orchards showing twig cankers and shoot blight symptoms (Figure 1) in five different regions across Spain (Andalucía (n = 56), Islas Baleares (n = 39), Cataluña (n = 43), Comunidad Valenciana (n = 76) and La Rioja (n = 11)). For isolation, wood segments with cankers were cut from the affected branches, washed under running tap water, surface disinfected for 1 min in a 1.5% sodium hypochlorite solution and rinsed twice in sterile distilled water. Small pieces of affected tissues taken from the margin of the lesions were plated on potato dextrose agar (PDA; Biokar-Diagnostics, Zac de Ther, France) supplemented with 0.5 g/L of streptomycin sulphate (Sigma-Aldrich, St. Louis, MO, USA) (PDAS). Plates were incubated at 25 °C in the dark for 7 to 10 d, and all colonies were transferred to PDA. All isolates were hyphal-tipped and maintained in 15% glycerol solution at −80 °C in 1.5 mL cryovials in the fungal collection of the Instituto Agroforestal Mediterráneo-Universitat Politècnica de València (IAM-UPV) (Spain) ( Table 1).

DNA Extraction, PCR Amplification and Sequencing
Mycelium was scraped from 10-day-old fungal cultures grown on PDA medium. Total fungal DNA was extracted using the E.Z.N.A. Plant DNA Kit (Omega Bio-tek, Norcross, GA, USA), following the manufacturer's short protocol instructions.

Phylogenetic Analyses
Sequences generated in this study were compared with reference sequences in the GenBank nucleotide database to determine the closest relatives for the phylogenetic studies. Fo each of the five loci (ITS, tub, tef-1α, cal and his), the DNA sequences obtained in this study (Table 1), together with those retrieved from GenBank (Table 2), were aligned using the ClustalW algorithm included in the MEGAX software package [29,30]. The alignments were analyzed and adjusted manually when necessary. Ambiguous sequences at either end of the alignments were excluded prior to analyses. Concatenated datasets were built in Sequence Matrix v.1.8 [31].  Phylogenetic analyses were based on Bayesian inference (BI), maximum likelihood (ML) and maximum parsimony (MP). Bayesian analyses were performed using MrBayes v 3.2 on the CIPRES Science Gateway v 3.3 [32,33]. The best-fitting model of nucleotide evolution for each partition was determined by MrModeltest 2.3 using the Akaike information criterion (AIC) [34]. Four simultaneous analyses were run for 100 million generations, sampling every 10,000, with four Markov chain Monte Carlo (MCMC) chains. The first 25% of saved trees were discarded and posterior probabilities were determined from the remaining trees. The ML analyses were done with the tool Randomized Axelerated Maximum Likelihood RAxML-HPC2 on XSEDE implemented on CIPRES Science Gateway v 3.3 [35]. ML tree searches were performed under the generalized time-reversible with gamma correction (GTR + Γ) nucleotide substitution model using 1000 pseudoreplicates. The other parameters were used as default settings. MP analyses were performed in MEGA X with the tree Bisection and reconnection (TBR) algorithm, where gaps were treated as missing data. The robustness of the topology was evaluated by 1000 bootstrap replications [36]. Measures for the maximum parsimony as tree length (TL), consistency index (CI), retention index (RI) and rescaled consistency index (RC) were also calculated.

Taxonomy
Agar plugs (6-mm diameter) were taken from the edge of actively growing cultures on PDA and transferred onto the center of 9-cm diameter Petri dishes containing one of the following culture media: malt extract agar (MEA; Sigma-Aldrich Laboratories), PDA, 2% tap water agar supplemented with sterile pine needles (PNA) or and oatmeal agar (OA; 60 g oatmeal, 12.5 g agar, Difco, Le Pont de Claix, France). Plates were then incubated at 21-22 • C under a 12 h/12 h near-ultraviolet light/darkness cycle to induce sporulation as described by Guarnaccia et al. (2018). Cultures were examined periodically for the development of ascomata and conidiomata. Colony colors were rated only on PDA after 15 days of incubation according to Rayner [37]. Morphological characteristics were examined using an Axio Scope A.1 microscope (Zeiss, manufacturer data) after mounting single pycnidia in lactic acid. Fungal structures were measured (30 measurements per type of structure) using the Zeiss AxioVision LE imaging device. Photos were captured using a Zeiss AxioCam MRm digital camera from images recorded with the 40× objective. Descriptions, nomenclature and illustrations of taxonomic novelties were deposited in MycoBank (MB 836048).
The effect of temperature on the mycelial growth of selected isolates of the species D. mediterranea (DAL24, DAL34 and DAL174) was measured on PDA. Fo this purpose, agar plugs (6-mm diameter) obtained from the growing edge of colonies were transferred to the center of PDA plates, which were incubated at 5, 10, 15, 20, 25, 30, 35 or 40 • C in darkness. Four replicates for each isolate and temperature combination were used. Growth was determined after 7 days in two orthogonal directions, and the mean growth rate was calculated in mm/day using a simplified version of the non-linear equation proposed by Duthie et al. [38]. Regression curves were fitted to the data using the R function "nls" included in the "stats" package [39,40].

Pathogenicity Tests
Pathogenicity tests were conducted as described by Diogo et al. [8]. One-year-old twigs of almond cv. Vayro, about 30 cm long, were inoculated with a set of 14 representative isolates of the five Diaporthe species found associated with P. dulcis in this study: D. amydgali Wounds were made in the center of each twig with a 6-mm cork borer. Colonized agar plugs with mycelium of about the same size, which were obtained from active 10-day-old colonies growing on PDA, were inserted underneath the epidermis and the wounds were sealed with Parafilm. Inoculated twigs were kept in an upright position with their lower ends immersed in 1 L jars with 500 mL of sterile water in a growth chamber at 23 • C with 12 h of light per day. The twigs were covered with a plastic bag during the first 4 days to keep a moist environment. Six twigs per isolate were used and a negative control was prepared using uncolonized PDA plugs. Jars were arranged in a completely randomized design and the water was changed every 3 days. The experiment was repeated once.
Lesion lengths were measured 15 days after inoculation. Immediately after lesion measurements, two representative shoots per inoculated isolate and replicate were surface sterilized as described above. Small internal fragments were cut from the margin of the healthy and necrotic tissue and placed onto PDA. Plates were incubated at 25 • C in the dark for 7 to 10 d, and all fungal growths resembling Diaporthe were transferred to PDA. A representative subsample, one culture from each of the 14 isolates and replicates, were subjected to DNA extraction and molecular identification as described above to satisfy Koch's postulates.
Significance levels for mean values of lesion length (cm), corresponding to different Diaporthe spp. isolates inoculated and control detached twigs, were determined. The analyses were performed considering individual isolates and groups of isolates from each Diaporthe spp. ANOVA assumptions were verified using Shapiro-Wilk and Levene's tests. The datasets did not meet ANOVA assumptions, thus the analysis was performed using the Kruskal-Wallis test. Control twigs were compared with the inoculated ones considering individual isolates, and different species were compared with D. amygdali using the Wilcoxon rank sum test (p < 0.01). The analyses were performed in R using the agricolae and stats packages [39,40].

Phylogenetic Analyses
Three loci (ITS region and fragments of tub and tef-1α genes) were sequenced in all Diaporthe isolates (n = 225) obtained in this study and compared with those in GenBank. The BLAST search showed high identity with D. amygdali, D. eres, D. foeniculina and D. phaseolorum accessions. Fo phylogenetic analyses, two representative isolates of closely related species, i.e., the ex-type together with one additional isolate when possible, were selected as references, and their corresponding sequences were retrieved from GenBank (Table 2). These sequences (n = 38), including Diaporthella corylina strain CBS 121124 which was used as outgroup, were added to those of the Spanish isolates (n = 225). The MP three-locus phylogeny showed that Spanish isolates of Diaporthe grouped into five distinct clades, four of them with known Diaporthe species (data not shown). The most abundant group, with 192 isolates, clustered with the ex-type isolate of D. amygdali (CBS 126679), the second (n = 23) with the ex-type of D. foeniculina (CBS 111553) and two single isolates each grouped with the ex-type of D. eres (CBS 109767), and with D. phaseolorum (CBS 116019). The remaining isolates (n = 8) clustered together, closely related to, but separated from, D. sterilis (CBS136969), suggesting that they could belong to a new species.
For accurate resolution of the species limits of our isolates, fragments of his and cal genes were sequenced in a set of 70 and 39 representative D. amygdali isolates, respectively, and for all isolates of the other groups (Table 1). The selection of the D. amygdali isolates was based on the province/region of origin and year of isolation. In addition, all GenBank sequences (ITS and tub) of two undescribed Diaporthe isolates (PMM 1657 and PMM 1660), which shared 100% identity with these loci of the potential new species, were included in the analyses (Table 2). Then, MP, ML and BI phylogenetic trees were constructed for the five-locus combined dataset, which included all taxa (n = 265) regardless of the level of completeness. A total of 2826 characters, including gaps (ITS: 1-564, tub: 565-1384, tef-1α: 1385-1814, his: 1815-2300 and cal: 2301-2826), were used in phylogenetic analyses, of which 1494 were constant and 817 were parsimony informative. The MP analysis yielded a single most parsimonious tree (TL = 2387; CI = 0.647; RI = 0.955; RC = 0.618). The ML analysis resulted in a single best tree with the final ML optimization likelihood = −15029.27737. In the BI analysis, the ITS/tub/tef-1α/his/cal partitions had 158/341/310/210/308 unique site patterns, respectively, and the analysis read a total of 40,004 trees, sampling 30,004 of them. The topologies and branching order of the inferred trees were compared visually, and they were fully congruent among themselves and with the previous ITS/tub/tef-1α multilocus phylogeny. The ITS/tub/tef-1α/his/cal ML tree is presented with the support of all phylogenetic methods at the branches (Figure 2). In the BI analysis, the ITS/tub/tef-1α/his/cal partitions had 158/341/310/210/308 unique site patterns, respectively, and the analysis read a total of 40,004 trees, sampling 30,004 of them. The topologies and branching order of the inferred trees were compared visually, and they were fully congruent among themselves and with the previous ITS/tub/tef-1α multilocus phylogeny. The ITS/tub/tef-1α/his/cal ML tree is presented with the support of all phylogenetic methods at the branches (Figure 2).  Diaporthe amygdali represented 85.3% of the studied isolates and they were obtained from all sampled regions. The second most frequent species was D. foeniculina, with 23 isolates (10.2% of total), and it was recovered in all sampled regions except in La Rioja. Diaporthe eres and D. phaseolorum, each with only one isolate, were recovered from the Andalucía region. The remaining isolates (n = 8, 3.6% of total) were grouped together with 92% and 98% bootstrap support for MP and ML, respectively, and with 1 of BI posterior probability, but not with any known Diaporthe species. Therefore, they were putatively identified as belonging to a novel species described here and named D. mediterranea. This new species was obtained from the Islas Baleares and Comunidad Valenciana regions.

Taxonomy
Based on both the results of the phylogenetic inference and morphological characters, one new species of Diaporthe is described below (Figure 3).
Agronomy 2020, 10, x FOR PEER REVIEW 20 of 26 current study and the novel taxa is shown in red. Ex-type strains are indicated in bold. The tree is rooted using Diaporthella corylina (CBS121124). The scale bar represents the expected number of nucleotide substitutions per site.
Diaporthe amygdali represented 85.3% of the studied isolates and they were obtained from all sampled regions. The second most frequent species was D. foeniculina, with 23 isolates (10.2% of total), and it was recovered in all sampled regions except in La Rioja. Diaporthe eres and D. phaseolorum, each with only one isolate, were recovered from the Andalucía region. The remaining isolates (n = 8, 3.6% of total) were grouped together with 92% and 98% bootstrap support for MP and ML, respectively, and with 1 of BI posterior probability, but not with any known Diaporthe species. Therefore, they were putatively identified as belonging to a novel species described here and named D. mediterranea. This new species was obtained from the Islas Baleares and Comunidad Valenciana regions.

Taxonomy
Based on both the results of the phylogenetic inference and morphological characters, one new species of Diaporthe is described below (Figure 3). Etymology. Named after the Mediterranean Sea, because this species was found on almond trees from orchards located in the Alicante province (Comunidad Valenciana) and Mallorca (Islas Baleares) in Mediterranean coastal areas of Spain.
Culture characteristics: Colonies covering the medium within 7 d at 25 °C, with moderate aerial mycelium. Colonies on MEA, PDA and OA white at first, becoming light cream, mycelium flat on Etymology. Named after the Mediterranean Sea, because this species was found on almond trees from orchards located in the Alicante province (Comunidad Valenciana) and Mallorca (Islas Baleares) in Mediterranean coastal areas of Spain.
Culture Notes: Diaporthe mediterranea was collected from P. dulcis in Spain. The BLASTn search showed 100% identity with the available sequences (ITS and tub) of two isolates named Phomopsis sp. 5 (PMM 1657 and PMM 1660), collected from Vitis vinifera in South Africa [41,42], which were not described as new species by any of the authors. Nevertheless, other loci are needed to better resolve the identity of these isolates. Phylogenetic analysis combining five gene loci showed that all the isolates of D. mediterranea clustered together in a highly supported clade (92/98/1) and displayed a close relationship but they were clearly differentiated from D. sterilis. Based on alignments of the separate loci, D. mediterranea differs from D. sterilis [43] in seven positions (6 nt and one indel of 1 nt) of 426 bp in tub (p-distance = 1.4%), 20 positions (4 nt and one indel of 16 nt) of 342 bp in tef1-α (p-distance = 1.5%), 21 nt of 434 bp in his (p-distance = 4.8%), and 3 nt of 469 bp in cal (p-distance = 0.6%). The ITS sequences of both species showed 100% identity. Morphologically, D. mediterranea mainly differs from D. sterilis in its capacity to produce alpha conidia, because all isolates representing D. sterilis could not be induced to sporulate on any of the culture media used by Lombard et al. [43], when this new Diaporthe species collected from Vaccinium corymbosum was described.

Pathogenicity Tests
All Diaporthe isolates inoculated on one-year-old twigs of almond cv. Vayro caused necrotic lesions of variable length (Figure 4). There was no effect of the experiment on the lesion length (p = 0.5032). Mean lesion length in canes inoculated with different Diaporthe isolates (n = 12 per inoculated isolate) ranged from 1.4 to 13.7 cm and control twigs treated with uncolonized PDA plugs showed a mean lesion length of 0.6 cm ( Figure 5). Statistical analysis revealed significant differences in lesion length between the control and twigs inoculated with all isolates, except those of D. foeniculina, namely DAL-27 and DAL-61 (p = 0.7224 and p = 0.0117, respectively) and D. phaseolorum DAL-222 (p = 0.0239). Notes: Diaporthe mediterranea was collected from P. dulcis in Spain. The BLASTn search showed 100% identity with the available sequences (ITS and tub) of two isolates named Phomopsis sp. 5 (PMM 1657 and PMM 1660), collected from Vitis vinifera in South Africa [41,42], which were not described as new species by any of the authors. Nevertheless, other loci are needed to better resolve the identity of these isolates. Phylogenetic analysis combining five gene loci showed that all the isolates of D. mediterranea clustered together in a highly supported clade (92/98/1) and displayed a close relationship but they were clearly differentiated from D. sterilis. Based on alignments of the separate loci, D. mediterranea differs from D. sterilis [43] in seven positions (6 nt and one indel of 1 nt) of 426 bp in tub (p-distance = 1.4%), 20 positions (4 nt and one indel of 16 nt) of 342 bp in tef1-α (p-distance = 1.5%), 21 nt of 434 bp in his (p-distance = 4.8%), and 3 nt of 469 bp in cal (p-distance = 0.6%). The ITS sequences of both species showed 100% identity. Morphologically, D. mediterranea mainly differs from D. sterilis in its capacity to produce alpha conidia, because all isolates representing D. sterilis could not be induced to sporulate on any of the culture media used by Lombard et al. [43], when this new Diaporthe species collected from Vaccinium corymbosum was described.

Pathogenicity Tests
All Diaporthe isolates inoculated on one-year-old twigs of almond cv. Vayro caused necrotic lesions of variable length (Figure 4). There was no effect of the experiment on the lesion length (p = 0.5032). Mean lesion length in canes inoculated with different Diaporthe isolates (n = 12 per inoculated isolate) ranged from 1.4 to 13.7 cm and control twigs treated with uncolonized PDA plugs showed a mean lesion length of 0.6 cm ( Figure 5). Statistical analysis revealed significant differences in lesion length between the control and twigs inoculated with all isolates, except those of D. foeniculina, namely DAL-27 and DAL-61 (p = 0.7224 and p = 0.0117, respectively) and D. phaseolorum DAL-222 (p = 0.0239).   When isolates of the different Diaporthe species were grouped, significant differences in mean lesion length (cm) were also observed (p < 0.01). Twigs inoculated with D. mediterranea showed significantly longer mean lesions (11.3 cm) compared with D. amygdali. (Figure 6). There were no statistical differences among mean lesion length values caused by D. amygdali (7.7 cm), D. eres (8.4 cm) or D. phaseolorum (6.2 cm). However, twigs inoculated with D. foeniculina showed significantly shorter lesions (2.6 cm) compared with the other Diaporthe spp., except for D. phaseolorum.

Discussion
The survey conducted on almond orchards showing twig cankers and shoot blight symptoms in five different regions of Spain from 2005 to 2019 resulted in a collection of 225 Diaporthe isolates, When isolates of the different Diaporthe species were grouped, significant differences in mean lesion length (cm) were also observed (p < 0.01). Twigs inoculated with D. mediterranea showed significantly longer mean lesions (11.3 cm) compared with D. amygdali. (Figure 6). There were no statistical differences among mean lesion length values caused by D. amygdali (7.7 cm), D. eres (8.4 cm) or D. phaseolorum (6.2 cm). However, twigs inoculated with D. foeniculina showed significantly shorter lesions (2.6 cm) compared with the other Diaporthe spp., except for D. phaseolorum. When isolates of the different Diaporthe species were grouped, significant differences in mean lesion length (cm) were also observed (p < 0.01). Twigs inoculated with D. mediterranea showed significantly longer mean lesions (11.3 cm) compared with D. amygdali. (Figure 6). There were no statistical differences among mean lesion length values caused by D. amygdali (7.7 cm), D. eres (8.4 cm) or D. phaseolorum (6.2 cm). However, twigs inoculated with D. foeniculina showed significantly shorter lesions (2.6 cm) compared with the other Diaporthe spp., except for D. phaseolorum.

Discussion
The survey conducted on almond orchards showing twig cankers and shoot blight symptoms in five different regions of Spain from 2005 to 2019 resulted in a collection of 225 Diaporthe isolates,

Discussion
The survey conducted on almond orchards showing twig cankers and shoot blight symptoms in five different regions of Spain from 2005 to 2019 resulted in a collection of 225 Diaporthe isolates, which were used to elucidate the diversity of Diaporthe species associated with this host using both phenotypical data and DNA sequence analyses. This is the first study in which a collection of Diaporthe isolates from almond has been characterized using multilocus DNA sequence analysis with five loci (ITS, tub, tef-1α, cal and his), which has been recommended in previous phylogenetic studies of the genus Diaporthe for species identification and separation [14,17,44]. This analysis allowed the identification of four known Diaporthe species, namely: D. amygdali, D. eres, D. foeniculina and D. phaseolorum. Moreover, it also confirmed that eight isolates represented a novel phylogenetic species, newly described here as D. mediterranea.
Diaporthe amygdali was the most prevalent species, due to the largest number of isolates collected from widely separated almond growing regions in Spain. This fungus has been described on this crop in other Mediterranean countries, such as France [11], Greece [45], Hungary [46], Italy [47], Portugal [8,48] and Tunisia [49], where it is considered the main pathogen associated with twig cankers and shoot blight symptoms. In Mediterranean areas, D. amygdali has also been reported as a damaging agent in other fruit and nut crops, such as apricot [50], peach [9,51] and English walnut [52]. Diaporthe amygdali is also present in other continents, affecting diverse hosts: on almond and peach in the USA [53,54]; grapevine in South Africa [55]; peach in Japan [56]; peach and nectarine in Uruguay [57,58]; and peach, pear and walnut in China [59][60][61].
Regarding the other Diaporthe species found in our study: D. eres was previously reported on P. dulcis in Portugal [8], and D. foeniculina is present on almond in Italy, with one isolate (CBS 171.78) deposited at the Westerdijk Fungal Biodiversity Institute (Utrecht, the Netherlands) [62]. To our knowledge, our study represents the first report of D. phaseolorum on almond.
The isolates described in our work as belonging to the new taxon, D. mediterranea, were found only in two almond-growing regions in Spain: coastal areas of Alicante province (Comunidad Valenciana) and Mallorca (Islas Baleares). It is interesting to note that the ITS and tub sequences of two Diaporthe isolates, namely Phomopsis sp. 5 (PMM 1657 and PMM 1660), which were collected from V. vinifera in South Africa [41,42], showed 100% identity with the ITS and tub sequences of D. mediterranea. Further studies including other loci would be needed to resolve the identity of the South African isolates (PMM 1657 and PMM 1660).
Pathogenicity tests were performed using one-year-old almond twigs, as described by [8], who determined the capacity of Diaporthe spp. isolates from Portugal to cause lesions on this crop. Except for D. foeniculina and D. phaseolorum, all Diaporthe species inoculated to almond twigs cv. Vayro were able to cause lesions significantly different from those developed on the uninoculated controls. The most severe symptoms were detected on almond twigs inoculated with D. mediterranea. Therefore, this study provides novel information about the ability of this species to cause disease on P. dulcis, being more aggressive than the well-known pathogen D. amygdali. Diaporthe eres was also pathogenic to almond, but the incidence of this species and D. phaseolorum in the survey conducted in this study was extremely low, with only one isolate found in each species.
The present study is the first comprehensive attempt to characterize Diaporthe species associated with P. dulcis in Spain, combining morphology and multilocus DNA sequence analysis. Our results confirm D. amygdali as a key pathogen of almonds in Spain. Moreover, the new species D. mediterranea should also be considered as a potentially important causal agent of twig cankers and shoot blight on this crop, according to the high virulence shown in the pathogenicity tests. In Spain, the lack of information regarding the identity of Diaporthe species on almond and their pathogenicity hinders the development of efficient control strategies and the development of resistant varieties. These aspects have been addressed for the first time in this work and will contribute to the development of improved integrated disease management programs against twig canker and shoot blight disease.  : Research funded by the Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), grants RTA2017-00009-C04-01, -02, -03 and -04 and with matching funds from the European Regional Development Fund (ERDF). G. Elena and C. Agustí-Brisach were supported by the Spanish post-doctoral grants "Juan de la Cierva-Formación" and "Juan de la Cierva-Incorporación", respectively. J. Luque and X. Miarnau were partially supported by the CERCA program, Generalitat de Catalunya. D. Gramaje was supported by the Ramón y Cajal program, Spanish Government (RYC-2017-23098).