RESEARCHFree Access icon

Etiology of Septoria Leaf Spot of Pistachio in Southern Spain

    Authors and Affiliations
    • Ana López-Moral1
    • Carlos Agustí-Brisach1
    • María del Carmen Raya1
    • María Lovera2
    • Carlos Trapero1
    • Octavio Arquero2
    • Antonio Trapero1
    1. 1Departamento de Agronomía, Unit of Excellence “María de Maeztu” 2020-23, ETSIAM, Universidad de Córdoba, Campus de Rabanales, 14071 Córdoba, Spain
    2. 2Departamento de Fruticultura Mediterránea, Andalusian Agricultural and Fisheries Research and Training Institute, 14004 Córdoba, Spain


    Septoria leaf spot (SLS) is the most prevalent disease of pistachio (Pistacia vera L.) in Spain. To elucidate its etiology, 22 samples of pistachio leaves showing SLS symptoms were collected mainly from 1993 to 2018 across southern Spain. Affected leaves from terebinth (P. terebinthus) were also collected for comparative purposes. Six Septoria-like isolates were recovered from pistachio leaves. They were identified as S. pistaciarum by sequencing internal transcribed spacers, partial RNA polymerase II second largest subunit locus, and 28S ribosomal RNA genes. The phenotypic characteristics of conidia and colonies were evaluated, confirming the identity of S. pistaciarum. Conidia were solitary, hyaline, and straight to curved. Large differences in length were observed between conidia from leaf samples, with those from terebinth being slightly larger than those from pistachio. Colonies showed slow mycelial growth on potato dextrose agar (PDA). The effect of temperature on conidial germination and mycelial growth was evaluated in vitro on PDA. For both characters, the optimum temperature was approximately 19 to 20°C. Eight culture media were tested, with oatmeal agar and Spezieller Nährstoffarmer agar showing the highest mycelial growth and pistachio leaf agar (PLA) showing the highest sporulation. A specific culture medium integrating lyophilized-powdered pistachio leaves into diluted PDA improved sporulation compared with PLA. Pathogenicity tests were conducted by inoculating detached and in planta pistachio and terebinth leaflets with conidial suspensions. Typical symptoms of SLS and cirri of S. pistaciarum developed at 10 and 21 days after inoculation, respectively, in both hosts. To our knowledge, this is the first report of S. pistaciarum causing SLS in pistachio and terebinth in Spain.

    The pistachio (Pistacia vera L.) crop on the Iberian Peninsula has undergone a major expansion during the last decade, with a total growing surface area of 39,456 ha and 13,106 t of total nut production in 2019 (MAPA 2020). The most important Spanish pistachio-growing areas are located in the central-eastern (Castilla-La Mancha) and southern (Andalucía) areas of the Iberian Peninsula, which represent 76.1 and 9.1%, respectively, of the national pistachio-growing surface. Pistachio is often considered to be the only cultivated and economically important species of the genus Pistacia (Tous and Ferguson 1996). However, it is worth mentioning that terebinth (P. terebinthus L.), a native plant from the western Mediterranean that is abundant in Spain, should also be considered as an economically important Pistacia species because it is commonly used as rootstock for pistachio attributable to its high drought resistance (Gijón et al. 2009).

    Several aerial fungal diseases of Pistacia species have been described in the main pistachio-growing areas worldwide, with Septoria leaf spot (SLS) being considered one of the most important diseases associated with pistachio fruit and leaf spot (Teviotdale et al. 2002). This disease is frequent in all countries in which the pistachio crop is established, and it is considered a prevalent disease in the Mediterranean basin, Middle East, and United States (Crous et al. 2013; Michailides et al. 1995; Teviotdale et al. 2002). In Spain, the disease is also widely distributed on pistachio plantations (López-Moral et al. 2017a), although there are no scientific reports of its occurrence.

    The main typical symptom of SLS is the development of many irregular dark-brown leaf spots (1 to 2 mm in diameter) on the leaf that can cover the entire surface of the leaflets. When the disease progresses, the affected areas become chlorotic and ultimately become brown and necrotic. The pathogen produces clusters of pycnidia, which can be observed as numerous black dots on necrotic lesions. Under high humidity conditions, pycnidia can release mucilaginous white cirri with abundant conidial masses. If conditions of high humidity persist, masses of cottony white mycelium develop over the lesions. Symptoms on petioles and fruit are similar, although fruit infection is not common. Severe outbreaks of the pathogen, associated with periods of high drought, lead to necrosis, premature defoliation, and reduced tree vigor (Teviotdale et al. 2002; Young and Michailides 1989).

    Since Desmaziéres (1842) described Septoria pistaciae Desm. as the causal agent of SLS of pistachio for the first time in northern France, three species belonging to the Septoria Sacc. genus have been traditionally associated with the disease: S. pistaciae, S. pistaciarum Caracc., and S. pistacina Allesch. (Crous et al. 2013). S. pistaciae and S. pistaciarum are widely distributed in the main pistachio-growing regions of the world, and they have been traditionally considered the main causal agents of the disease. S. pistaciarum has been reported in Texas, Arizona (Young and Michailides 1989), New Mexico (French et al. 2009), India (Ahmad et al. 2011), and across eastern Mediterranean and southeast Anatolian regions (Crous et al. 2013). S. pistaciae has been reported in California (Teviotdale et al. 2002) and Egypt (Haggag et al. 2006), and it is widely distributed in Asia and Europe (Crous et al. 2013). However, S. pistacina shows a more limited distribution because it has only been described in Greece, Syria, Turkey and Iran (Crous et al. 2013). Recently, the taxonomy of Septoria-like pathogens associated with SLP of pistachio has been reviewed (Crous et al. 2013). This study highlights the difficulty of identifying these three phylogenetic species well. Based on new insight into the molecular characterization of Septoria-like pathogens of pistachio, four Septoria-like species associated with SLS have been identified: S. pistaciae, S. pistaciarum, Pseudocercospora pistacina (Allesch.) Crous, Quaedvlieg & Sarpkaya and Cylindroseptoria pistaciae Quaedvl., Verkley & Crous. The identity of two species commonly associated with SLS, S. pistaciarum, and S. pistaciae, has been confirmed, and it has been determined that they belong to the S. protearum Viljoen & Crous species complex. S. pistacina clusters into the genus Pseudocercospora Speg. rather than Septoria, as previously believed, and it has been renamed Pseudocercospora pistacina. Finally, a fourth species, C. pistaciae, has been identified to be associated with SLS of P. lentiscus L. in Mallorca (Balearian Islands, Spain) (Crous et al. 2013).

    Although SLS and other serious foliar and fruit diseases of pistachio have been observed to be widely distributed in the main pistachio-growing regions across Spain, pistachio diseases have not been a major subject of study in this country or even throughout the Mediterranean basin. The lack of knowledge on this topic could exist because pistachios have been considered a secondary crop in Spain until recently. However, the great expansion of this crop in Spain makes studies of the major diseases of pistachio necessary to prevent serious infections. Most of the knowledge of pistachio diseases in Spain has been obtained from surveys conducted during the last two decades (López-Moral et al. 2017a, 2018a). However, few specific studies on Pistacia diseases have been conducted during the last few years. Likewise, in 2018, Alternaria alternata was reported for the first time in Spain to cause leaf blight in terebinth (López-Moral et al. 2018b). More recently, the occurrence of a wide diversity of fungal trunk pathogens has also been reported to be associated with branch dieback and panicle and shoot blight of pistachio in southern Spain (López-Moral et al. 2020). However, even though SLS is considered the most prevalent pistachio disease in southern Spain, its etiology has not yet been studied in this geographic area. Thus, the aim of this study was to elucidate the etiology of SLS of pistachio in southern Spain. To this end, the following specific objectives of the study were to: (i) identify by molecular tools the Septoria-like isolates associated with the disease; (ii) perform the morphological characterization of a large collection of conidia of Septoria isolates from affected leaves of pistachio and terebinth hosts, as well as those grown in culture media; (iii) characterize the effect of temperature on conidial germination and mycelial growth of Septoria isolates; (iv) evaluate the effect of culture media on mycelial growth and conidial production of Septoria isolates; and (v) evaluate their pathogenicity to pistachio and terebinth hosts under controlled conditions.

    Material and Methods

    Field surveys and fungal isolation.

    Between 1993 and 2017, 22 pistachio leaf samples showing typical symptoms of SLS were collected from seven pistachio orchards, with ages ranging from 5 to 25 years (Table 1). Most of them were collected from commercial orchards located in the Andalusia region of southern Spain (provinces of Córdoba and Málaga), with the exception of those collected in 1993, which came from an experimental pistachio field belonging to the current Andalusian Agricultural and Fisheries Research and Training Institute (IFAPA in Spanish, Córdoba). Additionally, six leaf samples showing SLS were collected from wild terebinth (P. terebinthus) to also determine the etiology of the disease in this host (Table 1). Samples included leaves showing circular necrotic black spots (from 2 to 3 mm in diameter) on the upper leaf and irregular dark-brown to black necrotic lesions on the underside of the leaf, with cirrus development being common. All samples were collected from affected fields, placed immediately in black plastic bags, and kept at 4°C until microscopy observations and fungal isolation were conducted.

    Table 1. Leaf samples of Pistacia vera and P. terebinthus showing Septoria leaf spot symptoms collected from 1993 to 2018

    Once in the laboratory, the leaves were carefully observed under a stereomicroscope to find cirri and conidia on the affected leaf surfaces. These fungal structures were plated on potato dextrose agar (PDA) (Difco Laboratories, Detroit, MI) acidified with lactic acid (2.5 ml of 25% [vol/vol] per liter of medium) to minimize bacterial growth. Petri dishes were incubated at 23 ± 2°C under a 12-h diurnal photoperiod of cool fluorescent light (350 μmol m−2 s−1) for at least 21 days until the fungal colonies of Septoria spp. were large enough to be examined, and hyphal tips from the margin of the colonies were transferred to PDA and incubated as described above to obtain pure cultures.

    We were able to recover a total of six Septoria-like isolates (PV-422, PV-423, PV-424, PV-659, PV-660, and PV-661) representative of different geographical origins across the Andalusia region, which were characterized molecularly before conducting additional analysis in this study (Tables 1 and 2). Single-conidium isolates were obtained following the serial dilution method described by Dhingra and Sinclair (1995), and they were maintained in the fungal collection of the Agronomy Department at the University of Córdoba, Spain.

    Table 2. Fungal isolates used in the phylogenetic analysis and their corresponding GenBank accession numbers

    Molecular characterization.

    DNA extraction.

    Genomic DNA was obtained from the active mycelium of the six fungal isolates of Septoria spp. selected in this study, growing onto PDA (Table 2). Mycelial tissues were ground using a FastPrep-24 grinder machine (MP Biomedicals, Irvine, CA, U.S.A.), and DNA was extracted with an E.Z.N.A Fungal DNA Kit (Omega Bio-tek, Norcross, GA, U.S.A.). The concentration and purity of the extracted DNA were measured with a MaestroNano spectrophotometer (MaestroGen, Hsinchu, Taiwan).

    PCR analysis and sequencing.

    The 5.8S nuclear ribosomal with two flanking internal transcribed spacers (ITS), the partial RNA polymerase II second largest subunit locus (RPB2), and the 28S ribosomal RNA (LSU) genes were amplified with the primer pairs ITS4/ITS5 (White et al. 1990), fRPB2-5F (Liu et al. 1999) and fRPB2-414R (Quaedvlieg et al. 2011), and LSU1Fd (Crous et al. 2009a) and LR5 (Vilgalys and Hester 1990), respectively.

    PCR was performed in a total volume of 25 µl containing the following mixture: 20 ng of genomic DNA, 5 µl of 5× My Taq Reaction Buffer, and 0.13 µl of My Taq DNA Polymerase (Bioline, Santa María del Águila, Almería, Spain). Additionally, each primer was used at 0.4 μM for ITS PCR, whereas for the RPB2 and LSU region PCR, each primer was used at 0.6 or 0.2 μM, respectively. A negative control was included in all PCR runs, using ultrapure water instead of DNA. The PCR cycling programs were as follows: for ITS, initial denaturation at 94°C for 3 min, followed by 30 cycles of 94°C for 30 s, 48°C for 30 s, and 72°C for 1 min, and a final extension at 72°C for 10 min; for RPBS and LSU, initial denaturation at 95°C for 3 min, followed by 35 cycles of 94°C for 15 s, 50°C (for RPB2) or 58°C (for LSU) for 15 s, and 72°C for 1 min, and a final extension at 72°C for 7 min. The PCR cycling program for RPB2 amplification was adjusted according to Hou et al. (2020). All PCR runs were carried out in a MyCycle Thermal Cycler (Bio-Rad, Hercules, CA, U.S.A.). Amplification products were checked by electrophoresis in a 1.5% (wt/vol) agarose gel stained with RedSafe (Intron Biotechnology) and visualized under ultraviolet light. A 100-bp DNA Ladder-GTP (gTPbio) was used as a molecular weight marker. Subsequently, the PCR products were purified using a MEGAquick-spin Total Fragment DNA Purification kit (Intron Biotechnology, Gyeonggi-do, Korea) following the manufacturer’s instructions. The resulting amplicons were sequenced in both directions by the Central Service Support Research of the University of Córdoba (Spain).

    Phylogenetic analysis.

    DNA sequences were generated with forward and reverse primers by means of SeqMan software (Lasergen SeqMan v. 7.0.0, DNASTAR, Madison, WI). Subsequently, consensus sequences were obtained and compiled into a single FASTA format file. All consensus sequences were subjected to BLAST searches against the NCBI GenBank nucleotide database to determine the most closely related species of Septoria spp. to our six fungal isolates, which were included in the molecular phylogenetic analyses. Additionally, sequences of Pseudocercospora pistacina, S. pistaciae, and S. pistaciarum, among other Septoria species, were included in the analysis as reference sequences, and Zymoseptoria verkleyi CBS 1336761 was included as an outgroup (Table 2). Reference sequences were selected based on their high similarity to our query sequences using MegaBLAST, downloaded from GenBank (, added to the data set, and aligned using CLUSTAL W v. 2.0.11 (Larkin et al. 2007).

    Neighbor-joining (NJ) analysis was performed individually for each locus (data not shown) using the maximum composite likelihood method and 2,000 bootstrap replications to determine whether the sequence datasets were congruent and combinable. Tree topologies of 70% reciprocal bootstrap generated individually for each locus were compared visually. Because no supported nodes were in conflict, the data of different loci were combined into single concatenated datasets. Therefore, a multilocus alignment of ITS and LSU was performed to identify our isolates of Septoria spp. by means of maximum parsimony (MP) and Bayesian inference (BI). The MP trees were obtained using the tree-bisection-regrafting algorithm with search level 1, in which the initial trees were obtained by the random addition of sequences (10 replicates). All positions containing gaps and missing data were eliminated. The robustness of the trees obtained was evaluated by 2,000 bootstrap replications. Tree length (TL), consistency index (CI), retention index (RI), rescaled consistency index (RC), and homoplasy index (HI) were recorded. BI analyses were performed with MrBayes v.3.2.6 (Ronquist et al. 2012), which uses the Markov chain Monte Carlo method to approximate the posterior probability of trees. Two analyses with four chains each were run at the same time for 1 × 107 generations, sampled every 100 generations and starting from a random tree topology. The “temperature” parameter was set to 0.2. For the consensus tree, the first 25% of the saved trees were discarded as the burn-in phase of the analysis. Each of the individual genes as well as a combined data set were aligned, adjusted manually if necessary, and analyzed by NJ or MP using MEGA v.7 (Kumar et al. 2016). In BI and NJ analyses, the best evolutionary model for each gene partition was also determined by MEGA v.7. The genes were concatenated in a single nucleotide alignment using Phylogenetic Data Editor. Sequences derived in this study were submitted to GenBank, and their GenBank accession numbers are shown in Table 2.

    Phenotypic characterization.

    Conidial morphology from leaf samples.

    The shape, size, color, and number of septa were described and measured directly by microscopy observations of conidia collected from the lesions on the underside of the affected leaves of P. vera or P. terebinthus because of the difficulty of inducing conidial production in vitro. For each sample (Table 2), 10 symptomatic leaves were carefully observed, lesions showing pycnidia (fruiting body) or cirri were scraped with a sterile scalpel, and the removed fungal structures were placed directly on slides with a 10-µl drop of 0.01% acid fuchsine in lactoglycerol (1:2:1 lactic acid-glycerol-water) and covered with a coverslip. Fungal structures were measured at ×400 magnification using an Eclipse 80i microscope (Nikon Corp., Tokyo, Japan). The length and width of 30 conidia were measured per sample, and the numbers of septa per conidia were recorded. Their averages and the length/width (L/W) average were calculated.

    Colony and conidial morphology of Septoria isolates on culture media.

    The characteristics of the mycelia (texture, density, color, and zonation) of the six Septoria isolates selected in this study (PV-422, -423, -424, -659, -660, -661; Table 2) were recorded after 21 days of growth on PDA at 23 ± 2°C in darkness, and color was determined by means of Rayner’s (1972) color scale. In addition, the six isolates were also grown on PLA (Agustí-Brisach et al. 2019; Chen et al. 2014) as described previously to induce sporulation. Conidial masses were deposited on slides with acid fuchsine and measured as described above. Thirty conidia per isolate were measured.

    Effect of temperature on conidial germination and mycelial growth.

    Conidial germination.

    Conidial suspensions were obtained from 42-day-old colonies of the six Septoria isolates (Table 2) growing on PLA at 23 ± 2°C under a 12-h diurnal photoperiod of cool fluorescent light (350 μmol m−2 s−1) for 42 days. For each isolate, two mycelial plugs (7.5 mm in diameter) were cut from the margin of the colony, placed in a sterile glass vial containing 9 ml of sterile distilled water (SDW), and vortexed for 5 s. Subsequently, the conidial suspensions of each isolate were adjusted to 2 × 105 conidia ml−1 using a hematocytometer, and a 10-μl drop of conidial suspension of each isolate was placed in the center of a microscope coverslip (20 × 20 mm). Coverslips were placed inside Petri dishes containing water agar, which were used as humid chambers, and incubated at 5, 10, 15, 20, 25, 30, and 35°C in darkness for 48 h. Then, a 5-µl drop of 0.01% acid fuchsine in lactoglycerol (1:2:1 lactic acid-glycerol-water) was added to each coverslip to stop conidial germination; they were then mounted on a slide. The percentage of germination was determined by observing 100 conidia randomly selected per coverslip at ×400 magnification using an Eclipse 80i microscope (Nikon Corp.), with germinated and nongerminated conidia being counted. Conidia were considered germinated when the germ tube was at least one-half of the longitudinal axis of the conidia. There were three replicate coverslips per fungal isolate and temperature combination, arranged in a completely randomized design. The experiment was conducted twice.

    Mycelial growth.

    The six Septoria isolates (Table 2) were grown on PDA for 42 days as described above. Mycelial plugs (7.5-mm diameter) obtained from the margins of actively growing colonies were placed in the center of Petri dishes filled with PDA and were incubated at 0, 5, 10, 15, 20, 25, 30, and 35°C in darkness for 21 days. Subsequently, the largest and smallest diameters of the colonies were measured, and mean data were converted to radial growth rate (mm day−1). There were four replicate Petri dishes per isolate and temperature combination, arranged in a completely randomized design. The experiment was conducted twice.

    Fungal growth and conidial production on culture media.

    Trial I.

    The following eight culture media were evaluated: (i) Czapek Dox agar (CDA, Difco Laboratories, Detroit, MI); (ii) pistachio leaf agar (PLA; 10 g of PDA, 10 g of agar [Rokoagar AF Lab, Roko Industries, Llanera, Asturias, Spain]), 1 liter of distilled water, one twice-sterilized pistachio leaf per Petri dish (Agustí-Brisach et al. 2019; Chen et al. 2014); (iii) malt extract agar (MEA) (20 g of MEA; Merck KGaA, Darrmstadt, Germany), 20 g of agar (1 liter of SDW) supplemented with 0.5 g l−1 of streptomycin sulfate (Sigma-Aldrich, St. Louis, MO) (MEAS); (iv) oatmeal agar (OA, Crous et al. 2009b); (v) PDA; (vi) Spezieller Nährstoffarmer agar (SNA) (Leslie and Summerell 2006); (vii) SNA with a 1 × 1-cm piece of filter paper added to the agar surface (SNAp); and (viii) yeast malt dextrose agar (YMDA, Seifbarghi et al. 2009). Most of these culture media were selected on the basis of favorable results previously obtained in other species or genera related to Septoria (Crous et al. 2013; Dhingra and Sinclair 1995; Seifbarghi et al. 2009).

    Trial II.

    The PLA evaluated in Trial I was slightly modified by integrating different amounts of powdered lyophilized pistachio leaves in diluted PDA instead of placing one twice-sterilized pistachio leaf on the culture surface. For this purpose, healthy pistachio leaves were collected from experimental fields (IFAPA-Alameda del Obispo, Córdoba, Andalusia Region, Spain), washed under running tap water for 5 min, air dried on a laboratory bench for 30 min, and frozen at −20°C for at least 24 h. Subsequently, frozen leaves were lyophilized at −43°C, ∼0.344 mBar (Telstar Cryodos, Azbil Group, Terrassa, Spain) for 14 days. After lyophilization, leaves were crushed using liquid nitrogen. The weight of the leaf powder from one leaf (n = 30) was estimated (average = 0.358 g of powder per leaf). To prepare the culture media, the corresponding amounts of pistachio leaf powder were added to Erlenmeyer flasks containing diluted PDA (as described in Trial I) to obtain Petri dishes filled with diluted PDA with one (PLA-1), two (PLA-2), three (PLA-3), or four (PLA-4) powdered leaves in each dish and sterilized at 120°C for 20 min. Additionally, PLA and the diluted PDA used to prepare PLA without pistachio leaf powder were also evaluated as controls in this trial.

    Medium inoculation, experimental design, and assessment.

    Mycelial plugs (7.5 mm in diameter) obtained from 42-day-old colonies of S. pistaciarum isolate PV-422 (Table 2) grown on PDA as described above were plated into sterile Petri dishes containing the different culture media tested in Trial I or Trial II. In each trial, there were four replicated Petri dishes per culture medium, arranged in a completely randomized design, and incubated as described above for 42 and 21 days for Trials I and II, respectively. Both experiments were conducted twice.

    The effect of the culture media on mycelial growth was assessed only in Trial I by measuring the diameter of each colony twice perpendicularly at 42 days of incubation, and the mean data were converted to radial growth rate (mm day−1). Conidial production was assessed using two mycelial plugs (7.5 mm in diameter) from 42- and 21-day-old colonies of PV-422 for Trials I and II, respectively. Both mycelial plugs were placed in a sterile glass vial containing 9 ml of SDW, vials with plugs were vortexed for 5 s, and the number of conidia per milliliter was counted using a hemocytometer. One vial per replicated Petri dish of each culture medium was obtained. A 10-µl drop of the conidial suspension of each culture medium was deposited directly on slides with a 5-µl drop of 0.01% acid fuchsine in lactoglycerol (1:2:1 lactic acid-glycerol-water) and covered with a coverslip. Conidia were measured at 400× magnification using an Eclipse 80i microscope (Nikon Corp.). The length, width, and number of septa from 30 conidia were measured in each replicated Petri dish of each culture medium, and their average values and average L/W were calculated.

    Pathogenicity tests.

    Fungal isolate and inoculum preparation.

    Inoculations were performed using the Septoria isolate PV-422 (Table 2). Because of the difficulty of producing conidia on PDA, the fungal isolate was grown on PLA at 23 ± 2°C under continuous fluorescent light for 42 days. Then, a conidial suspension was obtained by scraping the mycelial and pycnidia masses of the pathogen developed on the surface of PLA. It was filtered through sterile cheesecloth and adjusted with a hemocytometer to 5 × 105 conidia ml−1.

    Inoculation of detached leaflets.

    In May 2018, fresh asymptomatic detached leaflets of pistachio of cultivar Kerman were collected from a 10-year-old experimental orchard belonging to IFAPA (Alameda del Obispo, Córdoba province). Additionally, fresh asymptomatic detached leaflets of terebinths were collected from a natural forest in Córdoba. Leaflets were washed with 0.02% Tween-20 solution in tap water for 1 min, surface sterilized by dipping them in 10% solution of commercial bleach (chlorine at 50 g liter−1) for 1 min, and double washed with tap water. Subsequently, they were disinfected with 70% ethanol solution for 1 min, and the leaflets were air dried on sterile filter paper for 30 min. Disinfected unwounded leaflets were placed in humid chambers (56 × 18 × 41-cm plastic containers) and inoculated by spraying them until the run-off point with the conidial suspension of S. pistaciarum isolate PV-422. In addition, noninoculated leaves sprayed with SDW were included as controls. After inoculation, the humid chambers were incubated at 21 ± 2°C under 100% relative humidity (RH) in darkness until symptom onset. A completely randomized block design with three replicate humid chambers (block) per host and treatment (inoculated and noninoculated control) was used. Twelve leaflets per humid chamber and per inoculated host or control were used (72 pistachio leaflets and 72 terebinth leaflets in total). The experiment was conducted twice.

    Inoculation of potted plants.

    In June 2018, 2- to 3-year-old healthy potted plants of pistachio cultivar Kerman and healthy potted seedlings of wild terebinths were obtained from a commercial nursery. Inoculations were conducted by spraying the entire potted plants until the run-off point with a conidial suspension of S. pistaciarum. In addition, potted plants sprayed with SDW were included as controls. Immediately after inoculation, both inoculated and control plants were incubated at 21 ± 2°C and 100% RH in darkness for 72 h. After this first period of incubation, plants were maintained in a climatic room at 21 ± 2°C and 70% RH with a 12-h photoperiod until symptom onset. There were six replicate plants per host and treatment (inoculated and noninoculated control) combination, arranged in a completely randomized design. The experiment was conducted twice.

    Data analyses.

    All experiments were conducted twice, and data from the two experiments were combined after checking for homogeneity of the experimental error variances by the F test (P ≥ 0.05). In all cases, data were tested for normality and homogeneity of variances. Square root transformation of the data was conducted when necessary. In the conidial phenotyping experiment, conidial length, width, and L/W data were subjected to analysis of variance (ANOVA) to determine the differences between isolates in conidium morphology. Finally, to determine the effect of temperature on conidial germination and mycelial growth experiments, a nonlinear adjustment of the data was performed using the generalized Analytics Beta model (Hau and Kranz 1990) to evaluate the variation of the conidial germination percentage or the mycelial growth rate, respectively, for each isolate over temperature (López-Moral et al. 2017b). The average germination percentages or growth rates for each temperature were adjusted to a regression curve to estimate their optimum growth temperature and the maximum germination (MGeP) percentage or the maximum growth rate (MGR; mm day−1), respectively, for each isolate. Subsequently, optimum growth temperature, MGeP, or MGR data were subjected to ANOVA to determine the differences between isolates. In the culture media assay (Trials I and II), the dependent variable radial growth rate, conidial production, length and width of conidia, and/or L/W were subjected to ANOVA to determine the differences between culture media. Additionally, correlation and regression analyses were performed for the Trial II data, excluding PLA values. After these analyses, mean conidial production values were fitted to a regression curve over the pistachio leaf powder concentration (PLPC; g of leaf powder per Petri dish with diluted PDA). Additionally, the Pearson correlation coefficient (r) between the PLPC and conidial production, length of conidia, width of conidia, or L/W was calculated using the average of each dependent variable for each PLPC evaluated (n = 5). Mean values were compared using Fisher’s protected least significant difference test for experiments with independent variables with less than six levels or Tukey’s honestly significant difference tests for experiments with independent variables with six or more levels, both at α = 0.05 (Steel and Torrie 1985). All data from this study were analyzed using Statistix 10 (Analytical Software 2013).


    Field surveys and fungal isolation.

    Symptoms of SLS were observed in all of the surveyed pistachio orchards, but disease incidence was highly variable among orchards, years, and the use of fungicide treatments. The percentage of symptomatic leaves ranged from <1 to >75%. The two most affected orchards, with a percentage of diseased leaves >75%, were those of Puente Genil-2008 and Montilla-2018, which did not receive any fungicide treatment, whereas the remaining fields, which did receive fungicide treatments, showed an average incidence of infected leaves of <10%. Circular, dark-brown necrotic lesions from 1 to 3 mm in diameter were consistently found on both the upper and underside surfaces of the leaf (Fig. 1A to C). Necrotic areas were observed surrounding the spots when disease progressed, and the affected leaflets finally fell on the orchard floor. Similar spots were also observed on fruit, which became necrotic (Fig. 1D). Cirri formation was frequent on the underside leaf surface (Fig. 1E). Premature leaf fall was observed with a high level of defoliation mainly in years with high humidity and moderate temperatures and in orchards not treated with fungicides, in which important outbreaks occurred (Fig. 1F). Similar symptoms of SLS were observed in wild terebinth bushes, but no symptoms were observed in P. lentiscus bushes that were also common in the surveyed areas.

    Fig. 1.

    Fig. 1. Symptoms of Septoria leaf spot of pistachio. A to C, Circular, dark-brown necrotic lesions 1 to 3 mm in diameter on the leaf surface (red arrows); D, circular black spots on fruit (red arrow), and fruit becoming necrotic (blue arrow); E, details of cirrus formation (red arrow) on the underside surface of the leaf detail; E to F, commercial pistachio orchard (Córdoba province, Andalusia region, southern Spain) seriously affected by Septoria leaf spot showing a high level of defoliation.

    Download as PowerPoint

    Fungal structures (pycnidia and conidia) of Septoria spp. were found on both leaf surfaces in all samples of affected leaves collected in the surveyed pistachio orchards as well as in all samples of P. terebinthus. Pathogen detection was performed either directly on the leaves via conidial observations when the samples were collected in wet weather or after incubation of the leaves at room temperature in a humid chamber (100% RH) for 24 to 48 h. Isolation of the pathogen in culture media from the symptomatic leaf tissues was difficult and inconsistent because of the slow growth of Septoria spp. Frequently, the putative colonies of the pathogen were covered by several unidentified saprophytic fungi (i.e., Alternaria spp., Aureobasidium spp.) and bacteria (i.e., Bacillus spp.). Thus, the best time to obtain pure cultures of Septoria spp. was during wet weather periods in middle-late spring when cirri developed on the underside surface of the leaves. Directly plating cirrus structures on PDA helped in obtaining isolates of the pathogen characterized in this study.

    Molecular characterization.

    The phylogenetic analysis contained 28 taxa, including the outgroup Zy. verkleyi CBS 1336761. The gene boundaries were ITS, 1 to 487; RPB2, 488 to 814; and LSU, 815 to 1,521. All positions containing gaps and missing data were eliminated when MP analysis was run. There were 1,494 positions in the final dataset, of which 193 characters were parsimony informative, 214 were singleton sites, and 1,087 were conserved sites. The MP analysis showed the five most parsimonious trees (TL = 657; CI = 0.663, RI = 0.833, HI = 0.337, RC = 0.552), and one of those is shown in Figure 2. For BI analysis, K2+G models were used for the ITS, RPB2, and LSU regions. The tree obtained by BI analysis confirmed the topology obtained with MP. The six isolates used in this study clustered together (percent bootstrap support [BS; %]/Bayesian posterior probabilities [PP] = 100/0.9) with GenBank reference sequences of S. pistaciarum (Fig. 2; Table 2).

    Fig. 2.

    Fig. 2. One of the two most parsimonious trees (TL = 657; CI = 0.663; RI = 0.833; HI = 0.337; RC = 0.552) obtained from a heuristic search of the combined ITS, RPB2, and LSU sequences alignment of fungal species belonging to Septoria spp. The evolutionary history was inferred using the maximum parsimony (MP) method. Bootstrap support values (MP, >760%) and Bayesian posterior probabilities (>0.8) are shown at the nodes. Zymoseptoria verkleyi isolate CBS 1336761 was used as the outgroup.

    Download as PowerPoint

    Phenotypic characterization.

    Pycnidia and conidia morphology from leaf samples.

    Erumpent, brown, globose pycnidia were observed on infected leaves. They showed a central ostiole, exuding crystalline white cirri of conidia. In general, the conidia of S. pistaciarum from leaves of Pistacia spp. were solitary, hyaline, smooth, guttulate, straight to curved, 0 to 4-septate, apex subobtuse, truncated in the base, (25.0–)49.7(–57.5) × (2.1–)2.5(–2.8) μm, L/W = 19.8 ± 0.16. Conidia from terebinths were slightly larger (average length = 53.2 ± 0.91 μm) than those from pistachio (average length = 48.9 ± 0.61 μm) (Table 1).

    Colony, pycnidia, and conidial morphology of Septoria isolates in culture media.

    Colonies of the six isolates of S. pistaciarum (PV-422, -423, -424, -659, -660, -661) showed a very slow mycelial growth rate on PDA, reaching no more than 7.5 mm in diameter after 21 days of incubation at 23 ± 2°C in darkness. They were folded with lobate margins, surface olivaceous gray to brown with patches of beige to dirty white, reverse olivaceous gray to brown (Fig. 3A to F). On PLA, colonies were mostly beige to white with patches of light to dark-brown or olivaceous gray, felped mycelia, and lobulate margins (Fig. 3G). Pycnidia were induced on PLA. Pycnidia and conidia from colonies growing on PLA showed the same phenotype as that observed for leaf samples, and significant differences in conidial measures (P ≤ 0.0001 in any cases) were observed between isolates, with S. pistaciarum isolate PV-659 showing the largest conidia (length = 54.6 ± 1.5 μm; n = 30) and S. pistaciarum isolate PV-661 showing the shortest conidia (length = 37.8 ± 1.9 μm; n = 30). The mean of the conidial measures for the six fungal isolates was (22.0–) 44.5 (–65.0) × (1.83–) 3.1 (–4.8) μm, L/W = 15.0 ± 0.35.

    Fig. 3.

    Fig. 3. A to F, Twenty-one-day-old colonies of Septoria pistaciarum isolates PV-422, PV-423, PV-424, PV-659, PV-660, and PV-661, respectively, from P. vera growing on potato dextrose agar at 23°C in darkness; G, 43-day-old colonies of S. pistaciarum isolate PV-422 from Pistacia vera growing on pistachio leaf agar at 23°C with a 12-h photoperiod of fluorescent light (350 mmol m−2 s−1).

    Download as PowerPoint

    Effect of temperature on conidial germination and mycelial growth.

    Conidia of the six isolates of S. pistaciarum germinated between 10 and 30°C. Exceptionally, conidia of isolate PV-424 were able to germinate at the maximum temperature tested (35°C), although with a low germination percentage (1.1%). No germination of conidia was observed at 5°C in any cases. Significant differences in optimum germination temperature and MGeP (P < 0.0001 in any cases) were observed among isolates. The optimum germination temperature of S. pistaciarum ranged from 18.0 to 25.7°C for isolates PV-659 and PV-423, respectively. Four of the six isolates showed high conidial germination, with MGeP values ranging between 84.7 and 100% for isolates PV-659 and PV-422, respectively. However, isolates PV-424 and PV-661 showed low conidial germination, with MGeP values of 57.6 and 24.2%, respectively (Table 3; Supplementary Fig. S1).

    Table 3. Effect of temperature on conidial germination and mycelial growth of representative isolates of Septoria pistaciarum grown on potato dextrose agar at 0, 5, 10, 15, 20, 25, 30, and 35°C in darkness for 21 days

    The six isolates of S. pistaciarum were able to grow from the minimum temperature tested (5°) to 30°C. According to the Analytics Beta model, only isolate PV-659 was able to grow at the maximum temperature tested (35°C). Significant differences in the optimum growth temperature (P = 0.0005) and MGR (P < 0.0001) were observed among the isolates. The optimum growth temperature of S. pistaciarum was approximately 22°C because it ranged between 20.8 and 24.1°C for isolates PV-424, PV-661, and PV-423, respectively. As we described above, all isolates showed very low mycelial growth rates, which ranged between 0.27 and 0.5 mm day−1 for isolates PV-422 and PV-660, respectively (Table 3; Supplementary Fig. S2).

    Fungal growth and conidial production on culture media.

    Trial I.

    There were significant differences in mycelial growth rate (P ≤ 0.0001) between culture media. The highest growth rates were observed on OA (0.66 mm day−1) and on SNA (0.65 mm day−1), whereas the lowest growth rate was observed on PLA (0.26 mm day−1). Intermediate values of growth rate were observed for SNAp, PDA, YMDA, MEAS, and CDA, ranging from 0.59 to 0.48 mm day−1 for SNAp and CDA, respectively (Fig. 4).

    Fig. 4.

    Fig. 4. Mycelial growth rate (millimeters per day) of Septoria pistaciarum isolate PV-422 on the different culture media tested in Trial I. Columns represent the mean of two independent experiments of four replicated Petri dishes. Columns with the same letters do not differ significantly according to Tukey’s honestly significant difference test (α = 0.05). Vertical lines on the columns are the standard error of the mean. CDA, Czapek Dox agar; MEAS, malt extract agar with streptomycin sulfate; OA, oatmeal agar; PDA, potato dextrose agar; PLA, pistachio leaf agar; PLA-1, -2, -3, -4, diluted potato dextrose agar with one, two, three, and four powdered pistachio leaves, respectively; SNA, Spezieller Nährstoffarmer agar; SNAp, SNA with sterile filter paper added; YMDA, yeast malt dextrose agar.

    Download as PowerPoint

    All culture media tested were able to induce conidial production of S. pistaciarum isolate PV-422, with the exception of SNA and SNAp. Significant differences between culture media were observed for conidial production, the length of conidia, and L/W (P ≤ 0.0001 in any cases) but not for the width of conidia (P = 0.0528). Contrary to what was observed in mycelial growth, PLA was proven to be the most effective culture medium for inducing the conidial production of S. pistaciarum (3.2 × 106 conidial ml−1). PDA, CDA, YMDA, MEAD, and OA showed intermediate sporulation values, ranging from 2.6 × 105 to 3.4 × 104 conidia ml−1 for PDA and OA, respectively. The length of conidia ranged between 20.5 and 60.7 μm for PDA and OA, respectively, whereas the width of conidia ranged from 3.3 to 4.2 μm for YMDA and PDA, respectively (Table 4). Number of septa varied from 1 to 5.

    Table 4. Conidial production of Septoria pistaciarum isolate PV-422 on different culture media

    Trial II.

    All combinations of PLA, in addition to the diluted PDA used as a control, were able to induce a large amount of conidial production of S. pistaciarum isolate PV-422 at 21 days of incubation. A significant linear trend over PLPC was observed for all dependent variables analyzed, such as conidial production (r = 0.9918, P = 0.0009), the length of conidia, (r = –0.9185, P = 0.0276), the width of conidia (r = 0.9291, P = 0.0224) and L/W (r = –0.9026, P = 0.0360). Conidial production and width increased with increasing concentrations of pistachio leaf powder in diluted PDA, whereas conidial length and L/W decreased as the concentration of pistachio leaf powder increased. Conidial production ranged from 1.3 × 104 to 9.9 × 104 conidia ml−1 for diluted PDA and PLA-4, respectively, whereas PLA showed intermediate conidial production (4.5 × 104 conidia ml−1) compared with those observed for PLA-1 and PLA-2. The length of conidia ranged between 38.7 and 52.8 μm for PLA-4 and diluted PDA (control), respectively, whereas the range of variation in the width of the conidia was very scarce, from 2.0 to 2.3 μm for diluted PDA and the highest concentrations, respectively (Table 4). Number of septa varied from 1 to 4. The regression analysis of conidial production over the PLPC on diluted PDA showed a highly significant adjustment (R2 = 0.9837, P = 0.0009), with the following fitted equation: conidial production = 13,112 + 57,379 × PLPC (Fig. 5).

    Fig. 5.

    Fig. 5. Linear regression between conidial production (conidia per milliliter) of Septoria pistaciarum isolate PV-422 and pistachio leaf powder concentration (PLPC; g of leaf powder per Petri dish with diluted potato dextrose agar). The spots represent the average from the four replicated Petri dishes measured per each PLPC tested in two independent experiments.

    Download as PowerPoint

    Pathogenicity tests.

    Inoculation of detached leaflets.

    The pathogenicity of S. pistaciarum isolate PV-422 in detached pistachio and terebinth leaflets could not be evaluated because all plant tissues were completely colonized by saprophytes (Alternaria spp. mainly) at 3 to 4 days after inoculation in the two repetitions of the experiment. This could be attributable to the slow mycelial growth rate of S. pistaciarum, which requires a long time to cause infection and symptom development. Thus, pathogenicity tests in detached leaflets were not useful for this slow-growing foliar pathogen.

    Inoculation of potted plants.

    Typical symptoms of SLS developed at 10 days after inoculation in both pistachio (Fig. 6A to D) and terebinth (Fig. 6E and F) hosts. All sprayed leaflets of the inoculated plants showed many necrotic spots, whereas the control plants did not show any lesions (Fig. 6). Pycnidia and cirri of the pathogen (Fig. 6D) also developed on the underside of the inoculated leaves of both pistachio and terebinth hosts 21 days after inoculation. The pathogen was successfully reisolated from lesions of both P. vera and P. terebinthus hosts, with the consistency of isolation ranging between 62 and 43%, respectively. Even though these values seem low, they should be considered sufficient for S. pistaciarum because of the difficulty of growing this species on artificial culture media.

    Fig. 6.

    Fig. 6. Typical symptoms of A to D, Septoria leaf spot reproduced in potted plants of Pistacia vera cultivar Kerman and E and F, P. terebinthus at 10 days after inoculation by spraying a conidial suspension of Septoria pistaciarum isolate PV-422 compared with the respective noninoculated controls (arrow); D, cirrus of S. pistaciarum isolate PV-422 developed on the underside of the leaf spots at 21 days after inoculation.

    Download as PowerPoint


    Foliar diseases are among the major limiting factors of pistachio crops in the main growing regions of this nut tree worldwide, and SLS is one of the most important pistachio diseases causing fruit and leaf spot (Crous et al. 2013; Teviotdale et al. 2002). In fact, previous field observations conducted across central and southern Spain reveal that SLS is the most serious and prevalent foliar disease in Spanish pistachio plantings (López-Moral et al. 2017a). Although SLS has been traditionally observed in commercial pistachio orchards in Spain, the etiology of the disease has not yet been studied in this country, probably because pistachios have been considered a secondary crop in Spain until recently. Thus, to our knowledge, this work represents the first study elucidating the etiology of SLS of pistachio in Spain.

    The surveys conducted from central to southern Spain from 1993 to 2018 allowed us to characterize conidia of Septoria-like isolates from 28 leaf samples of pistachios or terebinths showing SLS symptoms from different geographic origins. According to the current literature on the taxonomy of Septoria-like species causing leaf and fruit spot in pistachio (Crous et al. 2013), the phenotypic characteristics of the conidia of the 28 fungal isolates collected from leaf samples were more than sufficient to identify all specimens as S. pistaciarum. This was possible because S. pistaciarum is morphologically distinct from other Septoria-like species associated with the fruit and leaf spot of pistachios, mainly because they produce larger conidia (Crous et al. 2013). However, despite the difficulty of isolating this pathogen on artificial culture media, we were able to recover six pure fungal isolates of S. pistaciarum from the lesions of affected pistachio leaves, which were used to confirm the identity of this fungal species by molecular tools. Likewise, the multilocus combined alignment of ITS, RPB2, and LSU sequences was useful to confirm the identification of all six fungal isolates as S. pistaciarum because they all clustered together with the reference sequence used in the phylogenetic analysis (BS = 100%; PP = 0.9). It was conducted based on Crous et al. (2013), who suggested this combined multilocus alignment as the best option for the molecular identification of Septoria isolates. Subsequently, the molecular identification was also validated by comparing the phenotypic characteristics of the colonies and conidia of our isolates growing on culture media with those reported by Crous et al. (2013). Indeed, their colony, pycnidial, and conidial characteristics (both from leaf samples and from PLA) corresponded closely to those described for S. pistaciarum by Crous et al. (2013), but they were quite different from those described for S. pistacina or Pseudocercopora pistaciae. Moreover, the angular dark-brown spots confined by leaf veins developed by the pathogen in both natural (Fig. 1) and artificial (Fig. 6C and D) infections make S. pistaciarum distinct from other Septoria species that attack pistachios (Crous et al. 2013). For all of these reasons, there is no doubt that all of the Spanish isolates recovered in this study are well identified as S. pistaciarum, being so far the only causal agent associated with SLS of pistachio in Spain.

    The optimum temperature for both the conidial germination and mycelial growth of the S. pistaciarum isolates was approximately 19 to 20°C. Exceptionally, isolate PV-423 showed the highest optimum temperatures for both conidial germination and mycelial growth, reaching 25.7 and 24.1°C, respectively. Only isolate PV-660 showed similarities in the effect of temperature on mycelial growth, with 23.7°C being the optimum growth temperature. Most likely, the differences in the optimum temperatures among isolates may be attributable to their specific genotypes and/or their geographic origin. In this context, additional research is needed using a larger population of S. pistaciarum isolates from a broad diversity of geographic pistachio areas across Spain to make a better approach to this biological aspect of the pathogen. The MGeP was usually high in most of the isolates, with the exception of isolates PV-424 and PV-661, which showed lower germination in vitro. Therefore, these two isolates should be discarded for additional analysis in vitro because of their low conidial viability. In contrast, isolate PV-422 showed 100% MGeP. For this reason, this isolate was selected to conduct the culture media assay and pathogenicity tests in this study because of the exceptional viability of its conidia in vitro. The MGR was no more than 0.35 mm day−1, as expected according to our previous observations on PDA. There is little information about cardinal temperatures of S. pistaciarum for comparison in the literature, with the exception of the studies conducted by Crous et al. (2013), who indicated that colonies of the pathogen reach 30 mm in diameter on PDA after 14 days at 24°C.

    The low mycelial growth rate and low sporulation of S. pistaciarum in vitro make it difficult to isolate and grow in artificial culture media. Consequently, we observed that conducting studies with this pathogen under laboratory-controlled conditions may be tedious. Thus, in the present study, we tested several culture media to evaluate their effect on mycelial growth and conidial production to improve the general management procedures for Septoria pathogens for future laboratory experiments. No important advances were achieved with regard to their effect on mycelial growth because no significant differences were observed between the best culture media, OA or SNA, and PDA, which are usually used as generic culture media for fungal growth. Nevertheless, relevant results were obtained regarding the effect of PLA inducing higher conidial production compared with the remaining culture media tested. In fact, PLA has already been described as a successful culture medium inducing pycnidial production in Botryosphaeriaceae and Diaporthaceae fungi, and it allows high concentrations of conidial suspensions to be obtained (Chen et al. 2014). The effect of this medium on inducing pycnidial and conidial production in S. pistaciarum was so relevant that we further modified a specific culture medium integrating powdered pistachio leaves into diluted PDA for future studies. The modified PLA media resulted in satisfactory induction of a major amount of conidia of S. pistaciarum after 14 to 21 days of incubation under fluorescent light. Indeed, a significant linear trend along with the PLPC was observed for all dependent variables analyzed. Among these variables, conidial production increased significantly over PLPC on diluted PDA. Interestingly, although the PLA-3 and PLA-4 treatments were not useful when the goal was to scrape the medium surface because this technique was impractical as a result of the soapy texture of the culture medium caused by the high PLPC, these treatments were very useful for obtaining highly concentrated conidial suspensions by directly transferring the mycelial surface to vials with SDW.

    Finally, inoculation tests in planta confirmed the pathogenicity of S. pistaciarum in both pistachio and terebinth hosts. According to Crous et al. (2013), the abundant angular dark-brown spots observed between leaf veins (Fig. 6) help to confirm the identity of S. pistaciarum. To our knowledge, the pathogenicity of S. pistaciarum in pistachios and terebinths has not been reported before. Only the pathogenicity of S. pistaciae to pistachio was demonstrated in Egypt by Haggag et al. (2006). It is worth mentioning that the protocol for inoculation used by these authors was adapted to conduct the pathogenicity tests in planta in the present study with successful results. Our results also reveal that the use of detached leaflets is not useful to conduct experiments with this pathogen under laboratory-controlled conditions.

    In summary, we report for the first time that S. pistaciarum causes SLS in pistachio and terebinth in Spain. Although S. pistaciarum is well known in other geographical areas of the world where pistachios are grown, it has only been reported once to infect P. terebinthus in Turkey (Crous et al. 2013). It is possible that the abundance of terebinth bushes in areas close to pistachio crops in Spain has led to its detection in the former host. The results generated here represent a relevant advance in knowledge of the etiology of SLS of pistachio and serve as the first step toward understanding the biology of S. pistaciarum. In addition, we set up laboratory protocols to improve the management of this pathogen to better conduct future experiments in vitro as well as in planta. Therefore, this work represents a basis for planning future studies investigating the epidemiology and control of the disease in Mediterranean pistachio-growing areas.


    The authors thank Crisolar and Mañán Organizations of Fruit and Vegetable Producers (OFVPs) and the private companies Pistachos Nazaríes, Almendras Francisco Morales, and Bain (Borges Group) for their collaboration. The authors thank F. Luque and F. González for their technical assistance in the laboratory.

    The author(s) declare no conflict of interest.

    Literature Cited

    Funding: This research was funded by the Junta de Andalucía (projects PPTRATRA-2016.00.6, Transforma de Fruticultura Mediterránea, and TRA.TRA20019.002, Transforma de Fruticultura Mediterránea, both from the Andalusian Institute for Research and Formation in Agriculture and Fishery), Juan de la Cierva-Incorporación fellowship (contract no. IJCI-2016-2810) from the Spanish Ministry of Science, Innovation and Universities (C. A.-B.), and the Spanish Ministry of Science and Innovation, the Spanish State Research Agency, through the Severo Ochoa and María de Maeztu Program for Centres and Units of Excellence in R&D (Ref. CEX2019-000968-M).

    The author(s) declare no conflict of interest.