1 Introduction

Endophytic fungi (EPF) colonize plant tissue without causing any visible disease symptoms at any particular moment (Schulz and Boyle 2005). Practically, EPF colonize almost any plant tissue, including leaf, stem, and root (Rodriguez et al. 2009). Since their discovery, EPF have been studied in many types of plants, including non-vascular, such as mosses (e.g., Schulz et al. 1993) and algae (e.g., Zuccaro et al. 2008), and vascular, such as shrubs (e.g., Schulz et al. 1993) and trees (e.g., Arnold and Lutzoni 2007). Most of the isolated EPF belong to the Ascomycota or Basidiomycota (Schulz and Boyle 2005).

A conservative estimate of worldwide fungal diversity has been proposed as 1.5 million species, which has been accepted as the working hypothesis and basis for the discovery of more fungal species (Hawksworth 2001). With only around 72,000 described species known so far, more than 1 million species are waiting to be found. EPF have been found in many plant species and are considered an important component of fungal diversity (Rodriguez et al. 2009). Despite the increasing number of studies of EPF in many countries, those based in the tropics are still lacking.

Arnold et al. (2000) isolated 418 EPF morphospecies colonizing the leaves of two understory tree species in a tropical forest in Panama, 59% of which were represented by single isolates. Cannon and Simmons (2002) isolated 64 EPF morphospecies colonizing the leaves of 12 tree species in a tropical forest in Guyana, 29 of which were from single leaf samples. These two studies reflect the high diversity of EPF colonizing tree leaves in tropical forests.

The different environmental factors in forests are expected to influence fungal diversity (Saikkonen 2007). However, most studies of the diversity of EPF colonizing leaves in tropical forests are limited to one forest site (e.g., Arnold et al. 2000, 2001; Cannon and Simmons 2002). Suryanarayanan et al. (2011) compared the EPF communities of 75 dicotyledonous trees belonging to 33 families from three tropical forest types in Southern India, namely, tropical dry thorn forest, dry deciduous forest, and montane evergreen forest. The type of forest appeared to have a larger effect on shaping the EPF community than the taxonomy of the host.

Existing studies of EPF in tropical forests are limited to EPF that colonize the above-ground part of plant, particularly leaf. EPF colonizing roots, further termed as root EPF, of tropical forest trees are rarely studied. Rodriguez et al. (2009) classified root EPF into two groups: class 2 EPF, which colonize the shoot, root, and rhizome, and class 4 EPF, also termed dark septate endophyte (DSE), which colonize the root only. In a review by Jumpponen and Trappe (1998), they noted that DSEs colonized approximately 600 plant species, representing 320 genera and 114 families, highlighting the abundance of DSE. However, DSE is not the only group of root EPF, indicating the possibility of an even higher abundance of root EPF in nature.

Root EPF are considered to play an important role in plant growth, similar to mycorrhizal fungi (Jumpponen and Trappe 1998). Researchers have recorded evidence about the growth promoting ability of several class 4 EPF. Exophiala pisciphila improved tolerance of maize (Zea mays L.) to cadmium toxicity (Wang et al. 2016). Leptodontidium orchidicola and Phialophora mustea improved growth of Tasmanian blue gum (Eucalyptus globulus) and birch (Betula pendula) (Berthelot et al. 2016). Veronaeopsis simplex suppressed Fusarium disease in Chinese cabbage (Brassica campestris) (Khastini et al. 2012). Heteroconium chaetospira promoted plant growth by increasing nitrogen uptake (Usuki and Narisawa 2007). Fusarium sp. improved plant resistance against fungal pathogen Gaeumannomyces graminis var. tritici (Maciá-Vicente et al. 2008). There have also been studies that recorded evidence about the growth promoting ability of class 2 EPF. Penicillium citrinum improved resistance of banana to a fungal pathogen, Fusarium oxysporum (Ting et al. 2012). Fusarium sp. improved resistance of eggplant to Verticillium wilt (Narisawa et al. 2002). Meta-analysis of data from temperate and boreal areas showed that root EPF colonization could have a negative, neutral, or positive effect on plant growth (Mandyam et al. 2013; Mayerhofer et al. 2013; Newsham 2011). Evidence of the positive effects of root EPF, along with the expectation of high EPF abundance in tropical forests, has underscored the necessity to conduct more studies of root EPF in tropical forests. However, studies of root EPF in tropical area remain a rarity (Mandyam and Jumpponen 2005).

The trap culture method is rarely used in the study of EPF, but is common in study of arbuscular mycorrhizal fungi (AMF). EPF and AMF have been found in the same plant community (Lingfei et al. 2005). Thus, utilization of the same plant species in trap culture for isolation of EPF is highly possible. Sorghum bicolor, Trifolium repens, and Allium fistulosum are crop species commonly used to isolate AMF (Del Val et al. 1999), thus can also be used to isolate EPF. Tree species are also a possible candidate to isolate EPF. Paraserianthes falcataria, Calliandra calothyrsus, Cassia siamea, and Sesbania grandiflora are common leguminous tree species in the tropics, particularly Indonesia. These tree species are fast-growing and are also candidate species for reforestation efforts (Otsamo et al. 1997, Wulandari et al. 2016). Considering the importance of the future utilization of EPF in reforestation, these tree species can also be used as host plants to isolate EPF. The objectives of present study were to (1) isolate root EPF from five forests soils, and identify them based on their rDNA ITS region, and (2) compare EPF community among different forests.

2 Materials and methods

2.1 Collection of forest soil

Forest soils were collected from five forests in Indonesia, with five replications in September 2012 (Table 1). Soils between 0 and 10 cm depth and within 0–5 cm distance from the root of representative seedlings, or around the main stem of representative tree species, were collected after removing the organic layer. Two kilograms of soil were collected for each replication. The distance between replications in one forest was ≥20 m. The latitude coordinates of each forest were recorded.

Table 1 Characteristics of forest sites used in this study
Full size table

2.2 Chemical analysis of soil

Soil pH was measured at the soil:solution ratio of 1:2.5 in deionized water and 1 M KCl solution (Table 2). Soil available phosphate was determined following the Truog method (Truog 1930). Briefly, air-dried soil was suspended in 0.001 M sulfuric acid solution, stirred by a mechanical shaker, and filtered immediately using a filter paper (Advantec Toyo No. 6, Japan). Color development solution (Olsen and Sommers 1982) was added to the filtrate, and absorbance was measured with a spectrophotometer (U-2900, Hitachi, Japan) at 880 nm.

Table 2 Chemical properties of forest soils collected from Indonesia
Full size table

Cation exchange capacity (CEC) was determined by the semi-micro Schollenberger method, which is based on the displacement of soil cations by leaching soil with an excess of 1 M neutral (pH 7.0) NH4CH3COO solution, as described by the United States of Agriculture (USDA) (Soil survey staff 1992). Briefly, the soil was leached with NH4CH3COO solution, allowing NH4+ to replace soil cations. The soil was leached again with 1.35 M KCl solution to replace NH4+. By titration with formaldehyde and thymol blue, the amount of NH4+ was measured and CEC was calculated. Leached exchangeable cations (Ca, Mg, Na, and K) from the soil were measured by an atomic absorption spectrophotometer (Z-5000, Hitachi, Japan).

2.3 Experiment 1. Isolation of EPF from tree and crop species

Sand was acidified and sterilized by autoclaving at 80 °C for 45 min. Forty grams of sterilized sand was mixed with 40 g of forest soil, then used as growth medium. Seeds of Calliandra calothyrsus, Paraserianthes falcataria, Sesbania grandiflora, and Cassia siamea were sown on sterilized sand and incubated in a growth chamber (Biotron LPH-350S, NK System, Japan) at 27 °C with a 16-h photoperiod. One of two-leaf-stage seedlings of C. calothyrsus, P. falcataria, S. grandiflora, and C. siamea was transplanted onto the medium in a 50 ml syringe pot. Three, five, or 20 seeds of Sorghum bicolor, Allium fistulosum, or Trifolium repens, respectively, were sown onto the same medium. Ten grams of sterilized sand was further added into the syringe pot to cover the root system of seedlings or seeds. All plants were grown in the growth chamber for 90 days at 27 °C, with a 16-h photoperiod. Five to ten milliliters of 1 mg P L−1 nutrient solution (based on Wagatsuma et al. 1988) was applied once every two days. Twenty-five pots were prepared for each plant species.

C. calothyrsus, P. falcataria, S. grandiflora, C. siamea, S. bicolor, A. fistulosum, and T. repens were harvested 90 days after transplanting or sowing. Fresh roots were extracted from each soil and washed under running tap water. The roots were surface-sterilized following the method of Verma et al. (2012) by dipping in 90% EtOH (1 min), 5% NaClO (5 min), and 90% EtOH (10 s) and rinsing three times with sterilized deionized water. The roots were dried with sterilized Kimtowels™ and left to air dry. The air-dried roots were cut into 5 mm. Five pieces were plated on half-strength malt extract agar (MEA) containing 100 μg mL−1 Penicillin-Streptomycin (Lonza Biowhittaker, Penicillin-Streptomycin Mixture) (modified from Verma et al. 2012) and water agar. Five replication plates were made for each medium, for roots from one pot, for a total of 50 plates per pot. The plates were sealed with Parafilm™ and incubated in the dark at 25 °C. Emerging colonies from roots were subcultured on new half-strength MEA medium. EPF isolated from this experiment were used to calculate the isolation rate. Isolation rates were calculated by dividing the number of isolates by the number of initial plates.

2.4 Experiment 2. Isolation of EPF from P. falcataria and S. bicolor

EPF were isolated again from P. falcataria and S. bicolor. P. falcataria was chosen because this species would be used as the target species in future studies and the EPF isolation rates for the four tree species were not markedly different (Table 3). This species is not only a candidate for reforestation, but also economically profitable in a mixed plantation with crop species, or as single-species plantation (Siregar et al. 2007; Krisnawati et al. 2011), and a candidate for energy production (Amirta et al. 2016). S. bicolor was chosen because it had the highest EPF isolation rate among crop species, and also among all species.

Table 3 Numbers of isolates and isolation rates of endophytic fungi from seven plant species
Full size table

EPF were isolated from P. falcataria and S. bicolor using the same method described above (Experiment 1). However, the nutrient solution for watering was changed to tap water to minimize nutrient input, mimicking the conditions in the forest. Seedling number per pot, plant height, leaf number, and symptoms of nutrient deficiency, on leaves of P. falcataria and S. bicolor were recorded before harvest (Tables S1 and S2). Plants showing good growth were harvested and EPF were isolated from the roots.

2.5 DNA extraction and amplification

EPF isolated in experiment 2 were subcultured on half-strength MEA for 1–3 weeks depending on the growth rate of each isolate. Hyphae of each isolate were collected using forceps and placed on the lid of a plastic tube containing 20 μL of InstaGene™ Matrix (Bio-Rad, USA). Hyphae were crushed with a pipet tip having a blunt end, further mixed with 180 μL of InstaGene™ Matrix, and vortexed. Next, rDNA was extracted following the manufacturer’s protocol for InstaGene™ Matrix. The extracted DNA was stored at −20 °C until use.

The internal transcribed spacer (ITS) region of the fungi was amplified using universal primers, ITS1F (5’-GTAACAAGGTTTCCGT-3′) and ITS1R (5’-CGTTCTTCATCGATG-3′) (Fujita et al. 2010), with an Expand High FidelityPLUS PCR system (Roche, Germany) at the following composition: 4 μL of 5× buffer with MgCl2, 0.2 μL of DNA polymerase, 2 μL of 2.0 mM dNTP, 0.4 μL of ITS1F, 0.4 μL of ITS1R, 11 μL of Milli-Q water, and 2 μL of DNA template. The reaction was performed in a Takara PCR Thermal Cycler Dice (Model TP600, Takara Bio, Japan) under the following conditions: initial denaturation at 94 °C for 120 s; 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 60 s, and extension at 72 °C for 60 s; and final extension at 72 °C for 420 s. The PCR products were separated on 1.0% agarose gel (D1 Agarose Low EEO, Conda, Spain) in 1× Tris-borate-EDTA buffer, stained with SYBR® Safe DNA Gel Stain (Invitrogen, USA), and viewed under blue light (470 nm, MBP-LED, Bio-Pyramid, USA). PCR-amplified fragments were purified using a MonoFas DNA Purification Kit (GL Science, Japan) following the manufacturer’s protocol. Purified DNA was ligated into pT7Blue T-Vector (Novagen, USA) using a DNA Ligation Kit Ver 1 (Takara Bio, Japan). Twenty microliters of IPTG (Takara Bio, Japan) and 35 μL of X-Gal (Takara Bio, Japan) were applied to Luria Bertani (LB) medium containing 100 mg L−1 ampicillin. T-Vector containing DNA was transformed into Escherichia coli JM109 (Takara Bio, Japan) by plating onto this LB medium. Plates with E. coli were incubated at 37 °C for 16 h.

Single colonies of E. coli were collected and DNA was amplified using primers T7 (5’-TAATACGACTCACTATAG-3′) and U19 (5’-GTTTTCCCAGTCACGACT-3′) (Ikenaga et al. 2016) with GoTaq® DNA Polymerase (Promega, USA) at the following composition: 2 μL of 5 x reaction buffer, 0.05 μL of DNA polymerase, 0.8 μL of 2.0 mM dNTP, 0.2 μL of T7, 0.2 μL of U19, 6.75 μL of Milli-Q water, and a single colony of E. coli. The reaction was performed in a Takara PCR Thermal Cycler Dice (Model TP600, Takara Bio, Japan) under the following conditions: initial denaturation at 94 °C for 120 s; 35 cycles of denaturation at 94 °C for 15 s, annealing at 50 °C for 60 s, and extension at 72 °C for 80 s; and final extension at 72 °C for 600 s. The PCR products were separated on 1.0% agarose gel in 1× Tris-borate-EDTA buffer, stained with SYBR® Safe DNA Gel Stain, and viewed under blue light. The PCR products were used for sequencing.

Sequencing reactions were performed in a Bio-Rad DNA Engine Dyad PTC-220 Peltier Thermal Cycler using an ABI BigDye™ Terminator v3.1 Cycle Sequencing Kit with AmpliTaq DNA Polymerase (FS enzyme, Applied Biosystems, Japan) following the manufacturer’s protocol. Single pass sequencing was performed on each DNA template using a T7 promoter. Fluorescent-labeled fragments were purified from the unincorporated terminators by adopting an ethanol precipitation protocol. The samples were resuspended in distilled water and subjected to electrophoresis in an ABI 3730xl sequencer (Applied Biosystems, Japan).

2.6 Phylogenetic analyses

Sequences of EPF isolates were submitted for BLAST analysis (Altschul et al. 1990). The sequences and their corresponding BLAST top hits were aligned by Multiple Alignment using Fast Fourier Transform (MAFFT) (Katoh et al. 2002) through http://guidance.tau.ac.il.. Maximum parsimony method was performed by MEGA 7 (www.megasoftware.net) with 1000 replications of bootstrap analysis (Fig. 1).

Fig. 1
figure 1

Maximum parsimony analysis of ITS1 region sequences of endophytic fungi. Values from bootstrap analysis (1000 replications) are shown

Full size image

3 Results

3.1 Soil chemical properties

Soil pH (H2O) ranged from 3.87 to 7.02 and soil pH (KCl), from 3.26 to 5.99 (Table 2). Based on USDA classification, soil pH (H2O) was neutral in both Tectona grandis and Swietenia macrophylla monoculture. Soil pH (H2O) was extremely acidic to very strongly acidic in Gmelina sp., Artocarpus champeden, and Dipterocarp mixed, Dipterocarp primary, and Macaranga sp. secondary forests. Available P concentration was 1.00 mg 100 g dry soil−1 in T. grandis monoculture and 0.99 mg 100 g dry soil−1 in S. macrophylla monoculture, and both values were higher than those in Gmelina sp., A. champeden, and Dipterocarp mixed, Dipterocarp primary and Macaranga sp. secondary forests. CEC ranged from 9.34 to 75.39 cmol kg−1; it was higher in T. grandis monoculture and S. macrophylla monoculture than in Gmelina sp., A. champeden, and Dipterocarp mixed, Dipterocarp primary or Macaranga sp. secondary forest. The same tendency as soil CEC was noted for exchangeable Ca, Mg, Na, and K.

3.2 Isolation of EPF

A total of 197 EPF were isolated from the roots of the seven host plant species grown in the five forest soils (Table 3). Among the 197 EPF, 13, 9, 17, 21, 75, 3, and 59 EPF were isolated from the roots of C. calothyrsus, P. falcataria, S. grandiflora, C. siamea, S. bicolor, A. fistulosum, and T. repens, respectively. S. bicolor and T. repens had higher EPF isolation rates than the other plant species. Moreover, among these, 197 EPF, 52, 55, 53, 13, and 24 EPF were isolated from T. grandis monoculture, S. macrophylla monoculture, Gmelina sp., A. champeden, and Dipterocarp mixed, Dipterocarp primary, and Macaranga sp. secondary forest, respectively. Soils of T. grandis monoculture, S. macrophylla monoculture, and Gmelina sp., A. champeden, and Dipterocarp mixed forest had higher EPF isolation rates than the other forest soils.

Leaf necrosis was detected in some seedlings of S. bicolor. Seedlings with higher plant height and leaf number, and less leaf necrosis, in 12 of a total 25 pots, were selected and used for isolation of EPF (Table S1). Twenty-one EPF were isolated from the roots of S. bicolor (Tables 4 and S1). Among the 21 EPF, 5, 2, 7, and 7 were isolated from soils of T. grandis monoculture, S. macrophylla monoculture, Gmelina sp., A. champeden, and Dipterocarp mixed, and Macaranga sp. secondary forest, respectively. EPF isolation rate for S. bicolor was 35%.

Table 4 EPF isolated from roots of Paraserianthes falcataria and Sorghum bicolor grown in forest soils of Indonesia
Full size table

Leaf necrosis was not detected in all seedlings of P. falcataria. Seedlings with higher plant height and leaf number in 10 of a total 25 pots, were selected and used for isolation of EPF (Table S2). Twelve EPF were isolated from the roots of P. falcataria (Tables 4 and S2). Among the 12 EPF, 2, 7, and 3, were isolated from soils of T. grandis monoculture, Gmelina sp., A. champeden, and Dipterocarp mixed, and Macaranga sp. secondary forest, respectively. The EPF isolation rate for P. falcataria was 24%, which was higher than that of the first isolation of EPF.

3.3 Identification of EPF

The sequences of the ITS regions of all isolates were submitted to BLAST including uncultured or environmental samples. The similarity scores between isolated EPF and their closest relatives in GenBank were between 84 and 100% (Table 4). All sequences of isolated EPFs were deposited in the DNA Data Bank of Japan (DDBJ) under accession numbers LC334061–LC334093.

Seven genotypes of EPF isolated from the roots of P. falcataria closely matched fungi, identified to the species level (Table 4). One genotype closely matched a fungus identified to the phylum level. The remaining four genotypes closely matched uncultured fungi. Based on the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov) database, the identified fungi were classified in orders Pleosporales, Hypocreales, Eurotiales, and Chaetothyriales.

Nine genotypes of EPF isolated from the roots of S. bicolor closely matched fungi identified to the species level (Table 4). One genotype closely matched a fungus identified to the genus level (Fusarium). One genotype closely matched a fungus identified to the family level (Clavicipitaceae). One genotype closely matched a fungus identified to the order level (Sordariales). One genotype closely matched a fungus identified to the class level (Dothideomycetes). There was one genotype that closely matched unidentified fungus (Fungal sp. voucher). The remaining seven genotypes closely matched uncultured fungi. Based on the NCBI database, the identified fungi were classified in the orders Pleosporales, Hypocreales, Eurotiales, and Magnaporthales.

4 Discussion

4.1 Specificity of EPF for host plants and forest sites

EPF from the orders Pleosporales, Hypocreales, Eurotiales, Magnaporthales, and Chaetothyriales were isolated in the present study (Table 4). EPF from orders Pleosporales, Hypocreales, and Eurotiales were isolated from both host plants P. falcataria and S. bicolor, whereas EPF from order Magnaporthales and Chaetothyriales were specifically isolated from P. falcataria and S. bicolor, respectively. EPF in these five orders were also isolated from the roots of subtropical Hordeum murinum in Ireland (Murphy et al. 2015). EPFs from two orders Pleosporales and Hypocreales were also isolated from the roots of plants growing in places with different environmental conditions, such as desert grass Bouteloua gracilis in New Mexico (Porras-Alfaro et al. 2008) and tropical shrub Sophora tonkinensis in China (Yao et al. 2017), showing the wide distribution of EPF from these two orders in nature.

EPF from the genera Acrocalymma, Fusarium, Tolypocladium, Penicillium, Talaromyces, Exophiala, Dictyosporium, Pseudochaetosphaeronema, Mariannaea, Trichoderma, and Mycoleptodiscus, have also been reported in other studies. Those EPFs were reported in the studies by Jin et al. (2017) (Acrocalymma vagum), Kwasna et al. (2016) Talaromyces verruculosus and Trichoderma spirale, Lin et al. 2007 (Dictyosporium sp.), Shubin et al. 2014 (Mycoleptodiscus sp. and Penicillium sp.), Yao et al. 2017 (Fusarium solani), Zhang et al. 2017a (Exophiala piscipila), Waipara et al. 1996 (Mariannaea sp.), Zhang et al. 2017b (Pseudochaetosphaeronema larense), and Sánchez Márquez et al. 2010 (Tolypocladium cylindrosporum). However, to our knowledge, the present study is the first to isolate EPF in these genera from the roots of P. falcataria. Amin (2013) isolated EPF belonging to a different genus, Nigrospora sp. (order Trichosphaeriales), from the roots of P. falcataria.

Sixteen of the 33 isolates had the closest match to fungi identified to the species level (Table 4). Ninety-seven percent similarity is widely used as the cut-off point (O’Brien et al. 2005) to determine whether the isolates are identical at the species level, or not. Thirteen of the 16 isolates were considered to be of the same species with the closest match. The remaining three isolates, 2354(1)-2, 2624(5), and 2655(2), were similar to Exophiala calicioides (84% similarity), Pseudochatoesphaeronema martinelli (95% similarity), and Mycoleptodiscus terrestris (94% similarity), respectively. These three isolates may be categorized in the same genera with the closest match.

Among the 16 isolates, 3 were specific to certain forest sites shared by P. falcataria and S. bicolor: Fusarium solani in T. grandis monoculture, Talaromyces verruculosus in Gmelina sp., A. champeden and Dipterocarp mixed, and Talaromyces aculeatus in Macaranga sp. secondary forest (Table 4). In addition, some of the isolates were specific to certain forest sites, but were not shared by the two host plants, examples of which are Dictyosporium heptasporum in T. grandis monoculture, Mariannaea camptospora in Gmelina sp., A. champeden, and Dipterocarp mixed, and Mycoleptodiscus sp. in Macaranga sp. secondary forest. These results indicated that EPF had low or high specificity for host plants, as well as forest sites. This phenomenon was also observed by Kernaghan and Patriquin (2011) in their study in a Boreal area. Kernaghan and Patriquin (2011) isolated root EPF from Betula papyrifera, Abies balsamea, and Picea glauca from two different sites and identified them by a molecular method. They revealed that the EPF communities were different among host trees in one site but not in another site. They also observed that some EPF were found only on certain host, showing the specificity of EPF to host tree species. Jumpponen and Trappe (1998) and Mandyam and Jumpponen (2005) clarified that root EPF, specifically DSE, colonized 587 plants, representing 320 genera and 114 families. Further inoculation experiments under natural and experimental conditions confirmed that DSE species had low host specificity.

The numbers of isolates in Gmelina sp., A. champeden, and Dipterocarp mixed forest and Macaranga sp. secondary forests were higher than those in the other forest sites (Table 4). The dominant species in Gmelina sp., A. champeden, and Dipterocarp mixed forest were Gmelina sp., A. champeden, and Dipterocarp sp. The dominant species in T. grandis monoculture forest and S. macrophylla monoculture forests were only one species each, T. grandis and S. macrophylla, respectively. The number of plant species in each forest might be the reason why the number of isolates was higher in Gmelina sp., A. champeden, and Dipterocarp mixed forest than the other forest sites. The dominant species in Macaranga sp. secondary forest was also one species: Macaranga sp. The reason why the number of isolates was larger in Macaranga sp. secondary forest than T. grandis monoculture forest and S. macrophylla monoculture forest is not known.

Our results demonstrate that the root EPF community differed among forest sites. Utilization of the trap culture method with the same host plant, P. falcataria or S. bicolor, still yielded different EPF among the five forests. Thus, differences in the root EPF community were mainly owing to differences in forest sites, involving the plant community and environmental factors, that resulted in the specific conditions in each forest.

4.2 Role of EPF from the same genera

Some of the root EPF isolated in the present study have also been isolated and studied for their importance, especially in relation with plant growth. Some were proven to promote plant growth under biotic and abiotic stress. Trichoderma sp. improved plant resistance against fungal pathogens (Vinale et al. 2008; Mukherjee et al. 2012), plant tolerance to salt stress (Brotman et al. 2013), and plant tolerance to drought (Bae et al. 2009). Exophiala sp. improved plant tolerance to heat stress (Khan et al. 2012) and cadmium toxicity (Wang et al. 2016). Talaromyces pinophilus promoted the growth of rice seedlings by gibberellin production (Khalmuratova et al. 2015). Penicillium citrinum improved plant resistance against fungal pathogens (Ting et al. 2012) and promoted growth by gibberellin production (Khan et al. 2008). Conversely, Mycoleptodiscus sp. was recorded to be a pathogen of Eurasian watermilfoil (Shearer et al. 2011). There is no information available regarding other remaining genera.

4.3 EPF isolation rate

The isolation of root colonizing microbes by trap culture is widely used in microbiological studies. However, the isolation from field-collected plants is a more common method in EPF studies, especially in studies that aim to isolate organic compounds for biotechnological applications (Strobel 2003). The trap culture method is common for studies that aim to clarify the role of EPF in protecting plants against soil pathogens and promoting plant growth (Amin 2013; Narisawa et al. 2002, 2007) or to clarify the existence of certain EPF in the field (Ahlich et al. 1998).

Studies of root EPF by some isolation methods might underestimate the number of species compared to studies that directly assess root samples by using molecular methods. However, the isolation method has one advantage, namely, the availability of culture for further studies. Brock et al. (2009) emphasized the importance of a herbarium or a fungal culture for the detailed assessment of unknown fungi.

In the present study, the trap culture method was adopted, and EPF were isolated from seven plant species grown on five forest soils. There was no difference in the EPF isolation rate among the four tree species (Table 3), but there was a difference among crop species, with S. bicolor showing the highest isolation rate. Among the seven plant species, S. bicolor showed the highest EPF isolation rate. Narisawa et al. (2002) isolated EPF by the trap culture method using eggplant, Chinese cabbage, tomato, melon, and strawberry as host plants. Among these five plant species, eggplant showed the highest EPF isolation rate. In our study, only representatives of EPF having the same morphology which emerged from root segments were isolated. In contrast, in the study of Narisawa et al. (2002), all emerging EPF were isolated. Thus, the EPF isolation rate in our study might be very low compared to that of Narisawa et al. (2002). In addition, there were also studies that isolated EPF from plant species that were taxonomically close to plant species of the present study: Sesbania bispinosa (Sreelalitha and Sridhar 2015), P. falcataria (Amin 2013), and Trifolium subterraneum (Mugerwa et al. 2013).

P. falcataria and S. bicolor were not originally part of the plant community in the five forest sites. Nevertheless, EPF were still isolated from their roots. Moreover, different EPF were isolated from different forests, indicating that the trap culture method can be used to evaluate differences in EPF community among forests. Considering that both host plants were not part of the plant community in the five forest sites, the effect of original plant species on the identity of isolated EPF was minimal. Therefore, differences in EPF isolates among the forest sites reflect the natural differences in the EPF community among the forest sites.

4.4 Relationship between soil chemical properties and EPF community

Soil chemical properties differ between forests in Java Island and Kalimantan Island. Different isolates were found in the two islands: Fusarium solani in Java Island and Talaromyces sp. in Kalimantan Island. There were specific isolates in each forest, that were not separated by the difference of soil chemical properties of two islands, such as F. solani and Talaromyces sp. Soil chemical properties may not be an important factor determining EPF community.

Root EPF were isolated from forest soil with pH (H2O) ranging from 3.87–4.82. The present study is not the only one to find EPF in soil with low pH. Hakim et al. (2015) isolated EPF from the roots of Shorea leprosula and Shorea selanica grown in reddish Latosol, a type of acid soil in the tropics. Göransson et al. (2008) observed EPF colonization in the roots of four woodland grasses (Elymus caninus, Poa nemoralis, Deschampsia cespitosa, and Deschampsia flexuosa) grown in subtropical forest soil with pH (H2O) ranging from 4.20 to 6.00. These findings emphasize the fact that EPF are commonly found in subtropical or tropical forest soils with low pH.

Root EPF were isolated from soil with available P lower than 1.00 mg 100 g soil−1. This value indicated low P availability for plant growth, according to USDA classification and Brady and Weil (2002). Della Monica et al. (2015) inoculated T. repens with four EPF isolates and cultivated the plant in soil:perlite (2:1 v:v) containing 3.42 mg 100 g−1 total P. One of the four EPF, Phialocephala glacialis, increased both P concentration in soil solution and shoot P content in T. repens. Hiruma et al. (2016) inoculated Arabidopsis thaliana with Colletotrichum tofieldiae and cultivated the plant in half-strength Murashige and Skoog medium, containing 0.68 mg 100 g−1 total P. C. tofieldiae increased the shoot fresh weight of A. thaliana. Root EPF isolated in the present study may have the potential to mineralize P and promote plant growth.

5 Conclusion

EPF were isolated from the roots C. calothyrsus, P. falcataria, S. grandiflora, C. siamea, S. bicolor, A. fistulosum, and T. repens, grown in five different forest soils in Indonesia. EPF isolation rates were not markedly different among the tree species. S. bicolor exhibited the highest EPF isolation rate among crop species, and also among all species used in the present study. Some EPF were isolated from more than one forest, and some specifically from only one forest. In addition, the number of isolated EPF from Gmelina sp., A. champeden, and Dipterocarp mixed were larger than that from other forests. These results indicated that EPF had specificity to the plant species. Future studies including EPF inoculation of plants is necessary to clarify the effect of EPF on plant growth.