The genus Castanediella

Abstract Two new species, Castanediellabrevis and C.monoseptata, are described, illustrated and compared with other Castanediella taxa. Evidence for the new species is provided by morphological comparison and sequence data analyses. Castanediellabrevis can be distinguished from other Castanediella species by the short hyaline conidiophores and fusiform, aseptate hyaline conidia, while C.monoseptata differs from other Castanediella species by its unbranched conidiophores and fusiform, curved, 0–1-sepatate, hyaline conidia. Phylogenetic analysis of combined ITS and LSU sequence data was carried out to determine the phylogenetic placement of the species. A synopsis of hitherto described Castanediella species is provided. In addition, Castanediella is also compared with morphologically similar-looking genera such as Idriella, Idriellopsis, Microdochium, Neoidriella, Paraidriella and Selenodriella.

During a survey of hyphomycetes in Thailand, two hyaline-spored hyphomycetes were collected. They were shown to belong to the genus Castanediella based on morphology and phylogeny analyses of ITS and LSU sequence data. The new species C. brevis and C. monoseptata are introduced.

Collection and isolation of fungi
Dead leaves from a variety of plants in two forests (Lampang province and Chiang Mai province) were collected in 2016 in Thailand. Samples were taken to the laboratory in Zip-lock plastic bags for examination. The specimens were incubated in sterile moist chambers and examined using a Motic SMZ 168 series microscope. Fungi were removed with a needle and placed in a drop of distilled water on a slide for morphological study. Photomicrographs of fungal structures were captured with a Canon 600D digital camera attached to a Nikon ECLIPSE Ni compound microscope. All measurements were made by the Tarosoft (R) Image FrameWork program. Photo-plates were made with Adobe Photoshop CS3 (Adobe Systems, USA). Isolation of the fungi on to potato dextrose agar (PDA) was performed by the single spore isolation method (Chomnunti et al. 2014 (Jayasiri et al. 2015;Index Fungorum 2018).

DNA extraction, PCR amplification and sequencing
Genomic DNA was extracted from fungal mycelium grown on PDA or malt extract agar (MEA) at room temperature using the Fungal gDNA Kit (BioMIGA, USA) according to the manufacturer's instructions. The internal transcribed spacer region of ribosomal DNA (ITS) and large subunit nuclear ribosomal DNA (LSU) genes were amplified via polymerase chain reaction (PCR) using the following primers: ITS5 and ITS4 (White et al. 1990) for ITS, and LR0R and LR5 (Vilgalys and Hester 1990) for LSU. The PCR products were sequenced with the same primers. The PCR amplification was performed in a 25 μL reaction volume containing 12.5 μL of 2 × Power Taq PCR MasterMix (a premix and ready to use solution, including 0.1 Units/μl Taq DNA Polymerase, 500 μM dNTP Mixture each [dATP, dCTP, dGTP, dTTP], 20 mM Tris-HCl pH 8.3, 100 Mm KCl, 3 mM MgCl 2 , stabilizer and enhancer), 1 μL of each primer (10 μM), 1 μL genomic DNA extract and 9.5 μL deionised water. The PCR thermal cycle program of ITS and LSU were followed as: initially 94 °C for 3 min., followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 50 s, elongation at 72 °C for 1 min., and final extension at 72 °C for 10 min.

Phylogenetic analyses
Original sequences were checked using BioEdit version 7.0.5.3 (Hall 1999), and most reference sequences were originated from previous publications. The remaining homogenous sequences were obtained by BLAST searches (Altschul et al. 1990) from GenBank. All sequences used in this study are listed in Table 1. Alignments for each locus were done in MAFFT v7.307 online version (Katoh and Standley 2016) and manually verified in MEGA 6.06 (Tamura et al. 2013). After alignment, the concatenation of different genes was done in SequenceMatrix 1.8 (Vaidya et al. 2011). The interleaved NEXUS files for Bayesian inference analyses were formatted with AliView v1.19-beta1k (Larsson 2014). Maximum parsimony (MP), maximum likelihood (ML) and Bayesian inference (BI) were used for phylogenetic analyses.
The best models of evolution for each gene region were determined using Akaike information criterion (AIC) as implemented in MrModeltest v2 (Nylander 2004). The analyses' results showed that the models GTR+I and GTR+I+G were the best ones for LSU and ITS sequence data, respectively.
For BI analysis, Posterior probabilities (PP) (Rannala and Yang 1996;Zhaxybayeva and Gogarten 2002) were determined by Markov Chain Monte Carlo sampling (BM-CMC) in MrBayes v 3.2.6 (Ronquist et al. 2012). For the combined dataset, the models were set to nst = 6 and rates = propinv for LSU and nst = 6 and rates = invgamma for ITS. Two independent analyses of two parallel runs and six simultaneous Markov chains were run for 1,000,000 generations, trees were sampled every 100 th generation and the temperature value of the heated chains was set at 0.15. The first 25% sampled trees of each run were discarded as "burn-in", and the remaining trees were used for calculating posterior probabilities (PP) in the majority rule consensus tree with the sumt command in MrBayes.
Phylogenetic trees were drawn with TreeView 1.6.6 (Page 1996). Bootstrap support values for maximum parsimony (MP, first set) and maximum likelihood (ML, second set) greater than 50% are indicated above or below the nodes. Ex-type strains are in bold, the new isolates are in red. The tree is rooted with Subsessila turbinata (MFLUCC 15-0831).

Molecular phylogeny
The aligned sequence matrix comprises LSU and ITS sequence data for 16 taxa (ingroup) and one outgroup taxon with a total of 1438 characters after alignment including the gaps, of which 120 were parsimony informative, 77 parsimony-uninformative, and 1241 characters constant. The dataset consists of thirteen species within the genus.
The tree was rooted with Subsessila turbinata (MFLUCC 15-0831). Maximum parsimony analysis resulted in two trees with TL = 391, CI = 0.657, RI = 0.642, RC = 0.422, HI = 0.343. For the Bayesian analysis, two parallel runs with six chains were run for 1,000,000 generations and trees were sampled every 100 th generation, resulting in 20002 trees from two runs of which 15002 trees were used to calculate the posterior probabilities (each run resulted in 10001 trees of which 7501 trees were sampled).The MP and ML (lnL = -4041.301739) analyses based on combined LSU and ITS sequence data provided similar tree topologies, and the result of MP analysis is shown in Fig. 1. The novelty of the species, Castanediella brevis and C. monoseptata, described in this study are supported by sequence data analyses as belonging to the genus Castanediella, but with low bootstrap support values. Isolates of Castanediella brevis and C. monoseptata formed separate clades in the phylogenetic inference, respectively. Castanediella brevis is sister to C. malaysiana and C. ramosa, while C. monoseptata shows close phylogenetic relationship to C. couratarii and C. malaysiana. Both the new taxa can be recognized as phylogenetically distinct species and are clearly novel based on the recommendations for molecular data (Jeewon and Hyde 2016).
Culture characteristics: Conidia germinating on PDA within 24 h. Colonies on PDA effuse, greyish white to dark from above and below, reaching a diam. of 5-7 cm in 30 days at 25 °C.
Among the species that produce more or less falcate and aseptate conidia, Castanediella communis, C. eucalypti, C. eucalypticola and C. eucalyptigena are most similar to C. brevis. However, Castanediella brevis differs from these species by its short, unbranched and 0-1-septate conidiophores.

Discussion
In this study, two new Castanediella species, C. brevis and C. monoseptata, were identified from decaying leaves in Thailand and a synopsis of hitherto described Castanediella species is provided (Table 2). Presently, the genus Castanediella contains 14 species, and is shown to be diverse in its habitats. Most of Castanediella species have been collected from plant leaves. Castanediella acaciae, C. camelliae, C. communis, C. eucalypti, and C. eucalypticola were isolated from disease symptoms on different host plant leaves (Crous et al. 2015Wanasinghe et al. 2018) whereas C. cagnizarii is the only species found on decaying leaves submerged in a stream (Castañeda Ruiz et al. 2005). Some Castanediella species were reported from decaying leaves, such as C. brevis, C. cagnizarii, C. diversispora, C. hyalopenicillata and C. monoseptata (Castañeda Ruiz et al. 2005;Hernández-Restrepo et al. 2016b;Costa et al. 2018). Castanediella couratarii was reported from dead wood (Hernández-Restrepo et al. 2016a).