Introduction

Fungi play a crucial role in leaf litter decomposition and contribute to soil respiration, nutrient recycling, and build-up of soil organic matter (Swift et al. 1979). This is because they decompose the lignocellulose matrix in litter that other organisms are rarely able to decompose (Cooke and Rayner 1984). Tropical forests have been investigated for the diversity and species composition of litter-inhabiting macro and microfungi (Bills and Polishook 1994; Lodge and Cantrell 1995; Lodg 1996, 1997; Sharma et al. 1995; Paulus et al. 2003, 2006b; Rambelli et al. 2004; Santana et al. 2005) and the succession of fungal assemblages during leaf litter decomposition (Promputtha et al. 2002; Tokumasu and Aoki 2002; Tang et al. 2005; Paulus et al. 2006a). Nevertheless, we still know little about the functional roles of fungi in the decomposition of lignin in tropical leaf litter (Osono 2007). This is despite the fact that the decomposition process of leaf litter in tropical soils is characterized by more rapid removal of lignin (Musvoto et al. 2000; Hirobe et al. 2004b), which leads to faster decomposition of leaf litter and lower accumulation of soil organic carbon than in temperate soils (Takeda 1998).

Osono (2006a) recently developed an approach to examine the abundance and activity of ligninolytic fungi in tropical and temperate forests with reference to the occurrence of bleached portions on leaves of broad-leaved woody species. The presence of bleached portions is associated with fungal colonization of leaf tissues and decomposition of lignin (Osono and Takeda 2001; Koide et al. 2005a, b). Osono (2006a) showed in a preliminary survey that the bleached area on the leaf surfaces was generally greater in tropical forests than in temperate forests. The bleaching agents in temperate forests included some ascomycetous genera and their anamorphs and basidiomycetes, but the agents in tropical forests were yet to be identified. More studies are necessary to explain the occurrence of the bleached area and fungal populations responsible for lignin decomposition in leaf litter on tropical forest soils.

The purpose of this study was to examine fungal populations responsible for lignin decomposition in fallen leaves in a subtropical evergreen broad-leaved forest in southern Japan. Leaves of Castanopsis sieboldii (Makino) Hatusima, a dominant fagaceous tree in the forest, were studied, because this litter is characterized by the occurrence of bleached portions (Fig. 1). Fallen leaves of C. sieboldii collected from the forest floor was examined for bleached area and chemical composition of the bleached and surrounding nonbleached portions of the same leaves to evaluate lignin decomposition by fungi. Fungi were isolated and hyphal abundance was determined in the bleached and nonbleached portions to describe fungal assemblages associated with the decomposition and bleaching of C. sieboldii leaves. Fungi were also isolated from green leaves to test the possibility that fungi from live leaves can persist after litter fall and take part in the decomposition of the litter. Finally, major fungal species were examined their ability to decompose lignin in sterilized leaves under pure culture conditions to identify fungal species responsible for the bleaching of the litter.

Fig. 1
figure 1

Bleached leaf litter of Castanopsis sieboldii. Bar = 1 cm

Full size image

Methods

Study site and sample collection

Sampling was conducted in an evergreen broad-leaved subtropical forest in the University Forest, University of the Ryukyus, in the northern part of Okinawa Island in the Ryukyu Islands, south-western Japan (26˚9′N, 128˚5′E, ca 250–330 m a.s.l.). The mean annual temperature was 22°C and the annual precipitation was 2456 mm during the years between 1992 and 1999 (Yona Experimental Forest, University of the Ryukyus). Minimum temperature was 5.4°C in January and minimum mean monthly precipitation was 130 mm in December. Typhoons frequently strike the island between July and October. Foliage litterfall peaks in March and August (Xu et al. 1998). The topography is hilly and dissected. The bedrock is composed of sandstone and slate, and yellow soil has developed. The forest stand was dominated by C. sieboldii, Schima wallichii ssp. liuliuensis (Nak.) Bloemb. with a maximum height of 20 m (Enoki 2003). Logging and other forest practices have not occurred for at least 50 years. Decomposition of leaf litter of the dominant tree species, including C. sieboldii, has been studied previously at the study site (Alhamd et al. 2004; Xu et al. 2004; Xu and Hirata 2005).

C. sieboldii leaves were collected in December 2004 and December 2006 in a forest stand adjacent to the study plot (200 m × 200 m) established by Enoki (2003). In December 2004, six quadrats (15 cm × 15 cm) were set on the forest floor along a 11-m transect at two-meter intervals, and C. sieboldii fallen leaves were collected from the forest floor within the quadrats and used for measurement of leaf area and leaf mass per unit area (LMA), and for chemical analysis. A total of 20 fallen leaves on which bleaching was evident were collected around the quadrats and used for isolation of fungus. A total of 20 healthy-looking green leaves attached to five C. sieboldii trees 2–4 m in height were also harvested around the transect and used for isolation of fungus. In December 2006, ten sets of bleached leaves of C. sieboldii, each including 30–40 leaves, were collected from the forest floor along the transect at one-meter intervals and used for hyphal length estimation. The leaves were placed in paper bags, preserved at room temperature (ca 15–20°C), and taken back to the laboratory.

Measurement of leaf area and leaf mass per unit area

Leaves for leaf-area measurement were pressed between board papers and oven-dried at 40°C for 1 week. The leaves were then photocopied and scanned with a photoscanner (Epson GT-8000). The area of bleached portions and total leaf area were measured (Koide et al. 2005a) by image analysis performed on a Macintosh computer using public domain NIH image software (written by Wayne Rasband, US NIH). The bleached area as a proportion of total leaf area was expressed as a percentage and mean values of the proportions were calculated. Leaf disks (6 mm in diameter) were then punched from the bleached area and from the surrounding nonbleached area of the same leaves, by use of a cork borer. Ten to 14 disks were obtained for each quadrat. The disks were oven-dried again at 40°C for 1 week to calculate LMA. The disks were then combined to make one sample each of bleached and nonbleached parts from each quadrat, ground in a laboratory mill to pass through a 0.5-mm screen, and used for chemical analysis.

Chemical analysis

The lignin content in the samples was estimated by gravimetry, according to a standardized method using hot sulfuric acid digestion (King and Heath 1967). Samples were extracted with alcohol–benzene at room temperature (15–20°C), and the residue was treated with 72% (v/v) sulfuric acid for 2 h at room temperature with occasional stirring. The mixture was diluted with distilled water to furnish a 2.5% sulfuric acid solution and autoclaved at 120°C for 60 min. After cooling, the residue was filtered and washed with water through a porous crucible (G4), dried at 105°C, and weighed as acid-insoluble residue. The filtrate (autoclaved sulfuric acid solution) was used for total carbohydrate analysis. The amount of carbohydrates in the filtrate was estimated by means of the phenol–sulfuric acid method (Dubois et al. 1956). One milliliter 5% (v/v) phenol and 5 mL 98% (v/v) sulfuric acid were added to the filtrate. The optical density of the solution was measured by use of a spectrophotometer at 490 nm, using known concentrations of D-glucose as standards. Total N content was measured by automatic gas chromatography (Sumigraph NC-900 NC analyzer; Sumitomo Chemical, Osaka, Japan). The methods are described in detail elsewhere (Osono 2004). The term “lignin” is commonly used for the acid-insoluble material as determined by the sulfuric acid digestion method. Although the lignin fraction contains not only true lignin but also lignin-like materials such as melanins and humic substances produced during decomposition, in this study, for the sake of simplicity, the term “lignin” is used for the suite of lignin and lignin-like materials.

Isolation of fungus

A surface disinfection method (Kinkel and Andrews 1988) and a modified washing method (Tokumasu 1996) were used for isolation of fungi from green and fallen leaves of C. sieboldii (Osono 2005). Fungus was isolated 4 days after harvesting. Two leaf disks were punched from a single green leaf with a sterile cork borer (6 mm in diameter), making a total of 40 disks from the 20 green leaves; 20 of the 40 disks were used for the surface disinfection method and the other 20 for the washing method. Using a sterile cork borer, two leaf disks were punched from the bleached portion of a single fallen leaf and another two disks from a nonbleached portion of the same leaf, making a total of 80 disks from the 20 fallen leaves; 40 of the 80 disks were used for the surface disinfection method and the other 40 for the washing method.

For surface disinfection, leaf disks were submerged for 30 s in 70% (v/v) ethanol, to wet the surface, then surface-disinfected for 15 s in a solution of 15% (v/v) hydrogen peroxide and submerged for 30 s in 70% ethanol. The disks were rinsed with sterile, deionized water, transferred to sterile filter paper in Petri dishes (9 cm in diameter), and dried for 24 h to suppress vigorous bacterial growth after plating (Widden and Parkinson 1973). The disks were placed on 9-cm Petri dishes containing lignocellulose agar (LCA) modified in accordance with Miura and Kudo (1970), two disks per plate. LCA contains glucose 0.1%, KH2PO4 0.1%, MgSO4.7H2O 0.02%, KCl 0.02%, NaNO3 0.2%, yeast extract 0.02%, and agar 1.3% (w/v).

For modified washing, leaf disks were washed in a sterile test tube and agitated in a vertical shaker for 1.5 min to isolate fungi growing actively on the surface. The disks were washed serially in five changes of 0.005% (w/v) Aerosol-OT (di-2-ethylhexyl sodium sulfosuccinate) solution and rinsed five times with sterile distilled water. The washed disks were treated in the same manner as that used in the plating-out procedure for the surface disinfected disks.

Plates were incubated at 20°C in the dark and observed 3 days and 2, 4, and 8 weeks after surface disinfection. Any hyphae or spores on the plates were subcultured on fresh LCA plates, incubated, and identified. Identification was based on micromorphological observations, with reference to Domsch et al. (1980) and Ellis (1971, 1976). Two fungal morphotaxa which were isolated frequently from live or fallen leaves were identified by DNA analysis according to the method described by Hirose and Osono (2006). The taxon of Lachnocladiaceae sp. and Valsaceae sp. were determined on the basis of the DNA sequences of 28S rRNA gene D1/D2 region of the isolates using primers D1 (Peterson 2000) and NL4 (O’Donnell 1993) (DDBJ accession numbers AB334107 and AB334108, respectively). The frequency of occurrence of individual species was calculated as a percentage based on the number of leaf disks with the species of the 20 leaf disks tested for each leaf type with each isolation method. A fungal species was arbitrarily regarded as frequent when the frequency of occurrence of the species determined by either of the two methods was more than or equal to 30% on any of green leaves, bleached portions of fallen leaves, or nonbleached portions of fallen leaves. Fungal species other than the frequent species are mentioned only when of special interest.

Hyphal length estimation

Bleached and nonbleached portions of leaves from ten quadrats were cut separately and used for hyphal length estimation by means of the agar-film method of Jones and Mollison (1948) but with several modifications (Koide et al. 2005a). Samples (1.0 g fresh weight) were homogenized in a blender at 10,000 rev/min in 49 ml deionized water for 3 min. The suspension (20 ml) was diluted with 20 ml molten agar solution (final concentration 1.5% w/v) and mixed at low speed on a magnetic stirring plate. Two agar films were prepared for each suspension in a hemocytometer (0.1 mm depth), transferred to glass slides, and dried for 24 h. The films were stained with fluorescent brightener (FB) for 1 h. The FB binds to chitin in the fungal cell wall (West 1988) and enables visualization of all hyaline hyphae.

The stained films were mounted between slides and coverslips with one drop of immersion oil (type DF; Cargille Laboratories, Cedar Grove, NJ, USA) and examined with a Nikon Microphot-SA epifluorescent microscope equipped with a high-intensity mercury light source. A Nikon UV-1A filter cube was used for examination of FB-stained hyphae. Darkly pigmented hyphae that were not stained with FB were observed by bright-field microscopy. Microscope fields were selected randomly and 25 fields were observed for each slide at 1,000× magnification. Hyphal lengths were estimated using an eyepiece grid and the grid-intersection method (Olson 1950). Total hyphal length was calculated as the sum of FB-stained hyphal length and darkly pigmented hyphal length. In this study, hyphae with clamp connections were classified as Basidiomycota. It was difficult, however, to estimated the hyphal length of individual basidiomycete species in decomposing leaves. The length of clamp-bearing hyphae in leaves was, therefore, estimated as total biomass of basidiomycetes, even though this may have resulted in underestimation of basidiomycete biomass because the frequency of clamp formation varies between species and because every hyphal segment observed does not necessarily bear clamp connections. Separate litter samples were dried at 40°C for 1 week to convert fresh weight to dry weight.

Pure culture decomposition test

Nineteen isolates of 19 species isolated from C. sieboldii leaves and obtained from culture collections (Table 1) were used for pure culture decomposition testing. Thirteen isolates were obtained from green leaves or the bleached or nonbleached portions of fallen leaves collected in December 2004 at the study site. In addition, Trametes versicolor, Collybia dryophila, and Mycena polygramma were obtained from a culture collection (IFO/NBRC, Chiba, Japan) and used in the previous decomposition tests (Osono and Takeda 2002, 2006; Osono et al. 2003, 2006). Microporus affinis, Xylaria sp.2, and Fistulina hepatica were isolated from sporocarps on decaying wood in Castanopsis forests. These isolates from culture collection and from wood were used for comparisons. Freshly fallen leaves of C. sieboldii were collected in December 2004 from the forest floor of the study site and used for the tests. The leaves were cut into pieces 1 cm in width, oven-dried at 40°C for 1 week, and preserved in a PVC bag until the experiment started.

Table 1 Mass loss of Castanopsis sieboldii leaf litter and organic chemical components, and lignin/carbohydrate mass-loss ratio (L/C) caused by fungi under pure culture conditions
Full size table

The leaves (300 mg) were sterilized by exposure to ethylene oxide gas at 60°C for 6 h. The sterilized litter was placed on the surface of Petri dishes (9 cm in diameter) containing 20 ml 2% agar. Inocula for each assessment were cut from the margin of the previously inoculated Petri dishes on 2% malt-extracted agar (malt extract 2% and agar 2% (w/v)) with a sterile cork borer (6 mm in diameter) and placed on the agar adjacent to the litter, one plug per plate. The plates were incubated for 12 weeks at 20°C in the dark. The plates were sealed firmly with laboratory film during incubation so that moisture did not limit decomposition on the agar. After incubation the leaves were retrieved, oven-dried at 40°C for 1 week, and weighed. The undecomposed initial litter was also sterilized, oven-dried at 40°C for 1 week, and weighed to determine original mass. Four plates were prepared for each isolate, and four uninoculated plates served as a control. Mass loss of litter was determined as a percentage of the original mass, taking the mass loss of control litter into consideration. The duplicated leaves were then combined and used for chemical analyses as described above. Chemical analyses were performed for those fungi that caused mass loss of litter more than or equal to 5% of original litter mass.

Lignin/carbohydrate loss ratio (L/C) is a useful index of substrate-use pattern of each fungal species (Osono et al. 2006). L/C of each fungal species was calculated according to the equation:

L/C = mass loss of lignin (% original lignin mass)/mass loss of carbohydrates (% original carbohydrate mass).

Results

Bleached area, leaf mass per unit area, and chemical composition

Bleached area accounted for 29.7 ± 3.2% (SE, n = 6) of the total area of C. sieboldii fallen leaves at the study site. LMA was significantly (P < 0.001) lower in the bleached area (8.1 ± 0.5 mg/cm2, SE, n = 6) than in the nonbleached area of the same leaves (10.0 ± 0.5 mg/cm2). Lignin content was lower and amounts of total carbohydrates and nitrogen were higher in the bleached area than in the nonbleached area (Table 2).

Table 2 Chemical composition and fungal biomass in bleached and nonbleached portions of Castanopsis sieboldii leaf litter
Full size table

Hyphal length and species composition of fungi

Total hyphal length was significantly greater in the bleached area than in the nonbleached area (Table 2). Darkly-pigmented hyphal length and darkly-pigmented hyphal length as a percentage of total hyphal length were significantly greater in the nonbleached area than in the bleached area (Table 2). In contrast, clamp-bearing hyphal length and clamp-bearing hyphal length as a percentage of total hyphal length were significantly greater in the bleached area than in the nonbleached area (Table 2).

A total of 47 fungal taxa were isolated from C. sieboldii leaves, of which 11 taxa were isolated frequently (Table 3). Geniculosporium sp.1, Valsaceae sp., Pestalotiopsis distincta, and P. neglecta were frequently isolated from green leaves. Lachnocladiaceae sp. was isolated frequently from the bleached area of fallen leaves. Pestalotiopsis neglecta and Cylindrocladium spp. were frequent on the nonbleached area of fallen leaves. Penicillium lividum, Trichoderma viride, T. harzianum, T. hamatum, and Pochonia suchlasporia were frequent on both the bleached and nonbleached area of fallen leaves.

Table 3 Frequency (%) of fungal taxa isolated from green leaves of Casatnopsis sieboldii and bleached and nonbleached portions of leaf litter by the surface disinfection (S) and washing (W) methods
Full size table

Decomposing ability of fungi

Mass loss of C. sieboldii leaves caused by 19 isolates ranged from 0.6 to 55.2% (Table 1). Bleaching of the litter was noticeable on leaves inoculated with Lachnocladiaceae sp. and Xylaria sp.1 obtained from C. siebodii fallen leaves and with T. versicolor, M. affinis, and C. dryophila obtained from culture collection and from wood. Of the 13 isolates from C. sieboldii leaves, the greatest mass loss was caused by Lachnocladiaceae sp., followed by Geniculosporium sp.1, P. distincta, Xylaria sp.1, and Beltrania rhombica. In contrast, mass losses caused by P. suchlasporia, P. lividum, and P. neglecta were lower than 5%. Valsaceae sp., three Trichoderma species, and Cylindrocladium sp. caused intermediate mass loss between 5 and 10%. Trametes versicolor, M. affinis, C. dryophila, and Xylaria sp.2 obtained from culture collection and from wood caused mass loss greater than or at a similar level to those caused by fungi isolated from C. sieboldii leaves, whereas M. polygramma and F. hepatica caused negligible mass loss.

Substantial mass loss of lignin was caused by Lachnocladiaceae sp., T. versicolor, M. affinis, and C. dryophila (Table 1). The lower mass loss of lignin was caused by P. distincta, Xylaria sp.1, T. harzianum, and T. hamatum. Mass loss of lignin was negative for B. rhombica, Valsaceae sp., and Cylindrocladium sp., indicating the net mass increase of acid-insoluble substances during the incubation. Mass loss of total carbohydrates was observed for all fungal isolates that caused mass loss of litter more than or equal to 5% (Table 1). Lachnocladiaceae sp. caused selective delignification with L/C of 2.2 that was greater than L/C of 0.7–1.0 for the other ligninolytic T. versicolor, M. affinis, and C. dryophila (Table 1).

Discussion

The results for frequency of fungal species in the bleached area and pure culture decomposition tests demonstrated that Lachnocladiaceae sp. was primarily responsible for selective delignification and bleaching of C. sieboldii leaves. This is supported by the greater hyphal length of basidiomycetes in the bleached area, suggesting that the bleaching was associated with the activity of basidiomycetes in the Lachnocladiaceae. L/C of 2.2 for Lachnocladiaceae sp. is at the highest of the range previously reported for litter-decomposing fungi from temperate forests (Osono and Takeda 2002, 2006; Osono et al. 2003, 2006). Previous studies also gave results consistent with those in this study that the bleaching of fallen leaves was caused by fungi capable of selective delignification—e.g., Xylaria sp. in Fagus crenata Bl. (Osono and Takeda 2001), Coccomyces sp. in Camellia japonica (Koide et al. 2005a, b), and Coccomyces sp. and Marasmius sp. in Gaultheria shallon (Osono et al. unpublished data). As far as we are aware, this study is the first to demonstrate fungal bleaching of leaves in the tropics.

Geniculosporium sp.1, Xylaria sp.1, Pestalotiopsis distincta, and three Trichoderma species had some ligninolytic abilities. Xylariaceous fungi are known to have ligninolytic ability (Whalley 1996). Pestalotiopsis species have been found to be cellulolytic (Osono et al. 2003; Koide et al. 2005a; Osono and Takeda 2006), but a species in Pestalotiopsis was recently shown to produce a ligninolytic enzyme (Hao et al. 2007). Thus, those fungi detected on the litter can partly participate in lignin decomposition in the bleached area and lignin can be decomposed to some extent in nonbleached portions simultaneously with carbohydrate decomposition. Beltrania rhombica and Cylindrocladium sp. were regarded as cellulolytic species and can utilize nonlignified cellulose in the nonbleached area and delignified cellulose in the bleached area. Penicillium lividum and Pochonia suchlasporia causing negligible mass loss of litter can be regarded as sugar fungi associated with the decomposition products and (or) intermediates produced by other ligninolytic and cellulolytic decomposers, for example Lachnocladiaceae sp. Saito (1960) demonstrated that the growth of sugar fungi can be supported by reducing sugars produced during the selective delignification of leaves by ligninolytic fungi.

The biodiversity and species richness of microfungi in tropical Asian forests have been addressed recently on fallen leaves of Castanopsis fissa (Tang et al. 2005) and other tropical trees (Promputtha et al. 2002; Paulus et al. 2003, 2006a, b). Tang et al. (2005) examined fungal succession on senescent untreated and autoclaved leaves of C. fissa over a four-month period in Hong Kong Island and identified a total of 38 fungal taxa. Paulus et al. (2006b) recorded 31–81 species per tree species on fallen leaves of six trees in an Australian tropical forest. The number of fungal taxa recorded on C. sieboldii is at the low end of this range, i.e. 47 taxa for both green and fallen leaves or 37 taxa for fallen leaves only. This may be partly because of the different method of fungal isolation and the location of study site in previous studies. Paulus et al. (2006b) reported that species accumulation curves against the number of leaves examined did not reach asymptotes when a total of 60 fallen leaves was examined for each of six tree species, suggesting that more species remain to be detected. Thus, the number of microfungal taxa recorded on C. sieboldii could increase when more leaves are examined.

Endophytic phyllosphere fungi of forest trees often take part in the decomposition of lignin in fallen leaves (Osono 2006b). Koide et al. (2005b) reported that endophytic fungi of live leaves of a temperate tree Camellia japonica L. were responsible for lignin decomposition in litter after leaf death. In the current study, Geniculosporium sp.1, Valsaceae sp., P. distincta, and P. neglecta were isolated from green leaves but occurred only infrequently on fallen leaves and (or) had limited ability to decompose lignin in fallen leaves. Thus, it is not probable that these fungi participate in lignin decomposition in fallen C. sieboldii leaves. The current result does not necessarily indicate the insignificant roles of tropical fungal endophytes in the decomposition. Biodiversity of tropical fungal endophytes has been addressed in recent studies (Arnold et al. 2001; Gamboa and Bayman 2001; Suryanarayanan et al. 2002; Arnold 2005; Promputtha et al. 2005; Duong et al. 2006), and extracellular enzymatic activity was detected for tropical endophytes (Lumyong et al. 2002). Promputtha et al. (2007) recently demonstrated that fungal endophytes of the tropical tree Magnolia liliifera existed as saprobes. Further studies are necessary regarding the roles of endophytic fungi in the decomposition in tropical forests.

Lignin content and LMA were lower in the bleached area of fallen C. sieboldii leaves than in the nonbleached area, indicating the removal of lignin by Lachnocladiaceae sp. enhanced the mass loss of leaf tissues and created a small-scale heterogeneity of decomposition within the single leaves. This difference between bleached and nonbleached area has been found on fallen leaves of other tree species on which the occurrence of a bleached area was evident (Osono 2006a). Lignin has been known as a major factor limiting decomposition of leaf litter not only in temperate regions (Berg and McClaugherty 2003; Osono and Takeda 2005) but also in tropical regions (Laishram and Yadava 1988; Arunachalam et al. 1998; Loranger et al. 2002; Hirobe et al. 2004a; Pandey et al. 2007). The level of colonization of fallen leaves by Lachnocladiaceae sp. (i.e. approximately 30% of leaf area available for fungal colonization) was at the highest of the range previously reported from tropical regions (Osono 2006a). Therefore, the relatively rapid decomposition of C. sieboldii leaves on the subtropical forest soil (Alhamd et al. 2004; Xu et al. 2004; Xu and Hirata 2005) is partly attributable to colonization of the litter by Lachnocladiaceae sp. This is in good agreement with the suggestion by Osono (2006a) that the colonization of fallen leaves by ligninolytic fungi plays a crucial role in decomposition processes in tropical forests.

Bleached areas can frequently be observed on fallen leaves of tree species other than C. sieboldii both at the study site and in several other tropical forests (T. Osono personal observation). More studies are necessary to examine whether the result of this study can be applied to fallen leaves of other tree species and to other tropical forest soils. Species richness and functional biodiversity of ligninolytic fungi and the succession of ligninolytic fungi during the decomposition processes also should be studied further in tropical forests.