ASSOCIATED WITH LEAF LITTER DECOMPOSITION OF MICHELIA NILAGIRICA AND SEMECARPUS CORIACEA AT HAKGALA MONTANE FOREST

In Hakgala forest, Michelia nilagirica Zenk. and In forest ecosystems the annual leaf litter Semecarpus coriacea T h w . are two o f t h e major lea f l i t ter is a major to the nutrient cycling contributors t o t h e forest floor. Michelia nilagirica leaves pathway in which a major proportion of the require about 2.4 years for 99% decomposition. I n contrast, Semecarpus coriacea leaves require more t h a n 20 years t o nutrients in the net primary production is complete i t s decomposition. Since fungi are t h e ma in biotic returned to the forest floor.2 The Hakgala forest components which regulate t h e decomposition rates o f lea f l i t ter, studies on t he fungal communities i n both lea f litter types a t various decomposition stages are o f importance. Leaves i n t h e freshly fallen, first, second and third decomposition stages o f both species were collected using t h e litter bag technique. Fungi were isolated from leaf material o f both types using washing and plating method and t h e identifications were carried out using identi f ication keys . Common as well as dif ferent fungal species were isolated from b o t h l e a f m a t e r i a l s . Broome l la a c u t a , C l a d o s p o r i u m cladosporioides, Trichoderma uiride, Pseudobotrytis terrestris, Curuularia lunata , Cylindrocarpon d i d y m u m , Mortierella vinaceae and Nectria coccinea were t he fungal species isolated i n t h e highest frequencies. B. acuta and C. cladosporioides were present throughout t h e decomposition process o f both types o f leaf litter. Aspergillus and Penicillium species were only occasionally isolated. Hence, t h e present s tudy provides a n account o f t h e fungal species associated w i t h t h e decomposition o f lea f litter o f Hakgala montane forest.


INTRODUCTION
Hakgala strict natural reserve is a n upper montane rain forest in the Nuwara Eliya District.The lower elevation of this forest has 97 tree species (over 15  floor receives a large amount of woody and nonwoody litter from various tree species including those mentioned above.The maximum total litter fall (33%) was recorded to be in the June-July period and the minimum (8%) was in the December -January period.The mean total litter fall for the forest floor per year was 7.76 t ha-ly-I of which 80.5% was represented by non -woody material^.^M. nilagirica and S .coriacea are two major leaf litter contributors to the forest floor which take first and third place respectively.The decomposition (t , , , , ) of the fallen dead leaves of M. nilagirica occurs within 2.4 years whereas decomposition of the fallen dead leaves of S .coriacea occurs within a period estimated to be 20.3 years.Decomposition rates of various substrates are regulated mainly by abiotic (climate) and biotic (substrate quality, micro-organisms and soil fauna) factor^.^,^^^Climatic or abiotic factors such as rainfall, temperature and humidity are not helpful in predicting the different decomposition rates of the various types of leaf litter at the Hakgala forest.This is because these various types of litter occur together in the forest floor where differences in moisture and temperature regimes a r e not significant.Therefore the causative factors for the differences in the decomposition rates between M. nilagirica and S .coriacea leaf litter are the biotic factors such as substrate quality, bacteria, fungi and soil fauna.and Semecarpus coriacea Thw.(Ana~ardiaceae).~Fungi a r e t h e most important decomposers of plant material and 75% of the decomposition of various organic substrates are done mainly by different fungal specie^.^No single fungal species is able to use all components of a substrate and i t is well established t h a t a succession of different groups of fungi will appear on different substrate^.^Since fungi play a major role in litter decomposition, a comparative study on the succession of fungi on the leaf litter of M. n i l a g i r i c a and S .coriacea a t various decomposition stages was carried out.There have been only a few recorded studies on the fungal succession on leaf litter in Sri Lanka.This investigation has provided an account of different fungal species (communities) involved in leaf litter decomposition in a Sri Lankan montane forest as well as their changes during the decomposition gradient.

S t u d y plots:
At the lower elevation (c.1750) of Hakgala forest five study plots (50 x 50 m) were located in order to represent five replicate samples.Of these, four plots were set up in pairs with a gap of 100 m between the pairs and the fifth plot was set up at an elevation 10 m below one of the pairs of plots.For the placement of litter bags a random 1 m x 1 m sub plot was marked on the forest floor in each of the five study plots (Figure 1).

Selection of the leaf litter decomposition stages:
To carry out the investigation on fun@ associated with the decomposition of selected leaf types it C7 1 m x 1 m sub plot

I I
Figure 1: Layout of the sampling plots was necessary to select analogous litter decomposition stages of both leaf types M .nilagirica and S .coriacea.This was easy with respect to the leaf litter of Michelia because of t h e rapid decay r a t e (99% decomposition occurring within 2.4 years).The selection of the decay stages for Semecarpus leaf litter posed a problem because of the slow rate of decomposition (99% decomposition in 20.3 year^).^Therefore the following procedure was used for the two types of leaf litter.
Freshly fallen leaves of the two species were initially collected separately in large nets set up under the trees a t each of the five study plots.They were then placed in large sterile polythene bags and transported to the University of Kelaniya.The following procedure was then followed.
For M. nilagirica: The freshly fallen leaves were placed in nylon mesh bags ( mesh size 2 mm, 6 leaves per bag).The bags were taken to the field and placed randomly in each of the five 1 m x 1 m sub plots.Thereafter these bags (2 bags from each sub plot) were sampled 3,6 and 9 months after placement in the field representing the l", 2nd and 3rd decomposition stages respectively.At each sampling time careful, detailed visual observations were made of the decomposing leaves and descriptions of colour, texture and condition were recorded.Mass remaining (%) was also calculated by drying the litter material in an oven a t 80 "C to a constant mass.
For S .coriacea: Using the visual observations made for each decomposition stages of Michelia leaves, the analogous stages of Sernecarpus leaf litter were selected from the forest floor of each of the five study plots.
Preliminary studies on the morphology and mass remainings (dry wt.) of the lsL, 2nd and 3 1 ~ decomposition stages of both leaf types proved that the analogous decomposition stages 'of both leaf types were similar.
Isolation of fungi: Using five replicate samples from each of the four decomposition stages (freshly fallen, lst, 2nd and 3rd) of both leaf types, fungal isolations were carried out according to the washing and plating m e t h ~d .~ Twelve leaf disks (5 mm diameter) were obtained randomly using 4 leaves of each type (3 disks from each leaf) from each of the 5 replicate samples with the help of a sterilized cork borer.They were then transferred into sterilized screw capped bottles (100 ml capacity) containing 25 ml sterilized distilled water and shaken a t 300 revolutions/min for 2 min.This washing process was repeated 20 times using 25 ml portions of sterilized distilled water for each washing.I t was determined t h a t 20 repeated washings were required for the efficient removal of surface contaminants associated with leaf material of both plant species.1°The excess water was then removed from the disks by blotting with sterile filter papers.
Two (1 mm) pieces of leafwere then taken from each disk using a sterilized scalpel (total of 24 1 mm leaf particles from each of the five replicate samples).Leaf particles were then plated (one particle per plate) on 2% malt extract agar supplemented with 0.01% streptomycin sulphate.The plates containing leaf particles were incubated a t room temperature (30 "C) for 7-10 d.Using a dissecting microscope (Olympus, 313368) fungi growing from the leaf particles were isolated into pure cultures using the same medium (2% MEA) without antibiotics.Using the sticky tape method1' and by preparing slide cultures the isolated fungal species were identified using fungal identification keys.l23l3To identify the fungi which did not sporulate u n d e r t h e given conditions, they were kept under UV light for 1,3,6 and 12 h to activate their sporulating abilities.The identification of these fungal spp.were confirmed by comparing with reference cultures available a t t h e D e p a r t m e n t of Biological Sciences, University of Calgary, Canada.

Calculations
Percentage frequencies of isolation of each fungal species were calculated as follows: % frequencey -No. of particles colonized x 100 of Total No. of particles plated Figure 2 shows the number of fungal spp.isolated from the different decomposition stages of the leaves of M. nilagrica and S .coriacea.In the case of freshly fallen leaves of both spp. a higher number of filngal spp.were isolated in lower frequencies.e.g.18 fungal species from Michelia and 14 fungal species from Semecarpus were isolated in percentage frequencies of 0-4% whereas 4 and 9 fungal spp., respectively from Michelia and Semecarpus freshly fallen leaves, were isolated in 4.1-16% frequencies.Three fungal spp.from Michelia a n d 2 fungal spp.from Semecarpus leaves were isolated in 16.1-64% frequencies.No fungal species were isolated from either spp. in a percentage frequency greater than 64%.

As indicated in
Among the fungi isolated from the lst decomposition stage of M. nilagrica.B. acuta was isolated in the highest frequency of (36.67%).

T r i c h o d e r m a uiride, C l a d o s p o r i u m sphaerospermum, Trichoderma pseudokoningii
and Hyaline sterile sp.16 were also isolated in the higher frequencies of 25.83%, 18.33%, 17.5% and 16.67% respectively.B. acuta was isolated in the highest frequency (44.17%) from t h e lst decomposition stage of Semecarpus.The other fungal spp.isolated in higher frequencies from S .coriacea were Mortierella vinacea (16.67%),Fusariunz sp. 3 (12.5%),Hyaline sterile sp.21 (12.5%),Nodulosporium sp. 2 (10.83%) and Cladosporium cladosporioides (10.83%).B. acuta and T .viride were the fungal species isolated from both leaf types a t the 1" decomposition stage (Appendix 1).
The frequency distribution classes of the fungal community in Figure 2 shows that the fungal species from the lst decomposition stage of both leaf types were isolated mostly in the lower frequencies.Seventeen fungal species from Michelia and 10 fungal species from Semecarpus were isolated in frequencies of 0-4%. 10 and 17 fungal species were isolated from Michelia and Semecarpus leaf litter respectively in frequencies of 4.1-16%.Only 5 fungal spp.from Michelia and 2 fungal species from Semecarpus were isolated in the frequencies of 16.1-64%.No fungal spp.were isolated in a mean percentage frequency higher than 64%.
As seen in Figure 2 fungal spp.from the 2nd decomposition stage of both leaf types were isolated in lower frequencies as well as higher frequencies.e.g. 9, 12 and 9 fungal spp.from Michelia were respectively isolated in frequencies of 0-4%, 4.1-16% and 16.1-64% respectively whereas 3,5 and 8 fungal spp.were isolated from Semecarpus leaf litter in the frequencies of 0-4%, 4.1-16% and 16.1-64% respectively.No fungal spp.were isolated in the frequencies greater than 64% from either leaf type.
With reference to the frequency distribution classes of the fungal communities shown in the Figure 2, most of the fungal spp.from the 3rd decomposition stage of both leaf types were isolated in higher frequencies.No fungal spp.from Michelia leaf litter were isolated in the frequency of 0-4% and only 3 fungal spp.were isolated between the frequencies of 4.1-16%.In Semecarpus leaf litter only one fungal spp.was isolated in the frequencies 0-4% whereas one was isolated in 4.1-16% frequency.Five and 4 fungal spp.from Michelia leaf litter were isolated in the frequencies 16.1-64% and greater than 64% respectively whereas 11 and 1 fungal species from Semecarpus leaf litter were isolated respectively in the frequencies of 16.1-64% and greater than 64%.
It was clearly observed that there was an increase of the isolations of fungal species common to both leaf types with the advancement of litter decomposition.

DISCUSSION
The results obtained for the fungal community structure and its changes during the 4 analogous decomposition stages (freshly fallen, l", 2nd and 31d) of M. nilagirica and S. coriacea leaf litter indicated the presence of certain fungal species (A.strictum, B. acuta, Dark sterile sp. 1, C. cladosporioides, N. coccinea, T. piluliferum, T. viride.. ... etc.) that was common to all four stages and other fungal species specific to the four decomposition stages of the two leaf types.Leaf materials of freshly fallen and l", 2nd and 3 1 ~ decomposition stages of both leaf types are present in the L, L1, F1 and F2 layers respectively of the  forest floor materials.Fungal succession on fallen quality such as nitrogen content and the presence leaves of Quetus, Betula and Fraxinus in a broad of inhibitory tannin concentrations.A comparative leaved UK woodland on mull humus, (until study on fungal succession and microbes on leaf macroscopic leaf structure was lost) has shown litter ( A g e r a t u m conizoides, Mallotus differences between the four litter types.14 The philippinensis, Holarrhena antidysenterica and differences between these succession patterns Vitex glabrata) in two degraded tropical forests were largely due to the differences in resource on northeast India has also shown high variations in the fungal communities in the four litter types.15 In the present study, the occurrence of different fungal species in Michelia and Semecarpus leaf litter might be also due to the differences in the leaf quality.The effect of climatic conditions is similar on the fungal spp.active on both types of leaf litter.In t h e freshly fallen a n d lst decomposition stages, the occurrence of fungal species common to both leaf types increased during the latter decomposition stages.This could be due to the differences in t h e structural components in the freshly fallen and slightly decomposed leaves.I n l a t e r stages of decomposition, the structural components become more or less similar and therefore a higher occurrence of fungal spp.common to both leaf types was observed.
Many of the studies on fungal succession on leaf litter have been carried out in climatic and geographical regions different to those from Sri Lanka.Therefore it is not very appropriate to compare the results obtained in the present study with the observations made from such studies.However the study on the fungal succession on Aspen poplar leaf litter in Kananaskis valley in the front range of rocky mountain in Canada where the mean annual temperature is 4"C, showed the presence ofAureobasidium pullulans only on live leaves.Other than this Cladosporium spp.including C. cladosporioides a n d C. sphaerospermum were observed in the live leaves and litter on the L layer and Mortierella spp.were observed in the litter of F layers.16 Studies on the micro fungi of decomposing red alder leaves in a forest north of Copenhagen showed the occurrence of Aureobasidium spp. in the early and middle stages of t h e decomposition process.Cladosporium spp.including C. cladosporioides persisted throughout the decomposition process of the Alder leaves.Mortierella spp.appeared from the middle of the decomposition process and persisted to the latter decomposition stage^.^As in the case of these studies, in the present study too Aureobasidium spp.were seen only in the early stages (freshly fallen and 1" decomposition stage) of Michelia leaf litter which are in the L layer of the forest floor materials but were not observed in the fungal succession on Semecarpus leaf litter.In the present study too, Cladosporium spp.especially C. cladosporioides were observed in several of the decomposition stages of both M i c h e l i a a n d S e m e c a r p u s leaf litter and Mortierella sp.(M.uinacea) was isolated from the later decomposition stages of Semecarpus leaf litter.Although the presence ofAcremonium spp.was observed throughout the decomposition stages of Aspen leaves16, in the present study isolation ofAcremonium spp.were made only from early decomposition stages of both leaf types.
Fungal succession studies carried out in India, have reported the occurrence of many Aspergillus and Penicillium spp.throughout the process of decomposition of different plant litter.Isolation of eight Aspergillus spp.were reported from the different decomposition stages of Chenopodium album leaf litter.17Four Aspergillus spp.were frequently isolated from decomposing materials of sugar cane bagasses.18Three Aspergillus spp.and five Penicillium spp.were frequently isolated from many of the decomposition stages ofMallotus leaf litter.In the decomposing leaves of Holarrhena antydysentrica and V i t e x g l a b r a t a leaf litter, P e n i c i l l i u m chrysogenum and Aspergillus niger were the rnost dominant fungal spp.occurring during the s u c c e ~s i o n .~~ In the present study the occurrence of Aspergillus spp.and Penicillium spp. in the decomposing leaves of both types was very low.Or_e A s p e r g i l l u s spp.was isolated in high frequency (50.84%)only from t h e 3rd decomposition stage of Semecarpus and only four Penicillium spp.were isolated occasionally except for Penicillium variabile.
According to the frequency distribution classes of the fungal communities in the four decomposition stages (Figure 2) it was clearly observed that the fungal isolations from the early decomposition stages of both leaf types occurred in lower frequencies.With the advancement of litter decomposition the fungal isolations occurred in higher frequencies from both leaf litter types.The distribution of fungal species present in the freshly fallen leaves a n d t h e leaves in decomposition stage 1, of both leaf types may be localized into some areas of the leaves.However in the latter decomposition stages of both leaf types, their distribution may be more or less even over the leaf surface in both leaf types.Therefore, this could be the reason for their isolation in lower frequencies from the early decomposition stages and isolation in higher frequencies from the latter decomposition stages of both leaf types.
From the present study it is clear that B. acuta, C. cladosporioides, C. Lunata, N. coccinea, T. piluliferum and T. viride were the dominant fungal spp.isolated from several decomposition stages of both leaf types with higher frequencies.Among them B. acuta was the most dominant fungal spp.isolated from all 4 decomposition stages of both leaf types with higher frequencies.These fungal spp.may have more competitive saprophitic abilities t h a n t h e other litter inhabiting mycoflora in the Hakgala forest floor and also they may have versatile substrate utilization abilities to utilize many of the labile and non labile complex structural cell wall components.
This study reveals t h e p a t t e r n of succession in fungal communities and gives an insight to the nutrient turnover in the ecosystem.An additional advantage is that the identified fungal cultures isolated during this study could be used as reference cultures.Therefore in addition to maintaining the live cultures, a database could be maintained where all the information on this collection of fungal isolates is available.

Decomposing leaves of
cm GBH) of which 62 a r e e n d e m i ~. ~ Some of the common tree species are Syzygium revolutum (Wight.)Walp.(Myrtaceae), Psychotria bisulcata Wight.& Arn.(Rubiaceae), Allophylus varians (Thw.)Redlk.(Sapindaceae), Michelia n i l a g i r i c a Zenk.(Magnoliaceae), Memecylonparvifolium Thw.(Melastomataceae), E u g e n i a m a b a e o i d e s Wight.(Myrtaceae), S y z y g i u m r o t u n d i f o l i u m Arn.(Myrtaceae), Cinnamomum ovalifolium Wight.(Lauraceae)

%
Frequency distribution classes

Figure 2 :
Figure 2: Frequency distribution classes of the fungal community in different stages of leaf decomposition of Michelia nilagirica and Semecarpus coriacea.