2.1. Plant collection and fungal isolation

In July 2021, leaf spot were observed on P. peregrina and P. tenuifolia in the natural habitats in Pirot (43^◦ 07’ N; 22^◦ 26’ E 666) and Deliblato Sands (44^◦ 57’ N; 21^◦ 03’ E 167) areas, Serbia. The spots appeared on leaves and sporadically on stems of P. peregrina, while identical spots, white mycelia and rot appeared on stems and lower branches of P. tenuifolia. The percentage of infected plants at a given locality were calculated by counting the numbers of healthy and symptomatic plants separately. In total, 197 and 100 individual plants of the single populations of P. peregrina and P. tenuifolia were found on about 0.5 ha. Sixty-three plants of P. peregrina and 50 plants of P. tenuifolia were symptomatic.

Symptomatic plants were transplanted into pots and transferred to the laboratory. Isolation of the pathogen was done from symptomatic leaves and stems. Leaf fragments (5x5 mm) of five symptomatic plants and stem fragments (5x5 mm) of ten symptomatic plants were cut from the margins of the lesions and the healthy tissue. Leaf and stems fragments were washed, surface disinfected with 75% ethanol for 1 min, 1% sodium hypochlorite solution (NaClO) for 2 min, rinsed three times in distilled water, dried on sterilized filter paper, and placed on potato dextrose agar (PDA) in Petri dishes (90 mm) (Sun et al., 2022). After seven days of incubation at 25 ^◦C, developed colony fragments were transferred to PDA to obtain pure cultures.

2.2. Pathogenicity test

The pathogenicity was tested by artificial inoculation leaves of plant origin. Two wounded and two non-wounded detached leaves of healthy P. peregrina obtained from the natural habitat were used. The leaves were placed onto the filter paper moistened in Petri dishes, wounded with a sterile needle, and mycelial fragments (5 mm) of 7-day-old culture on PDA were placed on the adaxial surface of the leaves. Control leaves were treated the same way, but sterile PDA fragments were used (Fang et al., 2023). Petri dishes with inoculated leaves and control dishes were incubated at 21±2 ^◦C under natural daylight conditions. The test was repeated once. Re-isolation of the pathogen was performed as described above.

2.3. DNA extraction, PCR amplification and amplicon sequencing

DNA extraction was performed by the cetyltrimethylammonium bromide (CTAB) protocol proposed by Day and Shattock (1997) from 7-day-old cultures grown on PDA. For molecular identification and characterization, the nuclear ribosomal internal transcribed spacer (ITS) region, partial translation elongation factor 1-alpha (EF-1α), beta-tubulin (TUB2), and histone 3 gene (H3) were amplified using primer pairs ITS1/ITS4 (White et al., 1990), EF1-728F/EF1-986R (Carbone and Kohn, 1999), T1/Bt2b (Glass and Donaldson, 1995; O’Donnell and Cigelnik, 1997), and H3-1a/H3-1b (Glass and Donaldson, 1995), respectively. Each 25 μL PCR mix contained 1 μL of template DNA, 1xPCR Master Mix (Thermo Scientific, Vilnius, Lithuania) and 0.4 μM of each primer. Conditions for amplification of ITS were: initial denaturation at 95 ^◦C for 2 min, followed by 35 cycles of denaturation at 95 ^◦C for 30 s, annealing at 55 ^◦C for 30 s and elongation at 72 ^◦C for 1 min. Conditions for EF-1α and TUB2 amplifications were: initial denaturation at 95 ^◦C for 8 min, followed by 35 cycles of denaturation at 95 ^◦C for 15 s, annealing at 55 ^◦C for 20 s and elongation at 72 ^◦C for 1 min. Conditions for H3 amplification were: initial denaturation at 94 ^◦C for 2 min, denaturation at 94 ^◦C for 60 s, annealing at 58 ^◦C for 1 min and elongation 72 ^◦C for 1 min; repeat protocol for 32 cycles. Final elongation occurred for 10 min at 72 ^◦C. PCR products (5 μL) were observed in 1.2% agarose gel, stained in ethidium bromide, and visualized with UV transilluminator. Amplified products were purified and commercially sequenced (Macrogen Inc., Seoul, South Korea) using the same reverse primers as for amplification. The obtained sequences were assembled using Pregap4 from the Staden programme package (Staden et al., 1998). For species identification, the obtained sequences were compared with those publicly available in NCBI’s GenBank database using MegaBLAST (http://www.ncbi.nlm.nih.gov/). Sequences for each locus were aligned using Clustal X (Thompson, 1997), under MEGA version 7 (Kumar et al., 2016). The sequences of isolate were deposited in the National Center for Biotechnology Information (NCBI) GenBank.

2.4. Phylogenetic analyses

Evolutionary history was inferred based on combined analyses of the two (ITS and EF-1α) loci of two representative isolates obtained in this study, reference isolates of Alternaria spp. and Alternaria alternantherae CBS 124392 as an outgroup,using Maximum likelihood (ML) method (MEGA X) (Kumar et al., 2018). Sequences used for phylogeny are presented in Table 1. For the ML, the best nucleotide substitution model was determined using the “find best model” option in MEGA X. For tree inference, the nearest-neighbor-interchange heuristic method was employed, with the initial trees being automatically generated by Neighbor-Join and BioN Jalgorithms. To estimate the statistical significance of the inferred clades, each tree was bootstrapped 1,000 times.

2.5. Morphological characterization

Macromorphological characteristics of two representative isolates (colony color, texture, and growth rate) were observed on PDA and PCA 40 cm below cool white fluorescent illumination at 23 ^◦C for seven days (Blagojević, 2020; Simmons, 2007). The colony diameter was measured every day in two diagonal directions. Mean values were obtained for every day from three replicates with three plates each. The growth rate was calculated and expressed as the diameter of the colony after seven days. Micromorphological characteristics of selected isolates (length, width, shape, size, color of conidia, conidiophores, dimensions of the conical beak, and sporulation pattern) were observed on seven-day-old cultures grown on PCA at 22 ^◦C under cool white fluorescent illumination and an 8/16 h light/dark regime (Simmons, 2007). Microscopic preparations were prepared in lactic acid and viewed under a microscope at 10x and 20x magnifications (Olympus CX43, Olympus, Hamburg, Germany). Photographs of conidiophore and conidia and measurements were obtained using the camera Axiocam ERc 5s, Zeiss, and software ZEN 2 (blue edition), Jena, Germany.

3.1. Symptoms, isolates and pathogenicity

Circular or irregular, light to dark brown spots with defined dark margins appeared on the leaves and sporadically stems of P. peregrina during the stage of follicle formation, whereas on P. tenuifolia identical spots first appeared from the base at the bud stage and finally white mycelia and rot appeared at the base of the stems and lower branches at the seed maturity stage (Figure 1). The spread of spots on leaves causing the plants to wilt and die. Disease incidences in the areas of Pirot and Deliblato Sands were 32% and 25-35%, respectively.

Following a seven-day incubation, five isolates of P. peregrina and five isolates of P. tenuifolia were obtained from symptomatic plants. In the pathogenicity test, on wound-inoculated leaves, leaf necrosis and mycelia appeared on the adaxial surface of the leaves, as well as necrosis along the leaf nerve, while non-wounded leaves and control leaves remained symptomless (Figure 2). The pathogen was reisolated from inoculated leaves and showed identical morphological characteristics as the original isolate, thus completing Koch’s postulates. From each host, a representative isolate with proven pathogenicity was selected for further molecular and morphological characterization (isolate PpPili from P. peregrina and isolate SBDPst from P. tenuifolia).

3.2. Molecular identification

Following the amplification of ITS, EF-1α, TUB2 and H3, PCR products of expected size (600, 300, 600 and 400 bp, respectively) were obtained for the two selected isolates. Comparison of the obtained sequences showed that isolates PpPili and SBDPst were 100% similar (identical) in ITS and EF-1α, but differed in 1 nt in TUB2 and 1 nt in H3 gene regions. BLAST analysis showed that both isolates were 100% similar with A. alternata ex-type CBS 916.96, established by Woudenberg et al. (2013) and Woudenberg et al. (2015), in ITS (AF347031) and EF-1α (KC584634). Also, sequences of both regions of obtained isolates were identical with several corresponding sequences of A. alternata isolates (e.g. CBS 115188, CBS 113013, CBS 117143, CBS 119115, CBS 918.96, IT44, IT61) (Matić et al., 2020; Woudenberg et al., 2015). In TUB2, isolate PpPili was 100% similar with A. alternata ex-type CBS 916.96 (=EGS 34.016; MF070244) (Siciliano et al., 2018) and other published A. alternata sequences of TUB2 region (Garibaldi et al., 2019; 2020), while isolate SBDPst was 99% similar with A. alternata ex-type CBS 916.96 and 100% similar with CBS 918.96 and CBS 115152 (Siciliano et al., 2018). In H3, isolate PpPili was 100% similar with A. alternata isolate IT44 (MG182429) (Garibaldi et al., 2018a), while isolate SBDPst was 100% similar with A. alternata isolate IT61 (MF997593) (Garibaldi et al., 2018b). Based on ITS, EF-1α, TUB2 and H3, obtained isolates were identified as A. alternata which is in accordance with Woudenberg et al. (2015), Garibaldi et al. (2019; 2020; 2018a), Siciliano et al. (2018), and Matić et al. (2020).

3.3. Phylogeny

The phylogenetic analysis, based on Maximum likelihood method and Kimura 2-parameter model, of concatenated ITS and EF-1α sequences of Alternaria spp. grouped two representative isolates, PpPili and SBDPst, with reference isolates of A. alternata (Figure 3), confirming the identification.

3.4. Macromorphology and micromorphology

After seven days, isolate PpPili formed circular, cottony, light brown to brown colonies, with a white border on PDA (diameter 60.3 mm), and circular, grayish-brown colonies on PCA (diameter 83.7 mm). Conidia were brown, ovoid, or ellipsoid with three to four transverse and one to three longitudinal septa (16.01-27.74 x 5.50-14.41 μm; average 20.83 x 9.41 μm, n = 50). Conidia formed in chains of 2-7, mostly 5-6. Dimensions of the conical beak were 0.70-5.27 x 1.53-4.57 μm; average 2.55 x 2.68 μm, n = 50. Conidiophores were simple, dark brown, and unbranched (27.86-187.31 x 1.91-7.73 μm; average 59.88 x 4.39 μm, n = 50) (Figure 4).

After seven days, isolate SBDPst formed circular, cottony, white, or light brown colonies on PDA (diameter 45.3 mm). On PCA, colonies were circular, grayish-brown (diameter 85.5 mm), and brown, ovoid or ellipsoid conidia formed in chains of 3-10, mostly 3-8. Conidia had zero to four transverse and zero to three longitudinal septa, and were 13.56-24.27 x 5.7-14.3 μm in size (average 18.93 x 9.67 μm, n = 50), with a conical beak 1.00-5.52 in long (average 2.82) x 0.65-3.34 in wide (average 1.73) μm (n = 50). Conidiophores were simple, dark brown, and unbranched, 12.64-53.15 x 1.29-5.97 μm in size (average 25.98-2.36, n = 50) (Figure 5).

Observed morphological characteristics corresponded to the genus Alternaria (Simmons, 2007). According to the sporulation pattern the isolates could be identified as an A. tenuissima morphospecies, but with regards to a comprehensive revision by Woudenberg et al. (2015) and recently, Dettman and Eggertson (2021) which show lack of phylogenetic divergence in support of morphological differences and the fact that morphospecies are synonymized with A. alternata, the isolates were identified as A. alternata.

In this study, Alternaria leaf spot was described on two herbaceous peony species P. peregrina and P. tenuifolia in Serbia. Isolated fungi whose pathogenicity was proven, were identified as A. alternata. In Serbia, there has been a report of Alternaria leaf spot affecting wild herbaceous peony (Mikić et al., 2023), but other diseases have not been identified in these plant species. In the United States, Garfinkel and Chastagner (2019) described Alternaria leaf spot and other diseases on the herbaceous peony species P. lactiflora Pallas.. Also, Park et al. (1997) examined the effect of different pathogenic fungi (Alternaria sp., Erysiphe aquilegiae, and Cronartium flaccidum) on the growth and root yield of P. lactiflora; brown, dark-brown to black spots were observed on leaves, and finally the withering of the above-ground part of the plant. The symptoms of A. alternata on tree peonies described by Shi et al. (2015) were similar with our findings of A. alternata on herbaceous P. peregrina. Zhang et al. (2008) reported A. suffruticosae, A. suffruticosicola, and A. tenuissima on woody peony – Paeonia suffruticosa cultivar Andrews.