Plant Succession Following Nuèes Ardentes of Mt. Merapi

Volcano, Java Indonesia

Sutomo Sarjana in Forestry Science (B.Sc. Hons.)

Gadjah Mada University, Indonesia

This Thesis is presented for the Degree of Master of Science

University of Western Australia

SCHOOL OF PLANT BIOLOGY

FACULTY OF NATURAL AND AGRICULTURAL SCIENCES

2010

ii

STATEMENT OF ORIGINAL CONTRIBUTION

The research presented in this thesis is an original contribution to the field of plant ecology.

The ideas presented and discussed here unless otherwise referenced or acknowledged, are

my own original thoughts. This thesis has been completed throughout my enrolment at the

University of Western Australia and has not been used previously for a degree at any other

institution.

Sutomo

iii

ABSTRACT

Tropical volcanic ecosystems, particularly in Asia, have been relatively

understudied. Mt. Merapi volcano, in Java Indonesia provides an opportunity to study

succession following nuées ardentes events which essentially reset the successional clock

of the montane forest. Information regarding the development of the vegetation on Mt.

Merapi during early succession is scarce. Using a chronosequence approach, my study

documented the patterns of vegetation succession following nuées ardentes and wildfires in

tropical montane forests on Mt. Merapi. I examined primary succession in areas directly

affected by nuées ardentes and secondary succession in areas affected by fire associated

with nuées ardentes.

There was a rapid colonization by vascular plants in both primary and secondary

succession as the sites aged. Imperata cylindrica, Eupatorium riparium, Anaphalis

javanica, Athyrium macrocarpum, Brachiaria paspaloides, Dichantium caricosum,

Selaginella doederleinii, Eleusine indica, Cyperus flavidus, Calliandra callothyrsus and

Acacia decurrens were the species mainly responsible in explaining the differences

between sites. In primary succession, the species richness and diversity reach their peak 14

years after disturbance. In secondary succession, the species richness was similar to the

reference site in only a little after two years - however, the peak of species diversity was 14

years after the nuées ardentes disturbance.

Native and exotic invasive species varied in abundance among sites. The native

invasive I. cylindrica dominated the early succession but then disappeared under the shade

of the emerging tree species. In contrast, invasive exotic species such as Eupatorium spp

and Brachiaria spp remained in the system long after the nuées ardentes had occurred and

forest structure had developed.

iv

In the secondary succession the forest structure developed throughout the

succession with the older sites, regaining nearly complete stratified forest vegetation after

14 years. In terms of species interspecific associations, positive associations were greater

than negative associations as time progressed in the primary succession. The nitrogen-

fixing species Calliandra callothyrsus had the highest number of positive interspecific

associations compared to the other species, which may suggest that this legume species has

a prominent role in facilitation.

An examination of soil nutrient status showed improvement in soil condition as

time progressed. There was a clear pattern of increase in N, C, and exchangeable cation

concentration with age of site, whereas the P concentration decreased with time. There was

also a significant relationship between species composition and the measured soil nutrients

(P, N, Ca++, Na+ Mg++ and K+).

Results from this study have shown that the ecosystem is resilient to volcanic

disturbance as shown by the significant increase in species richness and diversity, increase

in positive species association, and improvement in soil nutrients within 14 years of

disturbance. However, this study had also raised some concern regarding invasive alien

species in the succession. Hence, ecological intervention in the form of weed management

in Mt. Merapi succession should be considered.

v

ACKNOWLEDGEMENT

Praise is to “Allah Subhanahu wataàla” for His blessings and gift of health among

many others blessings for me to be able to finish this thesis. I have been fortunate and

grateful to have received an Australian Development Scholarship, a prestigious scholarship

awarded by the Australian Agency for International Development (AusAID). Gratitude also

goes to its staff, Ms. Megan Mayne and AusAID scholarship officer at the University of

Western Australia Ms. Deborah Pyatt and AusAID representative staff in Indonesia, Mr.

Riza Reinanto and Ms. Trilia Dianti for their kind support throughout the end of my study.

To my supervisors, Professor Richard Hobbs and Dr. Viki Cramer go my sincere

gratitude. It has been a great honour for me that they were willing to accept me as their

student and gave me the opportunity to work with them. Richard’s brilliant ideas in ecology

have always been influential to me and I enjoyed every minute of his presentations. I thank

Viki, for her patience for getting into the details of my research, field site visits to

Yogyakarta, Indonesia and her constructive criticisms and ideas throughout my research. I

would also thank to all the crew in the Ecosystem Restoration Laboratory, my office room

mates and fellow research students, Christina, Lauren, and Juan for the beautiful friendship.

My appreciation also goes to the head of school of plant biology Hans Lambers and all the

staff at the school, Dr. Krys Haq for her invaluable feedback on my writing and also to

UWA visiting professor, Pierre Legendrè for his lecture and tutorial on spatial analysis of

multivariate environmental data.

My special thanks to scientists, ecologists, botanists and staff at Murdoch

University. My gratitude goes to Dr. Fiona Valessini, for teaching me the basis of

multivariate ecology and the use of PRIMER software which was widely used in this study.

Dr. Philip Ladd, Dr. Joe Fontaine for the insightful discussions, Professor Ann Hamblin for

vi

her kind support and first comments, discussions and personal communication for Chapter

3 of the thesis. Thanks also to Dr. Helen Allison for the friendship and discussion on

resilience thinking.

To brilliant scientists Professor Pat Kennedy from Oregon State University, for

insightful discussions, her kind support, encouragement and friendship, Professor Franck

Lavigne, from the Universite´ Blaise Pascal, France, for his kind discussion by

correspondence on Mt. Merapi with his surprisingly fluent “Bahasa” and “Javanese”

languages, and Professor Bruce Clarkson from the Waikato University New Zealand, for

visiting my presentation and for a short discussion on volcanoes in the SERI world

conference on Ecological Restoration in Perth, 2009.

I would also like to extend my gratitude to the Indonesian government, Sri Sultan

Hamengkubuwono X, the King and the Governor of the Mataram and Yogyakarta Special

Province, Professor Endang Sukara and Dr. Irawati from the Life Sciences Department and

Bogor Botanical Garden, the Indonesian Institute of Sciences-LIPI, the heads of my

establishment in Indonesia, who permitted me to pursue my study in Australia, my advisor

Ir. Soewarno Hasanbahri MS. from the Faculty of Forestry, Gadjah Mada University for his

invaluable advice, encouragement and support while I was in Yogyakarta to do the

fieldwork, Ir. Arman Mallolongan M.M. from the Directorate of Forest Protection and

Nature Conservation (PHKA) the Indonesian Ministry of Forestry, and Ir. Tri Prasetyo, the

head of the Merapi National Park (BTNGM) for permission to enter the national park and

conduct the field data collections, Mbah Maridjan, the caretaker and gatekeeper of the

Merapi Mountain, Ir. Taufik Tri Hermawan M.Si from the Faculty of Forestry, Gadjah

Mada University for the invitation to host one of his lectures in ecology for undergraduate

students.

vii

I would like to thank also to the diligent volunteers from Gadjah Mada University

for their invaluable help, Gunawan, Ali, Iqbal, Indri, Kuntala, and Titi. We endured the

long and ascendant paths of the mountain, we defeated the biting cold of the mountain

climate, we shared the happiness at lunch time and we made it to cross the steep ravines

and survived being lost without GPS signals and safely made it to the base camp, and those

were memories I treasured.

Finally, I wish to thank my families, my wife Laily and our son Muhammad Hafidz

Imtiyaz Sutomo for their love, patience and company during the difficult periods of my

candidature, my parents Drs. Nana Suhana M.Si. and Susilowati, my sisters Kristina and

Tika Permata Sari for their continuous support and prayers, and lastly my aunt Dyah

Trimurti in Yogyakarta for her kind support ever since I started my journey as an

undergraduate student at Gadjah Mada University in 1999.

Perth, March, 2010

Sutomo

viii

TABLE OF CONTENTS

Abstract iii

Acknowledgements v

Table of Contents viii

List of Figures xi

List of Tables xiii

List of Appendices xiv

Chapter 1 General Introduction 1

1.1. Mt. Merapi Volcano 1

1.2. Nuèes Ardentes and Their Effects on Vegetation 2

1.3. Succession in Plant Communities 5

1.4. The Chronosequence Approach 8

1.5 Research Needs and Significance 10

1.6 Research Aims and Thesis Outline 12

Chapter 2 Mt. Merapi National Park and Study Sites Descriptions 14

2.1. Geographical Position 14

2.2. Management History 14

2.3. Climate 15

2.4. Geomorphology 15

2.5. Soils 16

2.6. Vegetation 17

2.7. Study Sites 18

2.7.1. Sites Selection 18

2.7.2. 2006 Site 21

2.7.3. 1998 Site 22

2.7.4. 1997 Site 24

2.7.5. 1994 Site 24

2.7.6. Reference Site 26

ix

Chapter 3 Establishment of Plant Communities on Nuèes Ardentes Volcanic

Deposits on Mt. Merapi: Composition, Diversity and Interspecific

Associations along a Chronosequence

27

3.1. Introduction 27

3.2. Method 31

3.2.1. Study Sites 31

3.2.2. Sampling 32

3.2.3. Statistical Analysis 32

3.3. Results 34

3.3.1. Species Diversity 34

3.3.2. Variation in Community Composition 35

3.3.3. Interspecific Association (co-occurrence) 40

3.4. Discussion 42

Chapter 4 Community Structure and Composition along a Chronosequence in

Forests Burnt by Nuèes Ardentes-Induced Fire in Mt. Merapi National

Park

49

4.1. Introduction 49

4.2. Method 53

4.2.1. Study Sites 53

4.2.2. Vegetation Sampling 53

4.2.3. Soil Sampling and Analysis 54

4.2.4. Data Analysis 54

4.2.4.1. Species Diversity and Turnover 55

4.2.4.2. Forest Structure 55

4.2.4.3. Variation in Community Composition 57

4.2.4.4. Variation in Soil Nutrients and Relationship with Floristic

Composition

58

4.3. Results 59

4.3.1. Species Diversity and Turnover 59

4.3.2. Forest Structure and Composition 60

4.3.3. Variation in Community Composition 63

x

4.3.4. Variation in Soil Nutrients 66

4.3.5. Relationship between Floristic Composition and Soil Nutrients 66

4.4. Discussion 70

Chapter 5 General Discussion 76

5.1. Introduction 76

5.2. Plant Succession on Mt. Merapi 76

5.2.1. Species Re-colonization and Diversity 76

5.2.2. Species Composition and Structure 77

5.2.3. Exotic Invasives Species 78

5.2.4. Interspecific Association 80

5.2.5. Soil Nutrients 80

5.3. Similarities to other Volcanoes in Asia 82

5.4. Management Implications 85

5.5. Conclusions and Recommendations 87

Appendices 89

References 96

xi

LIST OF FIGURES

Chapter 1

Figure 1.1 A Mt. Merapi nuèe ardente, known by locals as wedhus gembel, moves

down the volcano’s slope to Kaliadem area in Yogyakarta Province

3

Figure 2.1 The 2006 Mt. Merapi nuées ardentes resulted from collapse of the

gegerboyo lava dome at the southwest flank of the volcano that had

destroyed vegetation and houses in the village surrounding the Gendol

River in Kaliadem area

4

Chapter 2

Figure 2.1 Sketch of morphological unit of the Merapi strato-volcano 16

Figure 2.2 Map of sampling sites in Mt. Merapi National Park 20

Figure 2.3 Collapsed lava dome (the gegerboyo) at the southwest flank of Mount

Merapi in June 2006

21

Figure 2.4 The 2006 site condition showing the deposits of the nuées ardentes

(primary succession site) and the seared trees of Pinus merkusii (secondary

succession site)

22

Figure 2.5 Primary succession on the 1998 nuées ardentes deposits 23

Figure 2.6 The 1998 secondary succession site’s condition 23

Figure 2.7 The 1997 secondary succession site’s condition showing Imperata

cylindrica, few Melastoma sp and one tree fern of Cyathea sp.

24

Figure 2.8 The 1994 secondary succession site’s condition 25

Figure 2.9 The primary succession on the 1994 nuées ardentes deposit 25

Figure 2.10 An old secondary forest in Kaliurang as a reference site 26

Chapter 3

Figure 3.1. Deposit 2006-Steep and deep abrupt valleys, formed in the Kaliadem area

following the 2006 Mt. Merapi devastating nuées ardentes

31

Figure 3.2. NMDS of sites based on vegetation composition and abundance 35

Figure 3.3. Total number of interspecific association of species in each deposit of

primary succession sites of Mt. Merapi

40

xii

Figure 3.4. Seedling of Pinus merkusii on the 2006 deposit. This is the type of

situation at Mt Merapi in which recruitment from seed was taking place

45

Chapter 4

Figure 4.1. NMDS of sites based on vegetation composition and abundance 63

Figure 4.2. LINKTREE diagram 69

Chapter 5

Figure 5.1. Resilience of Merapi volcanic ecosystem 84

xiii

LIST OF TABLES

Chapter 2

Table 2.1. Site location, nuées ardentes history and environmental information in each

study site at Mt. Merapi National Park

19

Chapter 3

Table 3.1. Differences in species richness and diversity between sites on four nuées

ardentes deposits

34

Table 3.2. ANOSIM pairwise test of vegetation plots within sites 36

Table 3.3. SIMPER result 1. Percentage contribution of species to average Bray–

Curtis dissimilarities in all pairs of sites

38

Table 3.4. SIMPER result 2. Average abundance percentage in each site of the

selected species from table 2 (SIMPER 1)

39

Table 3.5. Results of the association tests using the chi-squared test statistic (χ2)

between the differentiating species and their co-occurring species

41

Chapter 4

Table 4.1. Differences in species diversity between the burnt sites and reference site

in Mt Merapi National Park

59

Table 4.2. Species turnover rates (D) between pairs of sites in the chronosequence on

Mt Merapi

60

Table 4.3. Number of individuals in each stratum for each site of secondary

succession at Mt. Merapi

60

Table 4.4. Importance Value Index (IVI), and shade tolerance for the most important

species in each stratum at each of the study sites

62

Table 4.5. ANOSIM pairwise test of NMDS vegetation plots ordination 64

Table 4.6. SIMPER result. Percentage contribution of species to average Bray–Curtis

dissimilarities in all pairs of sites

65

Table 4.7. Global test BEST. Combinations of edaphic factors that best constrain the

community composition within the Merapi National Park study sites

66

Table 4.8. Summary of differences in concentrations of soil nutrients in secondary

forest and the reference site on Mt. Merapi

67

xiv

LIST OF APPENDICES

Appendix 1. Mean abundance of species per family in the primary succession study

sites

89

Appendix 2. Mean abundance of species per family in the secondary succession

study sites

91

Appendix 3. Sketch of Eupatorium odoratum 94

Appendix 4. Sketch of Eupatorium riparium 95

1

Chapter 1

General Introduction

1.1. MT. MERAPI VOLCANO

Volcanic activity is a major natural disturbance that can catastrophically change an

ecosystem over a short time (Scheffer et al. 2001). Indonesia has 130 active volcanoes

(Weill 2004), and the island of Java is the most volcanically active island in the world, with

about 20 of its volcanoes having been active in the historical period (Whitten et al. 1996).

This study examines the patterns of vegetation succession on Mt. Merapi after volcanic

disturbance.

Mt. Merapi is one of the major active volcanoes on the Island of Java, located

approximately 30 kilometres north of Yogyakarta Province, in Central Java (Figure 2.2).

The name ‘Merapi’ is derived from the Javanese words ‘Meru’ (mountain) and ‘Api’ (fire).

The first eruption of Mount Merapi was noted in 1006. The eruption, described in an

ancient manuscript of Mataram Kingdom, the “Maha Pralaya Mataram”, covered all parts

of Central Java with ashes, and buried the Borobudur temple (Bemmelen 1970; Whitten et

al. 1996).

Mount Merapi has been reported as one of the most active volcanoes in the world

with 83 recorded eruptions (Gomez et al. 2009; Newhall et al. 2000). Hence, Mt. Merapi is

Indonesia’s most frequently erupting volcano (Simkin and Siebert 1994), with small and

frequent nuées ardentes occurring between large infrequent eruptions (Bardintzeff 1984).

Mount Merapi has varying intervals of eruptions. Small eruptions occur at an interval of 2-

5 years, medium scale eruptions occur every 7-10 years (MVO 2006) and large explosive

eruptions usually occur once in a 100 years (Newhall et al. 2000). Mt. Merapi is become

2

one of the most monitored Indonesian volcanoes (MVO 2006; Thouret and Lavigne 2005;

Voight et al. 2000).

1.2. NUÉES ARDENTES AND THEIR EFFECTS ON VEGETATION

Nuées ardentes are known amongst the local people of Mt. Merapi in Yogyakarta as

‘Wedhus Gembel’, which literally means ‘a sheep ran down from the mountain’ (Figure

1.1). A nuèe ardente (French for “glowing cloud”) is the “rapid movement of extremely hot

(often more than 700° C) turbulent gases and fragmental material across a land surface from a

volcanic vent. The denser, basal part of a pyroclastic flow hugs the ground and follows topography,

moving with great force and speed (up to 200 km/h)” (Dale et al. 2005b p. 249). Merapi has

different characteristic of nuées ardentes compared with other volcanoes, and it has become

the reference volcano for its nuées ardentes, known as the ‘Merapi type’. The nuèes

ardentes of Merapi are the product of a collapsed lava dome at the summit (Takahashi and

Tsujimoto 2000). In the Merapi-type nuées ardentes, gravity plays a significant role

(Bardintzeff 1984). In other volcanoes, such as Mt. St. Helens in the USA, the nuées

ardentes do not originate from a collapsed lava dome, but instead are blasted out from side

vents of the mountain due to the escalating volcanic activity (Bardintzeff 1984; Belousov et

al. 2007).

A dome is built continuously, and the topographic position of this ‘new’ dome

determines whether, where and when it will collapse. The topographic position also

determines the frequency of eruptions (Chiu et al. 2001). Even though nuées ardentes

usually descend along the land surface, they are also capable of moving uphill, leaping over

ridges as they move away from the volcano’s crater. The 1994 nuèe ardente on Mt Merapi

moved 6.5 kilometres south-southwest and deposited approximately 2.5–3 million cubic

metres of material (Abdurachman et al. 2000). The most recent nuées ardentes occurred in

3

2006. On 4th June 2006, the “Geger boyo” flank in Kaliadem (Sleman District, Yogyakarta

Province) collapsed and nuées ardentes occurred until 14th June. The flows moved down

the slope through Gendol River (Kaliadem area), and destroyed all vegetation and buildings

in its path (Figure 1.2).

Figure 1.1. A Mt. Merapi nuèe ardente, known by locals as wedhus gembel, moves down the

volcano’s slope to Kaliadem area in Yogyakarta Province. Photo courtesy of Merapi Volcanology Observatory/BPPTK (2006).

Volcanic eruptions have an important impact on the environment (Marti and Ernst

2005). Every year, approximately 60 volcanoes erupt on Earth, and even though almost

80% of them occurring under the oceans, terrestrial volcanic eruptions commonly cause

great impact on vegetation over large areas (Dale et al. 2005b). In East Java, Mt. Bromo-

Tengger’s violent eruptions created an un-vegetated area of ash near one of the best known

caldera called the “Sea of sand”. Similarly the 1963 violent eruption of Mt. Batur in Bali

4

created hard lava stream areas cleared of vegetation in which regeneration could take more

than one hundred years (Whitten et al. 1996).

Figure 1.2. The 2006 Mt. Merapi nuées ardentes resulted from a collapsed gegerboyo lava dome

at the southwest flank of the volcano that had destroyed vegetation and houses in the village surrounding the Gendol River in Kaliadem area. Photo by author.

Six types of volcanic activity can have complex effects upon vegetation: lava

formation, nuées ardentes, debris avalanches, lahar, tephra and blowdown (Dale et al.

2005b). Rough and hot heterogeneous mass material ejected from the nuées ardentes would

likely have an impact on any vegetation in its path due to poisonous gases, extreme heat or

by burial (Dale et al. 2005b; Kelfoun et al. 2000). Standing dead trees, damaged or buried

plants, and fine ash coverage of plant leaves are some of the impacts of nuées ardentes on

vegetation. In one of the few studies that has examined the effect of nuées ardentes on

vegetation, Kelfoun (2000) found that there were four types of effects to the tree vegetation

following the 1994 Merapi nuées ardentes: singed trees; broken trees; blown down trees;

and buried trees. Fine ash, stuck to tree foliage, can have a variety of effects. Antos and

5

Zobel (2005) which worked on tephra falls approximately 30 km from the cone of Mt. St.

Helens found that in the long term, fine ash coverage on tree foliage could reduce tree

growth due to the inhibition of the photosynthesis. The initial impact on understorey plants

is much more devastating, as they are buried under any material deposited with the flow.

Furthermore, Antos and Zobel (2005) also reported that the majority of herbaceous species

could penetrate upward through ≤ 4.5 centimetre deposits, but a 15 centimetre deep deposit

almost eliminated the herb layer. The heat released from nuées ardentes can ignite

wildfires. Intense fire will kill or damage some of the plants, but others are more resistant

and depend on fire for their regeneration (Bond and Wilgen 1996). Adaptations to fire

usually involve a plant’s capability to endure fire and to rejuvenate after the fire (Bond and

Wilgen 1996). According to Uhl (1990) species with thick bark and other protective

features exhibit greater survival. Following the devastating eruption of Mt. Semeru in East

Java in 1918, the singed Casuarina junghuhniana trees remained standing and re-sprouted

vigorously (Whitten et al. 1996). Albizia lopantha and Pinus merkusii are other examples

of tree species in Java where fire enforces their regeneration, breaks the hard seed case and

allows seeds to germinate (Whitten et al. 1996).

1.3. SUCCESSION IN PLANT COMMUNITIES

Succession is a central concept in ecology and the phenomenon of changes in

species composition over time has captured the interest of ecologists and naturalists for

centuries (McLean 1919; Smith 1914; Walker et al. 2007). Plant species establishment and

composition changes in a newly formed substrate with no biological legacy are known as

primary succession, while secondary succession refers to the species changes in an habitat

with relatively intact soil following a disturbance (Walker and del Moral 2003). Succession

6

model is a conceptual framework to elucidate the past successional trajectory and predict

the future trajectory by taking into account various mechanisms and the various stages of

the successional pathway (Walker and del Moral 2003).

The classical model proposed by Clements (1916) suggests that community

development occurs in distinct successional stages towards a “climax” community based on

processes of nudation (creation of a barren substrate following a disturbance); migration

(plant arrival on the already decaying site); ecesis (plant establishment, either generative or

vegetative); competition (plant interactions that eliminate some species) and, reaction (site

modification by plants). Later, Eggler (1954) proposed another model which emphasized

the importance of differential growth rates and longevities of species. In this model, early

successional species may be present in the late seral stages but probably will be less

abundant or slowed in their growth. Another model was proposed by Connell and Slatyer

(1977) which derived from the literature. The model was based on three concepts that affect

succession: facilitation, tolerance and inhibition. In the facilitation model, the establishment

of later species requires earlier species to ameliorate the site and thus these species become

the nurse plants for the later species. In the tolerance model, the later species that survive

will be the species that could tolerate the lower level of nutrients and other resources due to

competition with other species. Lastly, the inhibition model hypothesizes that the first

colonizer will continue to occupy the space and monopolize the resource, thus inhibiting

the establishment of the subsequently arriving species. This condition persists until the first

occupant species dies out and releases the resources that allow later species to establish and

reach maturity. More recent emphasis has shifted towards the idea of non-equilibrium

dynamics where it is recognized that disturbance can occur at any of those stages and at a

rate that makes a "climax" community unattainable (Suding and Hobbs 2009).

7

Li et al. (1999) stated that many succession theories were based on intensive work

in temperate forests. However, plant succession has also been described in the tropics

(Brearley et al. 2004; Finnegan 1996; 1981; Pena 2003). These studies have resulted in

various patterns of succession in terms of species richness, diversity, plant biomass and also

soil changes. However some general patterns may also appear in succession (Quesada et al.

2009; Radosevich et al. 2007). Increases in plant diversity, cover and biomass with time

have been observed in succession (Aplet et al. 1998; Dale et al. 2005c; del Moral 2000;

Finnegan 1996; Pena 2003; Tsuyuzaki and Hase 2005; Zhu et al. 2009). Common patterns

are also observed in the differences of the characteristics of plants which dominate each

stage of succession (Wills 2002). Early seral stages are usually dominated by plants with

rapid growth, small seeds, and a short life span (e.g. ruderal species). Late seral stages are

dominated by species with slow growth, often with large seeds and longer life spans

(Radosevich et al. 2007).

The way that species composition changes through time is described as the

successional trajectory. Walker and del Moral (2003) summarize some of the prominent

types of trajectory that may be found in a succession:

1) Convergence theory argues that random patterns of species composition at the

start of succession will be narrowed down and become more similar as time

progresses (Baer et al. 2005; del Moral 2007; Inouye and Tilman 1988).

2) A divergence pattern is identified when vegetation composition becomes more

dissimilar as the succession proceeds (Bush et al. 1992; Lepš and Rejmanek

1991).

3) Several trajectories can combine to form a network. Network trajectories occur

when there are several different invasive pioneer species that dominate and each of

the species can initiate different trajectories (del Moral et al. 2010).

8

With increasingly fragmented landscapes, convergence patterns are rare but can still be

found (although there are some limiting factors) in some primary succession sites such as

on Mt. St. Helens. Their chance of occurring increases where there are strong biotic

interactions (del Moral 2007; Walker and del Moral 2003). There are several methods

which can be used to examine a succession trajectory. A commonly used method is the

chronosequence approach (space for time substitution) and by the approach of direct

observation in the permanent plots (Walker and del Moral 2003).

1.4. THE CHRONOSEQUENCE APPROACH

There are various methods for studying succession, such as permanent plots,

chronosequence, air photo interpretation and palynology (Walker and del Moral 2003).

Permanent plots, and a chronosequence approach, are the most common methods used in

succession studies. Direct observation in permanent plots, combined with experimental and

manipulative study, is the most appropriate method to apply (when possible), in order to

understand the mechanism of succession. However, permanent plot studies which last

longer than 10 years are rare (Foster and Tilman 2000). When timing (i.e. the succession

measured in scales of decades or even centuries) and logistics are the main concerns and no

aerial photos are available, an indirect approach using a chronosequence method is

preferred. The idea is essentially to choose a series of plots showing various stages of

disturbance or abandonment and thus presume that they characterize different stages of an

analogous successional development (Foster and Tilman 2000; Gomez-Pompa and

Vazquez-Yanes 1981; Walker and del Moral 2003).

9

In order for the chronosequence approach to be useful in inferring patterns of the

succession over time, there are prominent assumptions that need to be considered. Firstly,

the environmental conditions such as the climate, substrates and topographical positions of

the sites under study need to be relatively similar (Durán et al. 2009; Johnson and

Miyanishi 2008; Pickett 1989). Secondly, these sites should also have the same disturbance

history and have similar pre-disturbance conditions. This means that the past disturbance

history should be relatively similar or at least randomised if it is unknown. When these

assumptions are met then the chronosequence approach can be a powerful tool to use in the

study of succession (Durán et al. 2009; Johnson and Miyanishi 2008; Wills 2002; Zhu et al.

2009).

There are potential disadvantages of the chronosequence approach (Johnson and

Miyanishi 2008; Milberg 1995; Pickett 1989). Pickett (1989) stated that the

chronosequence approach is disadvantageous because of its averaging property which could

hide the short term dynamics of the examined system. Meanwhile, Foster and Tillman

(2000) argued that it is this particular property which yielded broad successional trends

makes the chronosequence approach advantageous. Myster and Malahy (2008) conclude

that the decision whether to use the chronosequence approach or not will depend largely on

the parameters measured and that species richness, diversity and total plant abundance

perform best with the chronosequence approach. Foster and Tilman (2000) support the use

of a chronosequence approach to infer basic patterns in succession based by re-sampling a

chronosequence plots over an interval of time because the rate of the succession can be

measured.

10

A chronosequence approach has been applied in many types of ecosystems in both

primary succession and secondary succession. The method has been used on primary

successional seres, especially on volcanic terrain, by Vitousek et al. (1993) in Hawaii.

Meanwhile the method had also been widely applied in a wide range of sites following fire

(Gonzalez-Tagle et al. 2008; Watson and Wardel-Johnson 2004) and land abandonment

(Baniya et al. 2009; Myster and Pickett 1992a; b). The chronosequence approach has been

used to serve a specific purpose to examine the vegetation dynamics (Wills 2002; Zhu et

al. 2009), soil nutrient status and biomass (Bautista-Cruz and del Castillo 2005; Bormann et

al. 1995; Durán et al. 2009; Wang et al. 2009) and forest structure (Spencer et al. 2001).

The results of these studies suggest that using a chronosequence approach remains a robust

method to gain the fundamental patterns of changes in succession.

1.5. RESEARCH NEEDS AND SIGNIFICANCE

Montane forests such as those on Mt. Merapi have become increasingly important

for conserving Indonesia’s biodiversity (Whitten et al. 1996). The importance of the

tropical montane forest for Indonesia is understandable, considering that much of its

tropical lowland forest has been cleared for more than 30 years (Whitten et al. 1996).

Paddy fields and plantations are now common in lowland landscapes in Java, replacing the

primary lowland forest that previously dominated these landscapes. In the northern part of

Java no single forest remnant remains. In the southern part only a few patches of lowland

natural forest remain such as those in Ujung Kulon and Pangandaran National Parks in west

Java. In the southern parts of central and east Java all lowland forests were converted to

plantations. Consequently, pristine forests in Java can only be found on mountains (van

Steenis 1972; Whitten et al. 1996). Forests in Merapi are also subjected to destruction due

11

to increasing volcanic activity of Merapi as well as prolonged drought and human

disturbances (Anonym 1999; 2004; Heru 2002; MVO 2006; Newhall et al. 2000).

Increasing temperatures during the El Nino Southern Oscillation (ENSO) in 1992 caused

wildfires in this area and the escalating volcanic activity led to an eruption in 1994 which

caused catalytic destruction to the forest over 650.30 hectares (Dinas Kehutanan DIY

1999).

Rehabilitation and management of a degraded ecosystem, including a volcanic

ecosystem, will require specific knowledge of the process of succession (Walker et al.

2007; Walker and del Moral 2003). However, the scientific knowledge in this field is

limited, particularly in volcanic tropical montane forests, where scientific studies about the

effects and the ecological consequences of disturbances at high-elevation are still scant

(Horn et al. 2001; Tsuyuzaki and Hase 2005; Whittaker et al. 1999). In contrast, there is

abundant of evidence regarding the needs of ecological intervention to accelerate the rate of

succession in lowland tropics (Brearley et al. 2004; Lindig-Cisneros et al. 2006; Zahawi

and Augspurger 1999). Understanding the dynamics of forest ecosystems during succession

and how they interact with disturbances, will be an important component of forest

ecosystem conservation and restoration management practices (Hobbs et al. 2007; Swamy

et al. 2000). However, these key elements are often neglected in Indonesia and elsewhere.

In addition, although there is a large body of literature available on the subject of

the geology and vulcanology of Mt. Merapi (Berthommier and Camus 1991; Camus et al.

2000; Gomez et al. 2009; Lavigne 1999; Newhall et al. 2000), ecological research on Mt.

Merapi has received scant attention. To date there has not been any comprehensive research

on the dynamics of the vegetation (native and exotic species) following volcanic

disturbance (but see Hardiwinoto et al. 1998). Therefore, the present study is significant in

12

terms of generating useful baseline data for the management of volcanic ecosystems in the

region and elsewhere in Indonesia and Asia.

In summary, this study will:

1) Be the first project to carry out a study on plant community succession on Mt.

Merapi in both primary and secondary succession.

2) Increase the knowledge and theoretical background in succession and dynamics of

a tropical montane forest following volcanic disturbance.

3) Contribute to the efforts of linking ecological succession knowledge with

restoration/rehabilitation practices.

1.6. RESEARCH AIMS AND THESIS OUTLINE

This study uses a chronosequence approach to examine the vegetation succession on

Mt. Merapi after volcanic disturbance. A range of sites that have been affected by nuées

ardentes at different times are available on Mt Merapi. While acknowledging the

limitations imposed by site variables other than time since disturbance, the sites used still

provide a sequence of different ages which can form the basis for further studies

Overall, the aims of the present study are to:

1) Evaluate how the vegetation changes over time in terms of the structure, species

composition and diversity in primary and secondary succession sites.

2) Examine whether early interaction patterns of the species in the primary

succession can be identified by examining their interspecific association.

3) Examine whether there were significant differences in major soil nutrients

between sites of varying ages and whether these edaphic factors are correlated

with the species assemblages in the secondary succession sites.

13

4) Evaluate these findings and discuss the implications for the management plan of

the Mt. Merapi National Park.

This thesis consists of five chapters. In chapter 2, the regional context of the Merapi

National Park and the study sites is briefly described with a particular emphasis on

management history, conservation, vegetation formation and environmental condition such

as climate, soil and geology. In this chapter I also describe the criteria and rationale for the

choice of sites. The next two chapters are the topics that are covered by this thesis. Chapter

3 focuses on the establishment and succession of pioneer plant communities on the nuèe

ardente deposits (primary succession), which also includes detection of early interaction

patterns of the species in the succession by examining their interspecific association.

Chapter 4 examines changes in floristic composition in the secondary succession which

also takes into account the identification of a subset of edaphic factors that best correlates

with the species assemblages in the course of the secondary succession. Lastly, Chapter 5

summarizes the knowledge gained from the previous topic chapters and discusses the

findings in the framework of management planning of the national park. It also proposes

potential ecological restoration (intervention) activities and outlines further research.

14

Chapter 2

Mt. Merapi National Park and Study Sites Descriptions

2.1. GEOGRAPHICAL POSITION

Mt. Merapi (7º 35’ S and 110º 24’ E) is administratively located in two provinces,

Central Java (Magelang, Boyolali and Klaten Districts) and Yogyakarta (Sleman District).

In Yogyakarta Province, Mt. Merapi is located approximately 30 kilometres north of

Yogyakarta. Mt. Merapi is representative of the landforms, soils and vegetation on a

volcanic mountain that typify a large portion of montane ecosystems in Java (Whitten et al.

1996).

2.2. MANAGEMENT HISTORY

In 1931, the Dutch colonial government established the forest area located in

Central Java Province (6,962.1 ha) and Yogyakarta Province (1,510 ha) on the slope of

Merapi as protected forest (Act no. G. B. No 4197/B, 4 May 1931). The appointed

protected forest was meant to protect the local natural resources and the surrounding area.

In 1975, the Indonesian Ministry of Agriculture, the department that managed Indonesian

forests at that time, issued Act no. 347/Kpts/Um/8/1975, which assigned some parts of the

protected forest as nature reserve (198.5 ha) and tourism forest (30 ha) named CA/TWA

Plawangan-Turgo. Since 1975, villagers were prohibited from living in the protected forest

in the kumpulrejo and patuk areas because of the escalating threat from the volcano. At this

time, management was taken over by the joint Forestry and Agriculture Provincial office of

Yogyakarta. The Merapi Forest area located in Yogyakarta Province is managed by

Yogyakarta Forestry Provincial Office based on Governor Act no. 336/1974. On 4 May

15

2004, The Minister of Forestry issued decree no. 134/MENHUT-II/2004 regarding the

amendment of forest area functions of protected forests, nature reserves, and tourism forests

to become the Merapi National Park, which covers an area of 6,410 ha (Anonym 2004).

2.3. CLIMATE

Based on Schmidt and Fergusson’s climate classification (1951), the Merapi area is

classified as type B - tropical monsoon area - which is characterized by a high intensity of

rainfall in the wet season (November-March) with a dry season that can often be very dry

without any rainfall (April-October). Annual precipitation varies from 2,500–3,500

millimetres (Anonym 2004). The variation of rainfall on Mt Merapi’s slope is influenced by

orographic precipitation. As in many other tropical monsoon areas, there are minor

temperature and humidity variations during the year. Relative humidity on Mt Merapi

varies from 70%-90%, with daily average temperatures from 19° to 30° C (Dinas

Kehutanan DIY 1999).

2.4. GEOMOPRHOLOGY

Mount Merapi is a basaltic-andesite volcano whose formation is known through five

periods: the pre-Merapi period (>400,000 years ago), old-Merapi period (between 400,000

to 6,700 years ago), middle-period Merapi (6,700 – 2,200 years ago), young-Merapi

(2,200-600 years ago) and the Merapi we know today which began 600 years ago (Gertisser

and Keller 2003; Newhall et al. 2000). Based on its geomorphology, Mt. Merapi can be

divided into five units: volcanic cone, volcanic slope, volcanic foot plain and fluvio-

volcanic plain (Pannekoek 1949) (Figure 2.1). Mt. Merapi’s strato-volcano summit

continually changes due to dome growth and collapse (typical of the Merapi-type nuées

16

ardentes; see chapter 1) and in 2006, the Merapi summit was 2,911 metres (MVO 2006).

The topography of the Mt Merapi varies from undulating to mountainous. There are two

bedrock materials that comprise Mt. Merapi (Anonym 2004):

1. Bedrock of the new Merapi, consisting of tufa, lahar, endeictic and basaltic lava. This

deposit is widely spread on Mt Merapi.

2. Old Merapi volcanic bedrock, which is found locally in some places, especially in hill

areas such as on Turgo, Plawangan, Gono and Maron hills.

Figure 2.1. Sketch of morphological unit of the Merapi strato-volcano (Pannekoek 1949)

A spring belt occurs when groundwater moves from recharge area to the discharge

area at the slope break (Irawan and Puradimaja 2006). Figure 2.1.showed that there were

three break of slope-spring belt on Mt. Merapi which is the source of excellent quality

water resource for the lower areas (Irawan and Puradimaja 2006; Pannekoek 1949).

17

2.5. SOILS

Soils of the study area are mainly of young volcanic-ash origin (regosol) with

shallow and/or deep, low to medium fertility solums with a profile not yet developed

(Anonym 2004; Darmawijaya 1990). The soil textures are granulated, whereas the

structures are crumbly (Anonym 2004; Dinas Kehutanan DIY 1999).

2.6. VEGETATION

Based on his extensive work in Java’s volcanic mountain forests, van Steenis (1972)

described plant formation patterns that generally occupied the montane forests. Primary

forests occur both below and above 2,000 metres. High primary forest occurs at 1,000-

2,000 metre elevation. This forest contains tall and emergent trees species up to 25-35

metres such as Altingia excelsa, Schima wallichii and Vernonea arborea. Primary forest

that occurs at an elevation greater than 2,000 metres is named elfin or low primary forest.

This type of forest is structured by small trees and shrubs up to 10 metres in height such as

Pittosporum spp., Rhododendron spp., and Vaccinium spp.

When these primary forests experience a disturbance that damages or kills

vegetation (fire, volcanic activity), then secondary forests developed. Secondary forest that

occurs at elevation below 2,000 metres usually includes invasive grass and shrub species

such as Imperata cylindrica, Eupatorium odoratum, Lantana camara, and Saccharum

spontaneum and the tree species Homalanthus populneus, Macaranga javanica, Vernonea

arborea, Villebrunea rubescens, Casuarina junghuhniana and Pinus merkusii. Disturbed

elfin forests generally include species such as Albizia spp., Anaphalis spp. (some Anaphalis

are also can be found in lower elevation secondary forest), Engelhardia spp., Dodonaea

spp., and Lespedeza junghuhniana.

18

2.7. STUDY SITES

On Mt. Merapi, areas which have been completely buried by the nuées ardentes

deposits undergo primary succession. These areas usually occur along the streams, channels

or valleys created by the solid material flow paths of nuées ardentes. The secondary

succession areas were located adjacent to the primary succession areas. These areas are the

adjacent forest on either side of the valley, or deposit channel which escapes burial and

mainly scorched by the extreme temperature of the nuées ardentes.

2.7.1. Site Selection

Sites or areas of different ages (years since last nuée ardente) were selected to

obtain a chronosequence. Identification of site age was conducted by studying aerial

photographs, topography maps, and nuée ardente history maps (obtained from the Merapi

Volcanology Observatory) to date sites affected by recent nuées ardentes. Identification

was also conducted by reconnaissance study, interviewing long - term residents of the

surrounding villages, personal communication with the national park’s ranger and

managers, and also field site visits. Sites also had to show no obvious signs of human

disturbance and be at least 50 metres from any human activities or structures. Based on

these, I chose four sites that were affected by nuées ardentes at different times (2006, 1998,

1997, and 1994) and one forest area that was mostly undisturbed and had not been affected

by nuées ardentes for at least 50 years as a reference site (Figure 2.2). The five sites were

located in a lower montane zone and were located at a range of altitudes from 1,000 to

1,600 metres. Chronosequence assumptions were met within these sites as they had similar

environmental conditions such as climate, substrate, topography and geomorphology. Four

sites were low-slope (6-13°) while one site was of intermediate-slope (28-29°). The fieldwork

19

was conducted from March to August 2008. Summary of the average environmental

conditions in each site is described in table 1.

Table 2.1. Site location, nuées ardentes history and environmental information in each study sites at Mt. Merapi National Park.

Location Year of nuée ardente

Site age (years) Soil type Elevation

(m)

Average Slope

(°) Kaliadem 2006 2 Regosol 1,220 12.2

Kalilamat 1998 10 Regosol 1,579 28.26

Kalibedog 1997 11 Regosol 1,207 6.53

Kalikuning 1994 14 Regosol 1,180 6.23

Kaliurang - - Regosol 1,000 13

20

Figure 2.2. Map of sampling sites in Mt. Merapi National Park. Ellipses show the location of the sampling sites in the nuées

ardentes affected areas. The rectangle shows the location of the reference site. The green spot in the insert of the map of Java Island is the approximate location of Mt. Merapi on the island

21

2.7.2. 2006 Site

The 2006 site was located in Kaliadem, at an altitude of 1,220 metres. This site is

the youngest site (2 years old) in terms of time since last nuèe ardente. On 4 June 2006, the

Geger Boyo Flank in Kaliadem (Sleman District, Yogyakarta Province) collapsed (Figure

2.3) and nuées ardentes occurred until 14 June 2006. The flows moved down the slope

through the Gendol River, with areas dominated by Pinus merkusii heavily affected. The

area that was affected by the nuèe ardente flows formed steep and deep abrupt valleys.

Deposits from the nuées ardentes were estimated to reach 5.6 million cubic metres

(PVMBG 2006). In this site, the damage caused by the flows was still clearly visible -

seared and standing or buried dead trees of P. merkusii located around the centre, the area

that was near to the Gendol River (Figure 2.4).

Figure 2.3. Collapsed lava dome (the gegerboyo) at the southwest flank of Mount Merapi in June 2006. Arrow sign showing the remains of the track created by the collapsed dome

(MVO 2006)

22

Figure 2.4. The 2006 site condition showing the deposits of the nuées ardentes (primary succession site) and the seared trees of Pinus merkusii (secondary succession site)

2.7.3. 1998 Site

The 1998 Merapi eruption on 11 and 19 July 1998 produced 8.8 million cubic

metres of nuèe ardente deposits, in places up to 8 metres deep. Nuèes ardentes flow down

towards rivers such as Sat, Lamat and Senowo which are located west side of Mt Merapi, 6

kilometres from the crater. Located at an altitude of around 1,500–1,600 metres, the 1998

site was the most difficult to reach. This site is situated in an area of one of the most

dangerous tracks of volcanic flows - the Lamat River which belongs to the Central Java

section of the national park. The distance from the crater was around 2 kilometres and very

remote with a steep slope of 28°. Figure 2.5 and 2.6 shows sites that have undergone

primary and secondary succession.

23

Figure 2.5. Primary succession on the 1998 nuées ardentes deposits

Figure 2.6. The 1998 secondary succession site’s condition

24

2.7.4. 1997 Site

The nuées ardentes that occurred in 1997 flooded the Krasak, Bebeng, Boyong and

Bedog Rivers, which left deep and abrupt marks of valleys in the south-west flank and 2

million cubic metres of material deposits. The sampling sites for primary and secondary

succession were located in the areas surrounding the Bedog River at an altitude of 1,207

metres in the Yogyakarta Province. Figure 2.7 shows the secondary succession.

Figure 2.7. The 1997 secondary succession site’s condition showing Imperata cylindrica, few

Melastoma sp and one tree fern of Cyathea sp.

2.7.5. 1994 Site

In 1994, the collapsed dome initiated surges of nuées ardentes down to the Boyong

River which reached the middle of Turgo and Plawangan Hill and Kuning River,

Yogyakarta Province. Turgo and Plawangan are hills that are believed to be the remains of

what vulcanologists call old-Merapi, or the beginning of the Merapi that we see today

(Newhall et al. 2000). The estimated volume of the slide was 2.6 million cubic metres and

25

the deposits were almost 6 metres deep. The 1994 sampling site is located in an area

surrounding the Kuning River, at an altitude of ± 1,180 m. Figure 2.8 and 2.9 shows areas

that have undergone secondary and primary succession.

Figure 2.8. The 1994 secondary succession site’s condition

Figure 2.9. The primary succession on the 1994 nuées ardentes deposit

26

2.7.6. Reference Site

To compare the development of plant communities with a site that had not been

disturbed by the scorching and fire associated with nuées ardentes, I chose a reference site

named Kaliurang (Figure 2.10), which is located on the south flank of Mt Merapi, near the

Plawangan and Turgo Hills, approximately 4-5 kilometres from the crater in Yogyakarta

Province. Kaliurang is located at 7º35’ S and 110º24’ E, at an altitude approximately 1000

metres. Generally the forest condition around this site is intact and is characterized by the

presence of large trees (diameter between 40 and 100 centimetres) of primary species such

as Altingia excelsa. Some gaps due to natural tree felling were noted. In some locations, the

ground was covered by ferns, Araceae, and Palmae. However, this reference site is actually

an old growth secondary forest and it is not a “pristine” forest. This forest has been likely

been damaged or cleared (whether by volcanic and or anthropogenic disturbance) and has

re-grown at some time in the past.

Figure 2.10. An old secondary forest in Kaliurang as a reference site

27

Chapter 3

Establishment of Plant Communities on Nuèes Ardentes Volcanic Deposits

on Mt. Merapi: Composition, Diversity and Interspecific Association

along a Chronosequence

3.1. INTRODUCTION

Primary succession is the establishment of plant species and subsequent change in

community composition over time on a substrate with little or no biological legacy (Walker

and del Moral 2003). Volcanic eruptions pose a significant challenge to the study of

primary succession compared with other disturbances (i.e. erosion, landslides, floodplains

and glaciers) because of the absence of a biological legacy following the eruption (Franklin

et al. 1985). One type of volcanic disturbance is nuées ardentes or pyroclastic flows. Nuèes

ardentes are hot turbulent gas and fragmented material resulting from a collapsed lava

dome that rapidly moves down the volcanic slope (Dale et al. 2005b). The accumulation of

this material is called a nuées ardentes deposit and it may be up to ten metres thick

(Franklin et al. 1985). Such pure mineral deposits preserve no “memory” of previous

vegetation owing to the absence of a seed bank (Thornton 2007). Hence, colonization must

occur from other undisturbed places.

The phenomenon that creates the disturbance to a vegetation community and the

process of recovery that occurs plays a significant role in the dynamics of the community

and its species diversity (Connell and Slatyer 1977; Crain et al. 2008). The establishment of

vegetation on volcanic deposits has been documented in many parts of the world such as in

USA, Italy and Japan, with rates of establishment being shown to vary (Aplet et al. 1998;

Dale et al. 2005c; Eggler 1959; Tsuyuzaki 1991). For example, plant establishment and

spread on the debris-avalanche deposit of Mt. St. Helens were slow during the first years

28

after eruption (Dale et al. 2005c). In contrast, Taylor (1957) reported that six years after

Mt. Lamington in West Papua erupted, vegetation regeneration was very rapid, and

included species such as Saccharum spontaneum, Imperata cylindrica, Pennisetum

macrostachyum, Vitaceae and several ferns. Mt. Krakatau had at least 64 vascular plant

species (dominated by grass species such as Saccharum spontaneum and Imperata

cylindrica) that colonized the island three years after the eruption (Thornton 2007).

The pace of primary succession on volcanoes is usually thought to be slow,

especially in the early stage, but as abiotic processes develop on barren substrates, the rate

of succession increases (del Moral and Wood 1993). The initial conditions following

disturbance will determine the trajectory of plant establishment. Survival and recovery of

plants following volcanic disturbance will be severely affected by the type of volcanic

activity, nutrients transported by the volcanic disturbance, distance from seed sources and

the types of propagules from nearby undisturbed areas (Dale et al. 2005a; del Moral and

Wood 1993; Walker and del Moral 2003).

The study of succession may assist in recognizing the possible effects of species

interactions (i.e. facilitation or inhibition) (Connell and Slatyer 1977; Walker et al. 2007).

Facilitation promotes establishment and in the context of succession, facilitation can be

defined as any role plants play in influencing changes in species composition to the next

stage (Walker and del Moral 2003). In contrast, inhibition basically refers to the negative

effect that one species has on another species (Walker and del Moral 2003). A species

inhibits the growth or establishment of the co-occurring species through resource

competition or by directly excluding subsequent species by allelopathy effect (Collins and

Jose 2009; Soerjani et al. 1983; Walker and del Moral 2003). Previous studies have shown

that facilitation is more prominent in severely disturbed habitats, whereas competition tends

to be more significant in more productive and established habitats (Callaway and Walker

29

1997; Walker et al. 2007). The barren landscapes created by volcanic disturbance provide

excellent opportunities to examine the role of pioneer species in facilitating or inhibiting

the later species in the succession (Morris and Wood 1989; Walker and del Moral 2003).

However, much remains unknown about the initial interactions during succession that drive

the subsequent community composition (Bellingham et al. 2001; Connell and Slatyer

1977).

Observations of species co-occurrence may be seen as the first attempt to identify

species interactions and niche processes that structure the community (Walker and del

Moral 2003; Widyatmoko and Burgman 2006). By determining which species are strongly

associated with of co-occurring species (through field observation), experimental studies of

species interactions may become more effective, and may help in understanding the role of

species interaction throughout succession (Myster and Pickett 1992b).

A commonly used method in studying primary succession is the chronosequence

approach or space for time substitution (Walker and del Moral 2003). An alternative

approach is to establish permanent plots to carry out such research (Cramer et al. 2008; del

Moral 2007; del Moral and Wood 1993; Simbolon et al. 2003). Although there has been

some criticism of the chronosequence approach (see chapter 1) (Herben 1996; Johnson and

Miyanishi 2008), it is still a useful method, especially when timing and logistics are a

problem (Aplet et al. 1998; Myster and Malahy 2008). Thus, a chronosequence approach

can provide an initial approach to understanding the fundamental patterns of changes in

succession (Foster and Tilman 2000). The nuées ardentes deposits found on Mt. Merapi

present an opportunity to study succession using a chronosequence approach. The deposits

at these sites are relatively young, with the last known eruptions occurring between 1994

and 2006. The sites are not strictly analogous in terms of their slope and landscape position,

and hence care must be taken in interpreting the results (del Moral and Ellis 2004).

30

My goal was to understand the patterns of plant re-establishment in the early

successional stage on the nuées ardentes deposits on Mt. Merapi, in order to increase the

understanding of plant successional patterns caused by these unique, though ecologically

understudied Merapi-type nuées ardentes deposit materials. My research questions were:

1) How does species composition and diversity vary across deposits of different ages? 2) Which species contributed the most to the differences in composition between different

aged deposits?

3) Could early patterns of species interactions be identified by examining their

interspecific association?

31

3.2. METHOD

3.2.1. Study sites

My research sites were located in the south-west flank of Mt. Merapi, in the zone of

the Merapi National Park. These sites are more often affected by nuées ardentes, which

tend to flow down the hills in this direction. I chose four areas that were affected by nuées

ardentes between 1994 and 2006. The four deposit sites were located in a lower montane

zone at a range of altitudes from 1000 to 1500 metres (Montagnini and Jordan 2005) (for

detailed description and the map of the deposit sites, see Chapter 2). The condition of the

2006 deposit, two years after the nuées ardentes disturbance can be seen in figure 3.1.

Figure 3.1. Deposit 2006-Steep and deep abrupt valleys, formed in the Kaliadem area following

the 2006 Mt. Merapi devastating nuées ardentes. Photo was taken in April 2008.

32

3.2.2. Sampling

In 2008, I sampled vegetation on the four nuées ardentes deposits. I sampled ten

circular plots (diameter range approximately 10 metres) in each of the four deposit sites,

assigned at random to grid cells on a map (Dale et al. 2005a; Zimmerman et al. 2008). The

position and altitude of each site were recorded using a GPS (Garmin E-Trex legend) and

slope was measured using a clinometer (Suunto PM-5/360 PC Finland, clinometer).

I measured plant abundance as density, a count of the numbers of individuals of a

species within the quadrat (Endo et al. 2008; Kent and Coker 1992). Local plant names and

Latin names, when known, were noted. Whenever there was any doubt about a species

name, a herbarium sample was made. Drying and sample identification were carried out in

Dendrology Laboratory, Faculty of Forestry, Universitas Gadjah Mada. Vascular plant

nomenclature was based on “Flora of Java” (Backer and van den Brink 1963), “Mountain

Flora of Java” (van Steenis 1972), and International Plant Names Index Databases (IPNI

2008). .

3.2.3. Statistical analysis

Data were analysed using multivariate and univariate statistics. Multivariate

analyses such as Non Metric Multidimensional Scaling (NMDS), Analysis of Similarity

(ANOSIM), Similarity Percentage (SIMPER), were conducted using PRIMER V.6 (Clarke

and Gorley 2005). Univariate analyses such as ANOVA and post hoc tests were conducted

using SPSS package V.11.5.

Species richness and Shannon-Wiener species diversity on each deposit were

calculated. Changes in these values across the deposits were tested for significance using

one-way ANOVA test. I tested differences in community composition between deposits

using data on species abundance (density) per plot. The data was square root transformed

33

prior to constructing a resemblance matrix based on Bray-Curtis similarity (Valessini

2009). An NMDS ordination diagram was then generated based on the resemblance matrix.

Ordination results were then overlaid with the results of a cluster analysis, using 20%

similarity to show grouping of plots in each deposit and separation between different

deposits. Variation in community composition between deposits was subsequently tested

for significance using one-way ANOSIM (analysis of similarity). ANOSIM is basically

analogous to standard univariate ANOVA, and tests a priori defined groups against random

groups in ordinate space. The RANOSIM statistic values, generated by ANOSIM, are a

relative measure of separation of the a priori defined groups. A zero (0) indicates that there

is no difference among groups, while one (1) indicates that all samples within groups are

more similar to one another than any samples from different groups (Clarke 1993). The

SIMPER routine was then used to explore the relative contribution of individual species to

dissimilarity among deposits.

Interspecific association between species was measured using the chi-square (χ2)

test of the species presence/absence data on a 2 x 2 contingency table (Kent and Coker

1992; Ludwig and Reynolds 1988). The significance of the chi-square test statistic is

determined by comparing it to the theoretical chi-square distribution (P = 0.05, df = 1).

Based on this test, there are two types of association. A positive association occurs if χ2 test

> χ2 theoretical, that is, the pair of species occurred together more often than expected by

chance. Negative association occurs if χ2 test < χ2

theoretical, that is, the pair of species occurred

together less often than expected by chance. The strength of the association was measured

using the Ochiai index which gives a value of 0 at ‘no association’ and 1 at

‘complete/maximum association’.

34

3.3. RESULTS

3.3.1. Species diversity

Fifty two species belonging to 23 families were recorded in the four sites (Appendix

1). The highest number of species belonged to the Asteraceae (14), then Poaceae (11),

followed by Fabaceae (4). Species richness was significantly different between deposits

(ANOVA P = 0.05, table 3.1). Species richness was lowest in the 2006 deposit and was

highest in the oldest deposit (1994). Species diversity was lowest in the 1998 deposit and

highest in the 1994 deposit (Table 3.1). The two youngest deposits (2006 and 1998) had

significantly lower species diversity than the two older deposits (1997 and 1994) (ANOVA

P = 0.05; table 3.1). Despite a difference in deposit age of only one year, the species

diversity was significantly lower in the 1998 than the 1997 deposit.

Table 3.1. Differences in species richness and diversity between sites on four nuées ardentes deposits. Superscript letters (a-b) after mean values (±SD) indicate significant differences between sites as assessed with Tukey’s HSD test. Time of eruption/age of deposit at sampling

Species richness Species diversity

2006 site/2 years 4.0 (±1.49)a 1.03 (± 0.3)a

1998 site/10 years 6.7 (±1.41)b 0.95 (± 0.34)a

1997 site/11 years 7.4 (±1.07)bc 1.38 (± 0.16)b

1994 site/14 years 8.4 (±2.11)c 1.61 (± 0.22)b

35

3.3.2. Variation in community composition

Figure 3.2 showed NMDS ordination that clustered the sites into three groups . The

right cluster consists of the plots from the youngest deposit (2006), and the left cluster of

plots from the 1998 deposit. The middle cluster contains plots from the 1997 and 1994

deposits. This cluster is slightly intersected with the left cluster. An analysis of similarity

(ANOSIM) test confirmed that there were significant differences in Bray-Curtis species

similarities between deposits on Mt. Merapi (Global RANOSIM = 0.91, P < 0.001). Pairwise

comparison tests between sites generally showed R values ranging from 0.9 to 1.0 (Table

3.2) except for the comparison between the 1997 deposit and the 1994 deposits, which had

an R-value of 0.42.

Figure 3.2. NMDS of sites based on vegetation composition and abundance of the 2006 deposit (triangles), 1998 deposit (inverse triangles), 1997 deposit (squares), and 1994 deposit (diamonds).

Ellipses indicate groups resulting from cluster analysis.

36

Table 3.2. ANOSIM pairwise test of vegetation plots within sites. Sample statistic (Global R): 0.91, significance level of sample statistic P < 0.001, number of permutation: 999. A zero (0) indicates that there is no difference among groups, while one (1) indicates the groups are very different with each other. Groups R Statistic

2006-1998 1

2006-1997 0.97

2006-1994 0.99

1998-1997 0.94

1998-1994 0.96

1997-1994 0.42

Twelve species (Anaphalis javanica, Athyrium macrocarpum, Calliandra

callothyrsus, Cyperus flavidus, Eupatorium riparium, Imperata cylindrica, Panicum

reptans, Paspalum conjugatum, Pinus merkusii, Polygala paniculata, Polyosma ilicifolia,

and Polytrias amaura) were most responsible for the dissimilarity between the sites (Tables

3.3 and 3.4). The species that contributed to the dissimilarities among almost all pairs of

deposits were A. javanica, E. riparium, and Polytrias amaura, while I. cylindrica was the

species that was responsible for dissimilarities among all pair of deposit sites (Table 3.3).

The average of the Bray-Curtis dissimilarities between deposit 2006 and 1998 is

98% and this is made up of five species. Athyrium macrocarpum contribution is 10.27% of

the total 98%; E. riparium gives 6.85% of this total. I. cylindrica and Panicum reptans give

15.75% and 13.33% (Table 3.3). Therefore the highest dissimilarity contribution was made

by Panicum reptans as its abundance was 3.98% in the 1998 deposit and was absent in the

2006 deposit (Table 3.4). For the 2006 and 1997 pair of deposit sites, the dissimilarity is

93.13% (Table 3.3) characterized mainly by three species, namely A. javanica, Athyrium

macrocarpum and Polytrias amaura, with A. javanica providing the highest contribution

(16.35%) of the dissimilarities. A. javanica was absent in the 2006 and in the 1997 its

abundance was 4.19% (Table 3.4).

37

Five species were mainly responsible in explaining the difference between deposits

1998-1997 with C. flavidus having the highest contribution (19.51%) compared to the

other species (Table 3.3) as its abundance was 6.92% in the 1998 deposit and was absent in

the 1997 (Table 3.4). Differences between the 2006 and 1994 deposit were mainly due to

the contribution of three species, namely A. javanica, Athyrium macrocarpum and Polytrias

amaura with the highest contribution made by Polytrias amaura (16.04% Table 3.3) where

it was absent in the 2006 deposit. Its abundance in the 1994 was 3.78% (Table 4).

Deposits 1998 and 1994 were characterized by four grass species, namely C.

flavidus, I. cylindrica, Panicum reptans and Polytrias amaura, with the highest

contribution made by C. flavidus (19.31%). C. flavidus was abundant in 1998 (6.92%) and

less abundant in the 1994 deposit (0.20%). Lastly, in the 1997 and 1994 deposits the

highest contribution was shown by A. javanica (13.05%) where it was more abundant

(4.19%) in 1997 compared to the 1994 deposit (2.06%).

38

Table 3.3. SIMPER result 1. Percentage contribution of species to average Bray–Curtis dissimilarities in all pairs of sites. Only those species with a contribution to average dissimilarity of >5% are included. The average dissimilarity value (%) is also shown for each pair of the sites. Asterisks indicates exotic species.

Species Families

2006-

1998

2006-

1997

1998-

1997

2006-

1994

1998-

1994

1997-

1994

Anaphalis javanica Asteraceae - 16.35 9.99 8.08 5.11 13.05

Athyrium macrocarpum Polypodiaceae 10.27 9.86 - 12.25 - -

Calliandra callothyrsus Fabaceae - - - - - 5.00

Cyperus flavidus* Poaceae 23.89 - 19.51 - 19.31 -

Eupatorium riparium* Asteraceae 6.85 7.07 5.35 - 5.31 7.55

Imperata cylindrica Poaceae 15.75 8.21 13.06 6.72 13.07 8.16

Panicum reptans Poaceae 13.33 - 10.99 - 11.21 -

Paspalum conjugatum* Poaceae - - - 6.48 - -

Pinus merkusii Pinaceae - 7.11 - 7.83 - -

Polygala paniculata* Polygalaceae - - - 5.87 - 5.80

Polyosma ilicifolia Poaceae - 7.75 5.30 - - 7.26

Polytrias amaura Poaceae - 13.98 10.41 16.04 11.12 7.66

Average dissimilarity (%) 98.00 93.13 82.82 94.48 85.74 68.17

39

Table 3.4. SIMPER result 2. Average abundance percentage in each site of the selected species from table 2 (SIMPER 1). Asterisks indicate exotic species.

Species 2006 deposit 1998 deposit 1997 deposit 1994 deposit

Anaphalis javanica - 1.32 4.19 2.06

Athyrium macrocarpum 2.69 0.24 1.10 0.38

Calliandra callothyrsus 0.14 0.24 1.23 0.24

Cyperus flavidus* - 6.92 - 0.20

Eupatorium riparium* 0.10 2.18 2.77 1.28

Imperata cylindrica - 5.31 2.04 1.80

Panicum reptans - 3.98 - -

Paspalum conjugatum* 2.75 - - 0.56

Pinus merkusii 1.67 - - -

Polygala paniculata* 0.65 - - 1.70

Polyosma ilicifolia - 0.32 1.89 0.24

Polytrias amaura - - 3.52 3.78

Different species peak in abundance at different stages of early succession (Table

3.4). E. riparium, Athyrium macrocarpum and Calliandra callothyrsus were present at all

four sites. E. riparium abundance was lowest in the 2006 deposit, increased in abundance in

the 1998 and 1997 deposits, and then decreased in the 1994 deposit. E. riparium peaked at

the intermediate stage (1997 deposit). Athyrium macrocarpum abundance was highest in

the 2006 deposit whereas Calliandra callothyrsus abundance peaked in the 1997 deposit.

A. javanica, I. cylindrica and Polyosma ilicifolia were present in all but the

youngest (2006) deposit. A. javanica and Polyosma ilicifolia reached their peak of

abundance in the 1997 deposit whereas I. cylindrica reached its peak in the more or less

middle stages of the successional time range (1998 deposit). Paspalum conjugatum and

Polygala paniculata were only present in the youngest (2006) and the oldest (1994)

deposits. Paspalum conjugatum abundance was greatest in the 2006 deposit whereas

Polygala paniculata abundance was greatest in the 1994 deposit.

40

3.3.3. Interspecific association (co-occurrence)

The number of positive associations was greater in all sites than the number of

negative associations (Figure 3.3). Positive associations were highest in 1994, lowest in

1998 and at an intermediate value in the 2006 and 1997 deposits. Negative associations

were highest in the 1997 deposit, lowest in 1998 and similar in the 2006 and the 1994

deposits.

16

5

15

31

74

118

0

5

10

15

20

25

30

35

2006 1998 1997 1994

Deposits

Num

ber o

f ass

ocia

tion

Positive Negative

Figure 3.3. Total number of interspecific association of species in each deposit of primary succession sites of Mt. Merapi

A. javanica had the largest number of negative interspecific associations compared

with other species pairs, and in fact all of its associations were negative (Table 3.5). In

contrast, Calliandra callothyrsus had the highest number of positive interspecific

associations. I. cylindrica did not exhibit any association with other species at any of the

sites. Athyrium macrocarpum co-occurred with Polygala paniculata as shown by their

strong positive association (Table 3.5). E. riparium always co-occurred with Melastoma

affine whereas Calliandra callothyrsus was present together with Cyperus rotundus

41

Table 3.5. Results of the association tests using the chi-squared test statistic (χ2) between the differentiating species and their co-occurring species (as derived from SIMPER analysis; table 3). Association is significant at P values 0.05. Values of the Ochiai Index (strength of association) are equal to 0 at ‘no association’ and to 1 at ‘complete/maximum association’. Asterixis indicates the species pair that is also included in the most differentiating species. NA stands for no associations.

Species Paired species Type of association

Ochiai Index

Debregeasia longifolia - 0.40 Humata repens - 0.40 Anaphalis javanica Rubus flaxinifolius - 0.40

Polygala paniculata* + 1.00 Athyrium macrocarpum Polygonum chinense - 0.40

Crassocephalum crepidioides + 0.84 Cyperus rotundus + 1.00 Polygala paniculata* + 0.84 Panicum reptans* - 0.40 Eleusine indica + 0.77 Polytoca bracteata + 0.70

Calliandra callothyrsus

Polytrias amaura* - 0.40

Gnaphalium japonicum + 0.84 Stachytarpheta jamaicensis - 0.40 Eupatorium riparium Melastoma affine + 1.00

Imperata cylindrica NA NA NA

Polygala paniculata* - 0.28 Pinus merkusii Shuteria vestita - 0.40

42

3.4. DISCUSSION

In the first decade of primary succession, plant re-colonization on the nuées

ardentes deposits at Mt Merapi was rapid, with 52 species belonging to 23 families

recorded. Species abundance and composition were significantly different between the

younger and the older sites. Younger deposits were dominated by species such as Athyrium

macrocarpum, Polygonum chinense, Paspalum conjugatum and Cyperus flavidus. These

species were either present at very low abundance or were not present in the older deposits.

The older deposits were dominated by species such as Anaphalis javanica, Imperata

cylindrica, Polytrias amaura and Eupatorium riparium. These species were either present

at very low abundance or were not present in the younger deposits, which indicated that the

dominance of these species is related to the development of the habitat over a period of

several years at the sites.

E. riparium, Athyrium macrocarpum, Calliandra callothyrsus, I. cylindrica,

Polyosma ilicifolia and A. javanica were the most common species, although their

abundance varied throughout the succession. Native invasive species such as A. javanica

and I. cylindrica had lower abundance in the oldest deposit. A comparable phenomenon

was also found at Mt. St Helens, where the cover of the early pioneer pearly everlasting

(Anaphalis margaritaceae) increased after the eruption but then declined in the older phase

(Dale et al. 2005c). The decrease in I. cylindrica abundance may be due to the absence of

subsequent disturbance and fire that inhibited the stimulation of flowering (Collins and Jose

2009) and perhaps competition with other dominant species such as Polyosma ilicifolia and

E. riparium in the older sites. Invasive alien species like E. riparium were recorded in all

deposits, with relatively high abundance in the older deposits. According to Heyne (1987) ,

E. riparium is a fast growing species, usually found in steep slope in a wide range of soil

43

conditions. Thus, E. riparium may have indirectly facilitated the co-occurring species such

as Gnaphalium japonicum and Melastoma affine, perhaps by assisting in stabilizing and

preventing erosion on the deposit site. However, over-domination by this invasive species

could be a problem itself. The Mistflower or Eupatorium is native to South America, and

this unpalatable and highly competitive species has become a problem elsewhere such as in

Nepal (Kunwar 2003).

Stability in species richness and diversity after a decade of succession was

observed in Mt. Usu Japan (Tsuyuzaki 1991), while in Mt. St. Helens, the stabilization rate

was different in different locations or type of volcanic disturbances (del Moral 2007). At

the Mt. Merapi sites, species richness and diversity was still increasing after a decade of

succession. Species diversity was significantly different between the younger and the older

deposits, with the oldest deposit having the highest species diversity. The differences in the

rate of change in species diversity between the Mt Merapi sites and those of Mt Usu and Mt

St Helens was perhaps due to faster plant regeneration in the warm wet tropics than in the

cool or cold high mountain chains (A. Hamblin, personal communication, 2009).

Establishment of plant from seeds on volcanic deposits depends on a range of

factors. Plant establishment on Mt. St. Helens was influenced by factors such as distance

from seed sources, species-specific dispersal capabilities, germination and growth

characteristics of colonizing species and the substrate condition (Dale et al. 2005a; Dale et

al. 2005c). Distance from the safe sites that act as a seed source also plays an important role

in influencing the rate of plant establishment in primary succession (del Moral and Wood

1993; McClanahan 1986; Titus and del Moral 1998; Walker and del Moral 2003). From the

study of primary succession in a post-mining area, McClanahan (1986) found that distance

to a seed source was best in predicting the establishment of the subsequent species. Early

primary succession on Mt. St. Helens showed that although sites located near the

44

undisturbed vegetation still exhibited low plant cover, they retained higher species richness

comparable to the species richness in the undisturbed area (del Moral and Wood 1993). The

youngest (2006) deposit on Mt. Merapi provides a good example of the initial plant

establishment conditions. This site is located at 1,220 metres above sea level on a moderate

slope, 4 kilometres from the volcano’s crater. Seed source areas are located at a close

distance and so seeds are likely to disperse easily into the site - potentially by winds or

animals (mostly bird). The two major seed source areas for the 2006 deposit were a recently

harvested area of pine forest that was burnt in 2006 by fire associated with the nuées

ardentes, and an intact forest near the deposit. Rapid re-establishment in the 2006 deposit

was perhaps due to high abundance of perennial herbs in the understorey of the pine forest

which rapidly colonized the site after the forest was burnt. These perennial herbs species

were then perhaps dispersed by winds into the deposit site. The second seed source for the

2006 site was an intact forest on the west and east side of the valley that was minimally

affected by the eruption. Hence it was most likely that some tree seedlings found in the

2006 deposit site such as Pinus merkusii (Figure 3.4) and Calliandra callothyrsus may have

come from seed dispersal from the burnt and intact forests. Safe sites that act as seed

sources for the recovery of the vegetation in the affected areas were also observed in Mt.

Guntur volcano in West Java (van der Pijl 1939; van Steenis 1972).

The early establishment patterns recorded in this study also showed that there was a

convergence of floristic composition in the older deposits. A convergence can be identified

when two succession sites become similar in their floristic composition (Baer et al. 2005;

del Moral 2007). Chances of convergence occurring also increases where there are strong

biotic interactions and also similarities in climatic and edaphic conditions (del Moral 2007;

Walker and del Moral 2003). Pairwise comparison tests between deposits generally showed

an R value of 0.9 to 1.0, indicating large differences between the communities. Yet the

45

comparison between the 1997 deposit and the 1994 deposit was R = 0.42, indicating a

moderate degree of similarity between the sites. This result indicates that there was a

similar floristic composition in these two deposits. However, it is interesting to note that the

deposits that are only one year different in age - the 1997 and 1998 deposits - were not

more similar to each other. This finding indicates that factors other than age, such as sites

characteristics, are at play in influencing the species composition. The 1998 deposit was

located at the highest elevation (approximately 1,500 metres) compared to the other deposit

sites and hence the vegetation distribution and abundance in this area would also likely be

influenced by site characteristics such as elevation and the substrate’s nutrient status (van

Steenis 1972; Whitten et al. 1996).

Figure 3.4. Seedling of Pinus merkusii on the 2006 deposit. This is the type of situation at Mt Merapi in which recruitment from seed was taking place. Kaliadem, May 2008.

46

Plant establishment in primary succession is largely influenced by the development

of the site’s physical environment (Walker and del Moral 2003). Generally, nitrogen and

phosphorus are the most limiting essential macro nutrients in new volcanic soil materials

(Lambers et al. 2007). By means of physical weathering of new substrates over time,

phosphorus will become available for plants (Walker and Syers 1976). Nitrogen is

introduced to the ecosystem by N2 fixing organisms such as the Leguminoceae. Similar to

the primary succession on the debris-avalanche deposit at Mt. St. Helens (Dale et al.

2005c), there was also an increase in the abundance of nitrogen fixing species on the nuées

ardentes deposits at Mt Merapi. While Lupinus lepidus was the most abundant nitrogen-

fixing pioneer in early volcanic substrates in Mt. St. Helens (del Moral 2007), Mt. Merapi

had Calliandra callothyrsus as its nitrogen-fixing pioneer species during the early primary

succession. Calliandra callothyrsus is able to grow on a wide range of soil types, including

the moderately acidic volcanic origin soils that are a common feature in Southeast Asia

(Palmer et al. 1994). At Mt. Merapi primary succession, Calliandra callothyrsus was

present in all deposits of different ages. Its abundance was lowest in the youngest (2006)

deposit and was highest in the 1997 deposit but not in the oldest deposit (1994). The

decrease in Calliandra callothyrsus abundance was perhaps due to the dominance of grass

species Polytrias amaura in the oldest (1994) deposit. Given that nitrogen is generally one

of the most limiting essential macro nutrients in new soil materials (Lambers et al. 2007),

N2-fixing species are often regarded to have a facilitative function in succession

(Bellingham et al. 2001; Walker et al. 2003) because they can directly influence in assisting

establishment of subsequent species in the succession by providing nitrogen (Walker et al.

2003). Furthermore, the addition of organic matter by other pioneer species aids the

retention of water and nutrients to support the growth of co-occurring species (Hodkinson

et al. 2002).

47

Changes in species composition are not only driven by the changes in the physical

environment, but can also be driven by the effects of species on each other. This makes

species interaction an important indicator factor in succession and ecosystem development

(Muller 2005; Walker and del Moral 2003). Species co-occurrence observations may be

seen as the first attempt to detect species interactions, such as facilitation and inhibition,

and niche processes that structure the community (Walker and del Moral 2003;

Widyatmoko and Burgman 2006). At the Mt. Merapi sites, there was an increase in the

number of positive species associations over time. This observation might support the idea

that in the severely disturbed habitat in which primary succession occurs, the role of

facilitation will have more value for species change than competition (Callaway and

Walker 1997; Connell and Slatyer 1977; Walker and del Moral 2003).

At Mt. Merapi primary succession, Anaphalis javanica had the largest number of

negative interspecific associations with a low association index (0.4). This result may

suggest that Anaphalis javanica and Debregeasia longifolia, Humata repens and Rubus

flaxinifolius favour different environmental conditions to establish at a particular site. In

contrast, the nitrogen fixing legume, Calliandra callothyrsus, had the highest number of

positive associations (association index ranging from 0.7 to 1) with other species, such as

Cyperus rotundus and Eleusine indica grasses. This result may indicate that not only that C.

callothyrsus and its co-occurring species have similar requirements in terms of

environmental conditions, but it may also indicate the facilitative function of the legume

species (Walker et al. 2003). It is also interesting to note the absence of interspecific

association with Imperata cylindrica, which might indicate the aggressiveness of this

species in utilizing the resources, and possibly that, I. cylindrica, with its allelophatic

capability, may have inhibited other species from co-occurring together with it at a

particular site (MacDonald 2009).

48

Plant interspecific associations have also been found in other volcanic sites across

the globe. The association of Honkenya peploides, a low-growing, sand-binding pioneer,

with lyme grass, Elymus arenarius, and the lungwort, Mertensia maritima, has contributed

to the development of a relatively unstable ecosystem on Surtsey, a volcanic island in

Iceland (Fridriksson and Magnusson 1992; Thornton 2007). On the volcanic island of

Krakatau in Indonesia, the beach-creepers Ipomoea pes-caprae and Canavalia rosea, and

the grasses I. cylindrica (alang alang) or Saccharum spontaneum (glagah), have been

found to form associations in slow-growing sand dune communities on the island

(Thornton 2007). Furthermore, on a volcanic desert of Mt. Fuji, Japan, a dwarf pioneer

shrub, Salix reinii, was clumped together and positively associated with successional tree

seedling Larix kaempferi, demonstrating its role as nurse-plant in the primary succession

(Endo et al. 2008).

As a first attempt to detect species interaction (i.e. facilitation and inhibition) the

study of interspecific association is insufficient to determine the processes and mechanisms

of species interaction during succession. However, studies of the role of species interaction

during the succession can be done more effectively by targeting those species that are

strongly associated (Myster and Pickett 1992b). If the process and mechanism of recovery

and establishment in primary succession is to be investigated, a study of long-term

vegetation dynamics with manipulative controlled research is needed.

49

Chapter 4

Community Structure and Composition along a Chronosequence in

Forests Burnt by Nuées Ardentes-Induced Fire in Mt. Merapi

National Park

4.1. INTRODUCTION

In active volcanoes, volcanic activity remains the most significant threat to forest

vegetation (Lavigne and Gunnell 2006; Whitten et al. 1996). Fire is an integral part of

volcanic disturbance and has shaped community composition in montane forests of Java

(van Steenis 1972; Whitten et al. 1996). On Mt. Merapi, the intense heat (often more than

700° C) released from nuées ardentes ignites wildfires (Bardintzeff 1984).

Intense fire most likely kills or damages some plants, but others are more persistent

and even depend on fire for their regeneration (Bond and Wilgen 1996). Some species that

inhabit mountainous area of Java and Bali exhibit this phenomenon. In 1918, Mt. Semeru in

East Java erupted violently. Falls of hot ash stripped the branches of Casuarina

junghuhniana trees, but many trunks remained erect and re-sprouted. This species is a long-

lived pioneer, nevertheless, as with many other pioneers, it will be replaced by other

species when there are no subsequent fire disturbances (Whitten et al. 1996). Other

examples are Albizia lopantha and Pinus merkusii. Regeneration of these species is reliant

on fire, which breaks the hard seed case and allows seeds to germinate (Whitten et al.

1996).

The montane forests of Java and Bali are not resistant to fire (Marrinan et al. 2005).

The forests are easily ignited under conditions of prolonged drought, such as when

lightning strikes oil-rich species such as Vaccinium spp. On Mt Merapi, nuées ardentes are

the primary cause of forest fire (Simon 1998; Whitten et al. 1996). Recovery of the

50

montane forest following fire is usually slow (Horn et al. 2001). Fire destroys the

aboveground part of shrubs and some surviving species may be covered with ash, which

could slow the rate of the secondary succession (Antos and Zobel 2005; Whitten et al.

1996). Severely burned areas on mountains in Java and Bali are usually characterized by

the increase in abundance of invasive species, such as alang-alang grass (Imperata

cylindrica), and also white-leaved ‘edelweis’ (Anaphalis longifolia) and bracken fern

(Pteridium aquilinum) (Whitten et al. 1996). Homalanthus giganteus is also a common

pioneer tree species that occurs during secondary succession in these areas (van Steenis

1972).

Diversity measures are important indicators of successional processes in a plant

community (Hobbs and Huenneke 1992; Hobbs and Norton 1996; Magurran 1988; van der

Putten et al. 2000; Zhu et al. 2009). There are three contrasting hypotheses concerning

species diversity in a succession. In the first hypothesis, species diversity increases

continuously as succession progresses and the ecosystem becomes more complex (Odum

1969). The second hypothesis states that, due to plant density, species diversity is highest at

the beginning of the succession and decreases gradually as the succession proceeds

(Hubbell et al. 1999). The third hypothesis divides the succession into three stages: early,

mid and late. Species diversity, according to this hypothesis, will increase during the early

succession stage, reach a maximum in the mid succession stage and decrease in the late

succession stage (Aubert et al. 2003).

One of the main objectives of community ecology is to resolve the hypotheses

concerning the relationship between species assemblages and factors that may have

influenced the composition of the community (Pan et al. 1998). Temporal and spatial

heterogeneity of the abiotic environment is strongly correlated with the variation and

heterogeneity of the floristic assembly (Ruprecht et al. 2007; Walker et al. 2003). In the

51

tropical forests of Indonesia, the importance of edaphic factors that affect the floristic

composition of the forest ecosystem has been widely studied. However, these studies were

conducted on the lowland forests of Sumatra or Kalimantan Island (Brearley et al. 2004;

Herrera and Finegan 1997; Widyatmoko and Burgman 2006). Highland and/or mountain

rain forests, especially on Java Island, remain less studied (but see van Steenis 1972). A

significant proportion of the surface of the Indonesian archipelago is covered by highlands

or (volcanic) mountain regions. In Java, they are estimated to cover 21,950 square

kilometers, or 17% of the entire area of the island (Tan 2008). Soil changes are probably

the long-term impact of volcanic disturbances and are likely to have an effect on

vegetation.

Li et al. (1999) stated that many succession theories were based on intensive work

in temperate forests. Gomez-Pompa and Vasquez-Yanes (1981) and Chazdon et al. (2007)

studied secondary succession that occurs in the tropics, however their findings were based

on work on old fields or in lowland tropical forests. Other forest types such as volcanic

tropical montane forest have received little attention (Tsuyuzaki and Hase 2005; Whittaker

et al. 1999). Furthermore, we are now acknowledging that ‘one model fits all’ is not

appropriate for all communities and ecosystems due to the complexity of each system

(Hobbs et al. 2007).

52

The aim of this chapter was to describe plant species composition and diversity

along a chronosequence of sites that had been burnt by fires caused by nuées ardentes in

the tropical montane forest of the Merapi National Park. I then compared the successional

patterns to environmental variability. The research questions were:

1) Are there any differences in species diversity, turnover, and community structure and

composition across sites of different ages?

2) Which species contributed the most to the differences in composition between sites of

different ages?

3) Are there significant differences in major soil nutrients between sites of different ages?

4) Which subset of the soil nutrients best correlates with the species assemblages in the

course of secondary succession?

53

4.2. METHOD

4.2.1. Study sites

I chose four areas that were affected by nuées ardentes fire between 1994 and 2006

and one unaffected forest area in Kaliurang as a reference site. These study sites had acidic

soil and were located at an elevation from 1000-1500 metres above sea level. Note also that

although the sites had undergone nuées ardentes fire, their soils were of volcanic origin.

Detailed descriptions of the study sites are given in Chapter 2.

4.2.2. Vegetation sampling

In April 2008, vegetation was sampled in each of the four areas/sites burnt by fire

generated by nuées ardentes in 1994, 1997, 1998 and 2006. One area of unburnt forest (the

reference site) was also sampled. The position and altitude of each site were recorded using

a GPS, and slope was measured using a clinometer. At each site, an area of approximately

2.5 hectares was chosen and five circular plots (diameter range approximately 60 metres)

were randomly placed in the chosen area. In each of these larger plots, three sets of circular

plots of 10, 5 and 2 metres diameter were nested within each other to measure trees (10

metre plots), groundcover (5 metre plots) and seedlings (2 metre plots) (Isango 2007;

Supriyadi and Marsono 2001). The species name, height and diameter of trees (dbh ≥ 10

cm) and young trees (dbh 2-9.9 cm, height ≥ 1.3 m) were recorded. The number of

understorey plants and seedlings was counted (Kent and Coker 1992). All plants were

identified to species level when possible. Identification was conducted at the dendrology

laboratory, Faculty of Forestry, Gadjah Mada University Yogyakarta, Indonesia.

Identification was done using flora books such as “The Flora of Java” (Backer and van den

54

Brink 1963) and “Mountain Flora of Java” and the results were confirmed by a botanist in

the Faculty.

4.2.3. Soil sampling and analyses

To assess possible interactions between edaphic factors and species composition,

soils were sampled from each of the sites. Three to five soil samples were collected from

random points within the 60 metre plots. Soil samples were taken from 0-20 centimetre

depth using an auger (5 cm diameter), bulked and sealed in a plastic bag and transferred to

the laboratory. After removing stones, pebbles and large pieces of plant material, the

samples were sieved by 2 millimetre mesh screen and used for further physicochemical

analysis.

Soil organic matter was determined by the Walkey and Black (1934) method. Total

nitrogen and phosphorus were estimated following the Kjeldahl procedure (Bremmer and

Mulvaney 1982). The pH of the soil sample was measured in a soil-water suspension (1:2.5

w/v H2O) using a digital pH meter. Availability of exchangeable base cations (Ca++, Na+

Mg++ and K+) were extracted from the soil using a neutral (pH 7) salt extractant of 1 M

NH4-acetate in a mechanical vacuum extractor (Suarez 1996). The analyses were carried

out in the Soil Science Department, Faculty of Agriculture Gadjah Mada University

Yogyakarta, Indonesia.

4.2.4. Data analysis

Data were analysed using multivariate and univariate statistics. Multivariate

analyses such as Non Metric Multidimensional Scaling (NMDS), Analysis of Similarity

(ANOSIM), Similarity Percentage (SIMPER), and Biota-Environment matching (BEST)-

Linkage trees analysis (LINKTREE) were done using PRIMER V.6 (Clarke and Gorley

55

2005). Univariate analyses such as ANOVA and post hoc tests were conducted using SPSS

package V.11.5.

4.2.4.1. Species Diversity and Turnover

Species diversity at each site was calculated using the Shannon-Wiener diversity

index. Differences in diversity between sites were tested for significance using a one-way

ANOVA. To examine short term species turnover (beta diversity), a modified Sorensen’s

community correspondence index or CCI was used (Barbour et al. 1980; Cook et al. 2005)

with the formula as follows:

ba +=

2c CCI

Where a = the number of species present in the first community, b = the number of species

present in the second community, and c = the total number of species found in both

communities. I then calculated D, which is an index of how much a species list changes

across sites with the formula as follows (Cook et al. 2005):

CCI 1−=D

This index ranges from 0 to 1, and a low value indicates little change in the species

composition between sites whereas a high value indicates the opposite.

4.2.4.2. Forest Structure

In order to examine the vertical structure, forest vegetation was divided into five

strata (A, B, C, D and E), as recognized for humid tropical forests (Simon 1996). Stratum A

consisted of emergent trees more than 35 metres tall. Stratum B was the main canopy layer

with trees 18-35 metres in height. Stratum C consisted of young trees 8-18 metres tall.

Stratum D consisted of shrubs and sapling (of trees) with height ranges from 1.5-5 metres.

56

Stratum E was the groundcover layer, including grasses, herbs, tree seedlings and fern allies

(Simon 1996). The number of trees, young trees (poles), sapling and shrubs that have the

characteristics of stratum A, B, C and D were noted, while the number of groundcover

species was noted for the E stratum in each of the study sites.

Importance Value Index (IVI) (Curtis and McIntosh 1950; Kent and Coker 1992)

was used to describe the quantitative structure of the community. This statistic represents

the contribution that a species makes to the community in terms of the number of plants

within the quadrats (density), its contribution to the community through its distribution

(frequency), and its influence on the other species through its dominance. Importance Value

Index was calculated for each species of tree and groundcover in each of the study sites.

The formula for tree IVI is as follow:

IVI = RD + RF + RDom

Where RD = relative density of a species, RF = relative frequency of a species and RDom =

relative dominance of a species.

Relative Density of species A = 100% x species all of individualnumber Total

speciesA of individual ofNumber

Relative Frequency of species A = 100% x species all of valuefrequency Total

speciesA of valueFrequency

Relative Dominance of species A = 100% x species all of valuedominance Total

speciesA of valueDominance

57

Dominance values for a tree species were obtained by dividing the basal area of the tree

with the size of the plot (Simon 1996; Supriyadi and Marsono 2001). The IVI formula for

groundcover species (including seedlings) was similar to the tree layer but without the

calculation of relative dominance (Kusmana 1995), and so the formula is as follow:

IVI = RD + RF

Where RD = relative density of a species, and RF = is relative frequency of a species.

4.2.4.3. Variation in Community Composition

Species abundance data were square root transformed prior to all multivariate

analyses. A resemblance matrix based on a Bray-Curtis similarity index was generated as a

basis for the subsequent ordination and cluster analyses. Plant species composition and

abundance at each site were compared using non-metric multidimensional scaling

ordination (NMDS) (Clarke 1993). Statistically significant differences in species

composition and abundance between the sites were determined by analysis of similarity

(ANOSIM), which tests the null hypothesis that there is no difference in species

composition and abundance among groups (Clarke 1993). SIMPER, an analysis that

calculates the average Bray-Curtis dissimilarity between all samples, was used to identify

the species that differentiate sites (Clarke 1993).

4.2.4.4. Variation in Soil Nutrients and Relationship between Floristic Composition

and Soil Nutrients

Differences in soil nutrients between sites were tested for significance using one-

way ANOVA. To match species composition and abundance patterns in each site with the

soil nutrients, the BIOENV method in BEST routine was used (Clarke and Ainsworth

1993). BEST analysis looks at abiotic variables in combination, and tries to identify a

58

subset which is sufficient to ‘explain’ all the biotic structure capable of explanation.

Correlation values (ρ) in BEST describe how strongly these best soil nutrients affected

community composition. Statistical significance testing on BEST results was done by

generating global ρ value by 999 permutations. In addition, LINKTREE analysis (Clarke

and Gorley 2005; Mitchell et al. 2008) was conducted to demonstrate which variables take

high or low values for which samples. LINKTREE take the subset of abiotic variables

identified by BEST and use them to describe how best the assemblage samples are split into

groups, by successive binary division. Each division is characterized by a threshold on

environmental variables.

59

4.3. RESULTS

4.3.1. Species Diversity and Turnover

Sixty one species belonging to 29 families were recorded in the study sites. The

highest number of species belonged to the Poaceae (10), followed by Fabaceae (9) and then

Asteraceae (6). There were significant differences in species richness between sites

(ANOVA P = 0.05, table 4.1). Species richness was lowest at the 2006 site and highest at

the 1994 site. Species richness in the reference site (undisturbed site) was much lower when

compared to the 1994 site. Species richness in the reference site was significantly lower

than in all but the 2006 site. The changes in species diversity are not as distinct as the

changes in species richness over time (ANOVA P = 0.05, table 4.1). The reference site is

significantly different to the 1994 site, but not significantly different from 2006, 1998, and

1997. The 1998 site is not significantly different from 2006 and 1997 sites and also the

1997 site is not significantly different from the 1994 site.

Table 4.1. Differences in species diversity between the burnt sites and reference site in Mt Merapi National Park. Superscript letters (a-c) after mean values (±SD) indicates significant difference between sites as assessed with Tukey’s HSD test. Dates are those in which the site was burnt by fire generated by nuées ardentes. ANOVA group/years since fire

Species richness Species diversity

2006 site 9.20 (±1.48)a 1.91 (± 0.19)a

1998 site 14.0 (±3.39)b 2.13 (± 0.27)ab

1997 site 15.4 (±1.51)b 2.41 (± 0.19)bc

1994 site 19.4 (±2.96)c 2.7 (± 0.21)c

Reference site 10.6 (±1.67)a 2.21 (± 0.27)ab

60

Species turnover was highest (lowest species similarities) between the 2006 and the

1998 sites, and then between the 1997 and 1994 sites (Table 4.2). Species turnover between

the 1998-1997 sites was similar to the turnover between the 1994 site and the reference site.

Table 4.2. Species turnover rates (D) between pairs of sites in the chronosequence on Mt Merapi. 2006-1998 1998-1997 1997-1994 1994-Ref site

D 0.89 0.63 0.83 0.66

Sorenson Index 0.11 0.37 0.17 0.34

4.3.2. Forest Structure and Composition

In terms of vertical structure, the number of individuals found in each stratum

indicates the presence of the particular stratum in each site (Table 4.3). In the 2006 site,

stratums B, C, D, and E were recorded. The 1998, 1997 and 1994 sites also had four

stratums (B, C, D, and E) whereas in the reference site, all five stratums (A, B, C, D, and E)

were present.

Table 4.3. Number of individuals in each stratum for each site of secondary succession at Mt. Merapi. Stratum A refers to the number of trees that are more than 35 m in height. Stratum B is number of trees that are 18 to 35 m in height. Stratum C comprises of trees that are 8 to 18 m tall. Stratum D is the total number of saplings and Stratum E is the total number of groundcover species.

Stratums 2006 site 1998 site 1997 site 1994 site Ref site

A - - - - 41

B 3 5 16 25 28

C 5 45 18 17 1

D 4 10 15 6 11

E 12 20 23 25 16

In terms of quantitative structure, tree and groundcover species in the sites were

compared on the basis of the Importance Value Index, (IVI) (Table 4.4). In the 2006 site,

the tree layer was dominated by Pinus merkusii, whereas in the 1998 and 1997 site,

Homalanthus giganteus and Paraserianthes falcataria were the most important tree

61

species. In the oldest site (1994), Schima wallichii and P. merkusii were the most important

tree species whereas in the reference site, Altingia excelsa was the most important tree

species. In the groundcover layer, the 2006 site was dominated by Imperata cylindrica,

whereas Eupatorium riparium was the most important species in the 1998 and 1997 sites.

In the oldest site (1994) Brachiaria reptans was the most important species, while in the

reference site, Selaginella doederleinii was the most important species in the groundcover

layer. In the tree seedling layer, Acacia decurrens was the most dominant tree seedlings

species in the 2006 site, followed by P. merkusii. Albizia lopantha dominated the seedling

layer in the 1998 site, while in the 1997 and 1994 sites Calliandra callothyrsus was the

most important seedling. In the reference site, A. excelsa was the dominant seedling.

In addition to the IVI, Table 4.4 also shows the presence and absence of the most

important (dominant) species in each layer throughout the succession. In the tree layer, A.

excelsa and P. merkusii were present at the youngest site (2006) and then absent in the next

two older sites (1998 and 1997), and then reappeared in the oldest (1994) and the reference

site. Erythrina sp, H. giganteus, Albizia lopantha and Macaranga javanica were only

present at the 1998 site. In the groundcover layer, I. cylindrica was recorded in all four of

the burnt sites, but was more dominant in 2006, 1998 and 1997 sites than in the 1994 site,

and was absent in the reference site. In contrast, Brachiaria reptans was absent in the 2006

site and then present throughout the rest of the chronosequence and was at a very low

abundance in the reference site. Selaginella doederleinii started to appear in the 1997 and,

1994 sites and became dominant in the reference site.

62

Table 4.4. Importance Value Index (IVI), and shade tolerance for the most important species in each stratum at each of the study sites. Asterisks indicate exotic species.

Species Shading tolerance 2006 site 1998 site 1997 site 1994

site Reference

site Trees

Acacia decurrens* Intolerant 17.27 6.17 - 58.88 -

Albizia lopantha Intolerant - 33.68 - - -

Altingia excelsa Tolerant 60.63 - - 8.44 217.51

Erythrina sp Intolerant - 12.51 - - -

Homalanthus giganteus Intolerant - 148.65 - - -

Macaranga javanica Intolerant - 31.55 - - -

Paraserianthes falcataria Intermediate - - 116.20 46.83 -

Parkia sp Intolerant - 6.53 29.32 - -

Pinus merkusii Intermediate 222.09 - - 87.85 54.05

Schima wallichii Intermediate - - 104.92 89.0 21.80

Groundcover

Brachiaria reptans* Intermediate - 2.19 19.90 16.54 2.23

Eleusine indica Intermediate - - 8.82 14.49 3.11

Eupatorium riparium* Intermediate - 55.35 57.03 - 13.43

Eupatorium odoratum* Intermediate - 8.69 4.70 1.95 -

Imperata cylindrica Intolerant 73.77 14.18 23.51 9.55 -

Polygala paniculata* Intermediate 26.83 - 2.35 7.60 -

Selaginella doederleinii Tolerant - - 0.87 0.56 61.80

Seedling

Acacia decurrens* Intolerant 99.04 13.88 - 13.57 -

Albizia lopantha Intolerant - 77.77 - - -

Altingia excelsa Tolerant 9.52 - 11.05 7.46 33.35

Calliandra callothyrsus Intermediate - - 146.56 84.04 -

Pinus merkusii Intermediate 62.85 - - 41.31 -

Schima wallichii Intermediate - - 17.10 19.69 22.40

63

4.3.3. Variation in Community Composition

There was clear separation between the sites as shown by the NMDS ordination

result (Figure 4.1). Plots from the youngest site (2006) were separated from plots from the

older sites (1998, 1997 and 1994), and from the undisturbed site. An analysis of similarity

(ANOSIM) test confirmed that there were significant differences in Bray-Curtis species

similarities between sites (Global RANOSIM = 0.93, P < 0.001).

Figure 4.1. NMDS of sites based on vegetation composition and abundance: 2006 site (triangles), 1998 site (inverse triangles), 1997 site (squares), 1994 site (diamonds) and reference

site (circles).

Six pairwise comparison tests between sites (2006 and 1998, 2006 and 1997, 2006

and 1994, 2006 and reference site, 1998 and 1994, and 1994 and reference site) had an R

value of 1.0 (Table 4.5). The comparison between the 1997-1998 sites and 1997-1994 sites

had R-values of 0.72 and 0.86 and also, the comparison between 1997-undisturbed and

1998 undisturbed had R-value of 0.98 (Table 4.5).

64

Table 4.5. ANOSIM pairwise test of NMDS vegetation plots ordination. Sample statistic (Global R): 0.93, significance level of sample statistic P < 0.001, number of permutation: 999 Groups R Statistic

2006, 1998 1

2006, 1997 1

2006, 1994 1

2006, Reference site 1

1998, 1997 0.72

1998, 1994 1

1998, Reference site 0.98

1997, 1994 0.86

1997, Reference site 0.98

1994, Reference site 1

In Table 4.6, Eupatorium riparium contributed most to the dissimilarity between the

2006 and 1998 sites (21.27%), 2006 and 1997 sites (20.96%), 1998 and 1994 sites

(16.31%), and 1997 and 1994 sites (21.39%). Brachiaria reptans contributed most to the

dissimilarity between the 2006 and 1998 sites (9.96%) and 1998 and 1997 sites (9.89%).

Dichantium caricosum contributed most to the dissimilarity between the 2006 and 1994

sites (10.60%). Selaginella doederleinii was the most important species contributing to

dissimilarities between the reference site and the burnt sites. Imperata cylindrica was the

second most important species in the comparison between 2006 and 1998 sites and 2006

and the reference sites.

65

Table 4.6. SIMPER result. Percentage contribution of species to average Bray–Curtis dissimilarities in all pairs of sites. Only those species with a contribution to average dissimilarity of >5% are included. The average dissimilarity value (%) is also shown for each pair of the sites. Asterixis indicates exotic species.

Site comparison

Species

2006

and

1998

2006

and

1997

1998

and

1997

2006

and

1994

1998

and

1994

1997

and

1994

2006

and

Ref site

1998

and

Ref site

1997

and

Ref site

1994

and

Ref site

Brachiaria paspaloides* 6.21 - 6.51 - 5.07 - - 5.98 - -

Brachiaria reptans* - 9.96 9.89 8.01 5.42 8.45 - - 8.30 5.89

Calliandra callothyrsus - 8.01 8.68 - - 6.31 - - 6.80 -

Dichantium caricosum* - - - 10.60 7.81 7.22 - - - 8.08

Eleusine indica - - - 8.96 6.53 - - - - 5.97

Eupatorium odoratum* - - - 9.84 5.29 6.42 - - - 7.42

Eupatorium riparium* 21.27 20.96 5.76 - 16.31 21.39 6.15 13.94 13.60 -

Imperata cylindrica 9.89 - 7.41 9.02 - 5.63 15.14 5.83 9.89 5.13

Polygala paniculata* 6.82 - - - - - 6.81 - - -

Selaginella doederleinii - - - - - - 18.45 16.94 13.77 14.33

Average dissimilarity (%) 88.98 79.38 61.51 75.50 85.38 60.75 95.56 83.35 80.67 87.82

66

4.3.4. Variation in Soil Nutrients

Soil nutrients and characteristics were significantly different (ANOVA P = 0.05)

across the system (Table 4.8). Soil in all sites had pH values around 5. Soil organic matter

increased with time since the fire, and there was also a clear pattern of increases in N

content with the age of the site. The opposite was true with P; there was a clear and

statistically significant decrease in P concentration with the age of the site. The availability

of exchangeable potassium, magnesium and sodium content dropped slightly at first and

then increased throughout the succession. Calcium also increased significantly with the age

of the site.

4.3.5. Relationship between Floristic Composition and Soil Nutrients

BIOENV analysis in BEST showed a significant correlation between species

composition and soil nutrients (ρ= 0.6, P < 0.01) at the sites. BEST results showed that the

degree to which the chosen abiotic data ‘explains’ the biotic pattern (ρ) is optimized at

0.598 for the three soil nutrients P, Na and N, and slowly decreased beyond that, as more

variables were added or reduced (Table 4.7). In short, these three soil nutrients explain

about 60% of the variation in the species data.

Table 4.7. Global test BEST. Combinations of edaphic factors that best constrain the community composition within the Merapi National Park study sites.

∑ variables ρ Selections

3 0.598 P, Na and N

4 0.597 P, Na, K and N

5 0.579 P, Na, Mg, K, and N

3 0.578 P, Ca and N

2 0.571 P and N

67

Table 4.8. Summary of differences in concentrations of soil nutrients in secondary forest and the reference site on Mt. Merapi. Superscript letters after mean values (±SD) indicates significant different assessed with Tukey’s HSD test at p = 0.05 2006 site 1998 site 1997 site 1994 site Reference site

Soil organic matter (%) 0.75(±0.56)a 0.79(±0.09)ab 0.84(±0.22)ab 1.19(±0.19)b 1.05(±0.11)ab

Phosphorus (µg/g) 37.54(±5.17)d 23.72(±3.38)c 17.3(±2.25)bc 10.82(±3.37)b 1.05(±0.11)a

Nitrogen (%) 0.02(±0.01)a 0.05(±0.05)ab 0.11(±0.06)ab 0.15(±0.11)ab 0.21(±0.02)b

Magnesium (µg/g) 4.77(±0.53)a 2.67(±0.18)a 27.72(±1.03)b 21.29(±1.29)b 48.36(±12.56)c

Calcium (µg/g) 32.9(±2.43)a 57.78(±14.09)a 516.29(±45.49)b 567.77(±31.2)b 775.48(±77.71)c

Sodium (µg/g) 73.71(±1.92)b 28.22(±0.54)a 175.5(±11.14)c 221.12(±9.81)d 225.29(±22.7)e

Potassium (cmol(+)kg¹ֿ) 0.05(±0.02)a 0.02(±0.01)a 0.22(±0.12)a 0.14(±0.05)a 0.43(±0.2)b

pH 5.37 5.08 5.21 5.19 5.36

68

The LINKTREE diagram splits the study sites into four groups (nodes A, B, C, and

D) with abiotic inequalities. B% is an absolute measure of group differences and provides

the y-axis for the tree (Figure 4.2). LINKTREE analysis showed a large split for A

(B%=85%) that decline (split B at 69.6%) as groups got closer together. The first split (A)

in the species data was between the reference site (plots 13-15) and the burnt sites (plots 1-

12). This had an ANOSIM of R = 0.61. It was characterized by low or high P content (P<-

1.71 to the left, and >0.31 to the right). Alternatively, the same split of sample is obtained

by choosing high or low Mg or Na (left to right). The next split (B) divides the 2006 site

(plots 1-3) from the other remaining burnt sites (R = 0.65), based on Ca concentrations,

which were low at the 2006 site. The next split (C) divided the 1998 sites (plots 4-6) from

the 1997 sites (plots 7-9) and the 1994 sites (plots 10-12) (R = 0.94), with five possible

explanations: Na, Mg, P, K. Split D divided the 1997 and 1994 sites (R=0.93) on the basis

of Mg, Na and P.

69

Split B% R Soil factors characteristics A 85 0.61 P<-1.71(>-0.319) or Mg>0.853(<0.691) or Na>0.806(<0.794) B 69.6 0.65 Ca>-1.11(<-1.31) C 59.5 0.94 Na<-1.63(>0.41) or Mg<-1.29(>0.371) or P>0.864(<0.417) or K<-0.803(>-0.401) D 19 0.93 Mg>0.624(<0.478) or Na<0.56(>0.699) or P>0.182(<0.161)

Figure 4.2. LINKTREE diagram. 2006 site = plots 1-3; 1998 = plots 4-6 site; 1997 = plots 7-9;

1994 = plots 10-12; and the reference site = plots 13-15.

70

4.4. DISCUSSION

In the first decade after disturbance by fire there was a rapid recovery at the sites,

with 54 species belonging to 23 families recorded in the secondary forest at the study sites.

The highest number of species belonged to the Poaceae (10), followed by Fabaceae (9) and

then Asteraceae (6) (Appendix 2). Species richness and diversity increased with time since

the fire, however species richness and diversity in the reference site was not significantly

different from the youngest (2006) site. This pattern was similar to that reported in other

studies where species diversity reached its peak in older succession sites after most of the

climax species had entered the system, and then decreased with the loss of the species

present in early successional stages (Magurran 1988; Peet 1992; Zhu et al. 2009). The

results support the hypothesis of Aubert et al. (2003) that species diversity will increase

during the early succession stage, reach a maximum in the mid-succession stage and

decrease in the late succession stage. A decrease in the light availability at the forest floor

as the succession proceeds might be the cause of the decline of species diversity in the

reference site (Gomez-Pompa and Vazquez-Yanes 1981). Direct shading of overstorey

species inhibits the existence and regeneration or growth of less tolerant and intolerant

understorey species in the reference site (Lepš 1990).

There was progressive development of forest structure over time. Although all of

the burnt sites had four strata (B, C, D, E), the number of individuals in each stratum

differed. The number of individuals of stratum B (tall trees 18-35 m) was lowest in the

2006 site, greater in the older sites, but was the greatest in the reference site. The reference

site had five strata (A, B, C, D, and E) with the lowest number of individuals of stratum E

compared with the proportion of stratum E in the burnt sites. There were also differences in

the patterns of abundance of the most important species with different light requirement

71

characteristics (shade tolerant/intolerant) in the groundcover layer. The gradual decrease in

I. cylindrica (shade-intolerant species) abundance over time contrasted with the gradual

increase in the abundance of Selaginella doederleinii (shade-tolerant species), suggesting

that there was a decrease in the light availability at the forest floor as the canopy developed

and the succession proceeded.

Over the course of succession, the characteristics of species found at a site will

change (Wills 2002). In the Mt. Merapi sites, the younger sites were characterized by

shade-intolerant colonizer species with good dispersal capability. I. cylindrica is a widely

distributed invader grass that has a long record of colonising cleared lands in Indonesia

(Eussen and Soerjani 1975; Soerjani et al. 1983). I cylindrica has wide-spread rhizomes

and its seeds are wind-dispersed (Jonathan and Hariadi 1999), making it an effective

colonizer following fire (Murniati 2002). A. decurrens, however, is a nitrogen-fixing shrub

that is usually recruited after fire. At Mt. Merapi, it may have regenerated following the

nuées ardentes fire from a soil seed bank (Hardiwinoto et al. 1998; Spurr and Barnes

1980). I cylindrica and A. decurrens can also be found in other degraded areas on

volcanoes in Java, such as in Mt. Bromo-Tengger and Mt. Semeru (Anonym 2009; Whitten

et al. 1996). The species that occurred in the older sites and reference site on Mt. Merapi

were characterized as intermediate to shade-tolerant species with greater longevity. In the

older sites, A. decurrens was replaced by the leguminous tree, C. callothyrsus, which

occurred with other tree species such as Altingia excelsa. A. excelsa is a native emergent

tree species and its seedlings are shade tolerant. Older sites were also characterized by the

presence of the fern Selaginella doederleinii and the exotic invasive Eupatorium spp. in the

groundcover layer. Eupatorium spp. is a fast growing species, usually found on steep slopes

in a wide range of soil conditions and light availability (Heyne 1987).

72

Many studies have shown that generally species composition changes with time

after a fire (Clearly et al. 2006; Reilly et al. 2006; Ross et al. 2002; Spencer and Gregory

2006). The result of NMDS ordination was notable in that species composition differed

among each site, suggesting that the species composition changes with time after a fire. The

2006 and 1998 sites were different in terms of floristic composition and abundance with

highest species replacement rate when compared with the replacement rate in the other sites

(D = 0.89). Altingia excelsa, Pinus merkusii, and Polygala paniculata which were present

in the 2006 site, dropped out in the 1998 site whereas there was an increase in the number

of species from the Fabaceae family in the 1998 site with the colonization of Albizia

lopantha, Erythrina sp and Parkia sp. The increase in the number of N-fixing species from

the 2006 site to 1998 site seems to be followed by an increase in soil nitrogen. The

differences in species assemblages between the 1998 and 1997 sites (short interval) was the

lowest in all of the site pair-wise comparison, but they were still significantly different from

each other. Consistent with this, the species replacement rate was also lower (D = 0.63)

when compared with the replacement rate in the other site comparisons. Although

ANOSIM showed that the reference site and the 1994 site were significantly different, the

turnover rate between these sites was more or less the same as the rate in the 1998-1997

sites (D = 0.66). This result indicates that some of the species that characterized the

reference site, such as Altingia excelsa, Schima wallichii and Selaginella doederleinii, had

appeared earlier in the 1994 site and thus suggested convergence of floristic composition in

these sites.

In the Mt. Merapi succession, the changes in abundance of some invasive species

such as I. cylindrica, Brachiaria spp., and Eupatorium spp. are important to note. I.

cylindrica is an invasive native of south-east Asia. I. cylindrica dominated the early

succession sites, but was not recorded in older sites as it was most likely shaded out by

73

increasing canopy cover. In contrast, invasive exotic species Eupatorium spp. remained in

the system long after the fire had occurred and forest structures had developed. Eupatorium

is native to South America, and this noxious and highly competitive species has become a

problem elsewhere in Asia, such as in Nepal (Kunwar 2003). In the longer periods,

domination of invasive exotic species may limit the chance of recruitment of other native

species including seedlings of woody species, thereby reducing diversity and even changing

the successional trajectory and ecosystem functioning (Dale et al. 2005c; Hobbs and

Huenneke 1992; Raghubanshi and Tripathi 2009; Standish et al. 2009).

Changes in the physical environment, especially soil properties, can also be crucial

factors in driving the species diversity and ultimately the successional trajectory (Dzwonko

and Gawrofiski 1994; Velazquez and Gomez-Sal 2007; Whittaker 1960). Examination of

soil nutrients status generally showed improvement in soil condition as time progressed.

Nitrogen concentration was generally lower relative to P concentration in the early stage of

succession (Peet 1992; Walker and Syers 1976). At Mt. Merapi, total nitrogen content was

lowest in the 2006 site and gradually increased with time since fire. An increase in N may

also be the result of the interplay between biotic factors such as the occurrence of nitrogen-

fixing legumes, changes in other abiotic factors such as water content, soil pH, and soil

organic matter, or atmospheric deposition (Lambers et al. 2007; Le Brocque 1995a; Walker

and del Moral 2003). At the Mt Merapi sites, P concentration showed an opposite pattern to

N (Walker and Syers 1976), with highest concentrations in the earliest stage of succession

that declined over time. Kennards and Gholz (2001) reported that following intense fire and

extremely high temperatures (as is the case with nuées ardentes), there is a gradual increase

in extractable P. Soil organic matter showed a pattern of increase with time since fire. Fire

burns soil organic matter in the upper layer of the soil. Debano and Conrad (1978) found

significant losses in soil organic matter following intense fire. Exchangeable base cations

74

(Ca++, Na+ Mg++ and K+) generally increased with time. Capogna et al. (2009) also reported

an increase in Mg after fire in the Mediterranean basin and a gradual increase in the other

exchangeable base cations with time since fire has also been reported (Kennard and Gholz

2001). Fire burns the soil organic matter, and Mg that was previously complexed with the

organic matter is released into the soil through mineralization processes (Capogna et al.

2009).

BEST test result suggested that there was a significant correlation between species

composition and soil nutrients (ρ= 0.6, P < 0.01) with nitrogen, phosphorus and sodium

explained about 60% of the variation in the floristic composition. Furthermore, LINKTREE

analysis results indicate which soil variables take high or low values for which biotic

samples (Clarke and Gorley 2005; Mitchell et al. 2008). Phosphorus and exchangeable

cation concentrations differentiated the floristic composition in the reference site from the

burned sites. Phosphorus concentration was much lower in the reference site than in the

burned sites. The reverse was true for the exchangeable cations. It was also notable that the

increase in the number of N-fixing species such as Albizia lopantha, Erythrina sp and

Calliandra callothyrsus, Paraserianthes falcataria and Parkia sp. in the older sites appear

to be followed by an increase in soil nitrogen. Similarly, there were also increases in soil

organic matter and species richness as the burnt sites aged. In older sites, soil organic

matter is expected to be higher than in the younger sites since more accumulation or

decomposition of plant debris of the more vegetated old sites. These findings may support

the hypothesis that temporal and spatial heterogeneity of the abiotic environment is

correlated with the variation and heterogeneity of the floristic assembly or vice versa

(Ruprecht et al. 2007; Walker et al. 2003). However, it should also be noted that the

changes in community composition during succession is the result of complex abiotic and

biotic processes (Temperton et al. 2004). Hence, it would be rare for one environmental

75

variable to adequately explain floristic assemblages. Other biotic factors, such as

competition, are perhaps also at play (Le Brocque 1995b; Peet 1992; Temperton et al.

2004; Walker and del Moral 2003).

This study suggested that the Merapi forest exhibited a high resilience for site

recovery following nuées ardentes-induced wildfire with the rapid re-colonisation of

species and improvement in soil conditions in the burnt sites. However, unlike N, P

concentrations continued to decline as the sites aged. It is also important to consider the

potential problems of invasive species Eupatorium spp. as weeds, as these species remain

abundant even in the much more developed sites. These findings may have important

consequences for forest management as there is still much to learn about the capability of

alien invasive species to change soil chemical properties, which can be crucial factors in

driving the successional trajectory (Collins and Jose 2009; Dzwonko and Gawrofiski 1994;

Hughes and Denslow 2005; Velazquez and Gomez-Sal 2007).

76

Chapter 5

General Discussion

5.1. INTRODUCTION

The research presented in this study fills a gap in our understanding of an important,

though understudied, Indonesian tropical volcanic ecosystem. The aim of this study was to

describe patterns of plant community succession that occur in the tropical volcanic montane

ecosystem at Mt. Merapi, Java, Indonesia. The major findings of the research are identified

in the next sections of this chapter. The detailed results and discussions on each topic are

presented within the relevant chapters. Overall, results from this study have shown that the

succession progress is relatively rapid, as shown by the significant increase in species

richness and diversity, increase in positive species association and improvement in soil

nutrients over more than a decade of succession.

5.2. PLANT SUCCESSION ON MT. MERAPI

5.2.1. Species Re-colonization and Diversity

There was a close similarity in the number of species and families found in the

primary and secondary succession. Fifty-two species belonging to 23 families were

recorded in the four nuées ardentes deposits along the chronosequence, whereas 54 species

belonging to 23 families were recorded in the four secondary succession sites. Species

richness and diversity increase progressively from early to late successional stages in

tropical forests (Brearley et al. 2004; Pena 2003), though some studies have found that

diversity is often greatest in the mid-successional stage rather than in the late successional

stage (Aubert et al. 2003; Zhu et al. 2009). In this study, species richness and diversity

77

increased significantly with time since nuées ardentes in both primary and secondary

succession. However, species diversity decreased in the late successional stage in the

secondary succession sites, with species richness and diversity in the reference site not

significantly different from the youngest (2006) site. This result may be caused by the loss

of the species present in early successional stages due to decrease in the light availability at

the forest floor in the reference site (old secondary forest) (Pena 2003).

5.2.2. Species Composition and Structure

Shade-intolerant colonizer species with good dispersal capability have been

observed to dominate the early successional stage of succession, whereas intermediate to

shade-tolerant species with greater longevity characterize late successional stages

(Mahecha et al. 2009; Powers et al. 2009). In this study, species composition differed

significantly with site age in both the primary and secondary succession sites. In the

primary succession, younger deposits in Mt. Merapi were dominated by species such as

Athyrium macrocarpum, Polygonum chinense, Paspalum conjugatum, and Cyperus flavidus

whereas older deposits were dominated by species such as Anaphalis javanica, Imperata

cylindrica, Polytrias amaura and Eupatorium riparium. In the secondary succession sites,

I. cylindrica (alang-alang grass) dominated the sites soon after the fire and decreased in

abundance with site age. The lowest abundance of I. cylindrica was recorded 14 years after

the nuèe ardente fire. At the same time as I. cylindrica abundance declined, the invasive

shrub species E. odoratum began to dominate the sites and other tree species such as

Calliandra callothyrsus, Schima wallichii and Pinus merkusii were also recorded.

In the secondary succession, the oldest site (14 years after fire) had similar

structural characteristics (i.e. almost complete stratums and similar numbers of individuals

occupying the B stratum) as the reference site. Similarly, in 2 to 40 year-old

78

chronosequence sites in the Bolivian Amazon, stand structures also developed as the sites

aged (Pena 2003). In tropical dry forests in Costa Rica, recovery of the forest structure

comparable to those of mature forest required 4 to 5 decades of successional period

(Powers et al. 2009) therefore it is evident that the Merapi forests recover relatively rapidly.

Walker and del Moral (2009) proposed their view that stability and predictability in a

succession is maximum if there is a minimum or maximum frequent and severe

disturbance. Rapid recovery of plant communities found in this study may be due to the

ecosystems being adapted to severe and frequent volcanic disturbance.

Acacia decurrens, Altingia excelsa, Schima wallichii and Pinus merkusii were the

most important species in the tree layer. Brachiaria reptans, Eleusine indica, Eupatorium

riparium, Imperata cylindrica, Polygala paniculata and Selaginella doederleinii were the

dominant species present in the groundcover layer. The gradual decrease with time since

fire of shade-intolerant invasive species such as I. cylindrica and the opposite pattern

shown by a shade-tolerant fern species, Selaginella doederleinii, suggested that there was a

decrease in the light availability at the forest floor as the succession proceeded. This further

supports the view of a development in forest structure and canopy cover over time.

5.2.3. Exotic Invasive Species

Imperata cylindrica and Eupatorium spp. (E. odoratum and E. riparium, appendix 3

and 4) were the most important invasive species in this study. While the native invasive I.

cylindrica is not a problem at the moment because its abundance decreased as canopy cover

developed, the exotic invasive Eupatorium spp will likely continue to spread on Mt Merapi

because it tends to remain in the system even after a canopy had developed. The dominance

of Eupatorium riparium and Eupatorium odoratum in the groundcover layer of an old

secondary forest of Kaliurang (also termed as a reference site in this thesis) (Sutomo 2004),

79

the wide spread of Eupatorium riparium and Eupatorium sordidum in a closed forest of an

extinct volcanoes of Mt. Gede-Pangrango National Park in West Java (Wuragil 2009) and

the domination of Eupatorium inulifolium and Eupatorium adenophorum in Mt.

Papandayan Nature Reserve also in West Java (Setiawan 2008) support this view.

Eupatorium is native to South and Central America, and for decades, Eupatorium

spp. has gained attention as a noxious weed in Asia and Africa (Rogers and Hartemink

2000). In Indonesia, Eupatorium spp. were first introduced by the Dutch colonials in the

early 19th century and were used to reduce erosion in the large scale plantations of tea

(Camellia sinensis) and quinone (Cinchona spp.) in West Java (van Steenis 1972). Wind

dispersed seeds, an allelophatic strategy, and the capability to adapt to a wide range of soil

conditions has made this species a successful invasive weed (Grashoff and Beaman 1970;

Heyne 1987; Kunwar 2003). Eupatorium spp. also has the capability to benefit from

increased human disturbance (Kunwar 2003).

Fifty years following the eruption of the Paricutin Volcano in Mexico, Eupatorium

glabratum dominates the tephra deposits known as “arenales” and slowed the rate of the

succession (Lindig-Cisneros 2009; Lindig-Cisneros et al. 2006). Eupatorium spp. are not

exclusively a problem of high elevation only, but also in low elevation forest such as in

India and Indonesia (Swamy et al. 2000). In Indonesia, Eupatorium spp. were also found in

rubber plantations and savannas such as in Pangandaran Nature Reserve and in East Nusa

Tenggara (Prawiradiputra 2007). Invasion of Eupatorium spp. has also been a severe

problem elsewhere, such as in Nepal, where this species has been observed to have a

tendency to become over-dominant and thus decrease species diversity (Kunwar 2003;

Raghubanshi and Tripathi 2009). Hence, ecological intervention in the form of weed

management in Mt. Merapi succession should be considered.

80

5.2.4. Interspecific Association

Observation of interspecific associations in this study suggested that in a harsh

environment such as in early primary succession, plant facilitation is more prominent than

competition (Callaway and Walker 1997; Walker et al. 2003; Walker and del Moral 2003).

In the Mt. Merapi primary succession, the number of significant associations fluctuated as

succession proceeded, with positive associations being more apparent than negative

associations. Similarly, observation of interspecific association in an old field succession in

Romania resulted in no apparent trend, but instead species associations fluctuated as time

progressed (Ruprecht et al. 2007). In Mt. Merapi primary succession, the nitrogen-fixing

species Calliandra callothyrsus had the highest number of positive interspecific

associations compared with the other species, which may suggest that this N2-fixing species

has a prominent role in facilitation. In other volcanic ecosystem such as in Mt. St. Helens,

Lupinus lepidus was the most abundant nitrogen-fixing pioneer in early volcanic substrates

and was reported to have facilitated subsequent colonizers (del Moral 2007; del Moral and

Wood 1993). As a consequence of the paucity in nitrogen, leguminous species were

prominent in the succession on Mt. St. Helens.

5.2.5. Soil Nutrients

Succession involves changes in both the composition and structure of vegetation,

and changes in soil properties (Peet 1992). In a secondary succession of a tropical montane

cloud forest area in Mexico, there were significant changes with time since abandonment of

the majority of measured soil nutrients (Bautista-Cruz and del Castillo 2005). Similarly in

this study, most soil nutrient concentrations generally increased over time in the secondary

succession sites. Soil N and P showed contrasting patterns in concentration over time. In

just over a decade since fire, total N concentration increased significantly whereas the

81

reverse was true for total P concentration, which was greatest in the earliest stage and then

declined over time. The pattern of N:P found in this study was similar to what is generally

recorded during succession and therefore supports the generalization that in early years of

succession N is the limiting major soil nutrient, whereas P is limiting at the late stages of

succession (Walker and Syers 1976). Similarly, in a primary succession in Hawaii

Volcanoes National Park, N was reported to be the most important factor affecting plant

growth in the younger sites (Vitousek et al. 1993). The N concentration is expected to

increase with the invasion of nitrogen-fixing species (Walker and del Moral 2003). In this

study, invasion by nitrogen-fixing species does occur and is presumably associated with the

increase of N in the system. In contrast, in a secondary succession in a tropical montane

cloud forest in Mexico, N concentration remained low due to the absence of nitrogen-fixers

(Bautista-Cruz and del Castillo 2005). P becomes limited as the succession proceeds due to

the demanding usage by growing vegetation in each progressive stage of the succession,

and also the remaining P is also disappearing whether by erosion or becoming complexed

with other minerals (Bautista-Cruz and del Castillo 2005; Lambers et al. 2007; Walker and

Syers 1976). Lambers et al. (2007) discuss the changes in total soil P and N in a geological

time scale chronosequence whereas, in this study, the pattern of changes in total soil P and

N were observed over a very short time period.

Examination of the relationship between vegetation composition and environment,

especially soils, has been an important aspect in ecology (Pan et al. 1998; Zuo et al. 2009).

BIOENV analysis in this study showed a moderate but significant correlation between

species composition and soil nutrients, suggesting that soil organic matter, P, N, Ca++, Na+

Mg++ and K+ may be important abiotic factors in the species composition, and accounted

for approximately 60% of the species-environment relationships among the sites. In a

tropical forest in Sumatra, Indonesia, soil organic C, total N, Ca++, Mg++ and K+ were

82

reported to influence the distribution and abundance of a rare palm, Cyrtostachys renda. In

a dune succession in China, 68.1% of the species-environment relationship was correlated

with soil organic C, total N, electrical conductivity, pH, slope, very fine sand content and

soil water content (Zuo et al. 2009). However it should also be noted that the changes in

community composition during succession is the result of complex abiotic and biotic

processes and that in addition to the edaphic factors, biotic factors, such as facilitation and

competition, are also at play in shaping the community composition (Pan et al. 1998;

Temperton et al. 2004).

5.3. SIMILARITIES TO OTHER VOLCANOES IN ASIA

Knowledge gained from this study may also be relevant for other volcanoes in Asia,

as there may be similarities between the conditions on Mt. Merapi with other Asian

volcanoes in terms of the characteristics of volcanic activity, floristic composition,

environmental conditions and interaction between the biotic and abiotic factors in the

volcanic ecosystem.

Papua New Guinea has three active volcanoes, Mt. Lamington, Mt. Waiowa, and

Mt. Victory. Although nuées ardentes are not characteristic of these volcanoes, they have

an abundance of I. cylindrica and also Dysoxylum spp. in their younger deposits and

secondary succession sites (Taylor 1957). As is the case at Mt. Merapi, at Mt. Waiowa and

Mt. Lamington, there was a rapid recovery of plant communities and the communities

showed significant response to edaphic factors (Taylor 1957). In contrast, at Mt. Victory,

climate was the prominent driving factor of the succession.

83

Mt. Pinatubo in the Philippines is one of the few other examples of Asian volcanoes

which exhibit the phenomenon of nuées ardentes (Scott et al. 1996). However, information

regarding the development of its vegetation following its 1991 eruption is scarce. The

differences between Merapi and Pinatubo are that, unlike Pinatubo, Merapi is a very active

volcano with small and frequent nuées ardentes that occur between large infrequent

eruptions (Bardintzeff 1984). Once thought to be extinct, Mt. Pinatubo erupted violently in

1991 and created a local increase in carbon dioxide that rapidly boosted the regeneration of

its surviving and colonizing plants (Dale et al. 2005b).

Rapid recovery is also apparent in the Mt. Merapi ecosystem (Figure 5.1). Results

from this study have shown that the ecosystems on Mt Merapi are largely resilient to this

type of volcanic disturbance. However, as the threat of habitat and ecosystem destruction

due to the consequences of climate change and anthropogenic disturbance increases, these

volcanic-forests highlight the continuing need and importance of research on plant

community succession and restoration on a volcanic terrain in Indonesia and Asia in

general.

84

(A)

(B)

Figure 5.1. Resilience of Merapi volcanic ecosystem. (A) Kaliadem nuées ardentes deposit areas three months after May 2006 eruption (retrieved from the internet http://omhanif.multiply.com)

(B) Kaliadem nuées ardentes deposit areas, two years since eruption (photo by author).

85

5.4. MANAGEMENT IMPLICATIONS

In line with Lavigne and Gunnel’s (2006) view that vegetation in Java is resilient to

repeated volcanic activity, the results from this study show that the Merapi ecosystem is

resilient to nuées ardentes disturbance. However, concerns regarding problematic invasive

alien species such as Eupatorium odoratum and Eupatorium riparium mean that weed

management is an important management activity in the national park. Still, there is limited

information on the control of E. odoratum and E. riparium (Norgrove et al. 2000; Yadav

and Tripathi 1985; Yadav and Tripathi 1981).

Weed management to control species such as Eupatorium spp. can be done by plant

or plant part removal, or by changing resource availability (i.e. nutrient status, water and

light) (Luken 1990). Mechanical control, primarily by mowing has been used traditionally

to suppress the spread of Eupatorium capillifolium in pasture areas. This way, the regrowth

of this species is hindered and the seed production is reduced (MacDonald et al. 1994).

Mowing and/or slashing are physically demanding and impractical on Mt Merapi so other

options are preferred, such as chemical control using herbicides. Nevertheless, cover of

Eupatorium compositifolium was reduced by only 5% one year following the application of

Dicamba + 2, 4-D (Meyer and Bovey 1991). Thus, according to MacDonald (1994) a better

result is obtained when mechanical and chemical methods (mowing and herbicide usage)

are combined, as mowing and the use of Dicamba + 2,4-D or Triclopyr + 2,4-D herbicide

successfully reduced over 94% of the regrowth of Eupatorium capillifolium.

Changing resource availability can also be used in weed management (Luken 1990)

and to be able to perform this, first we have to understand what factors limit their growth

and development. Soil moisture stress was proposed by Yadav and Tripathi (1985; 1981)

as one of the possible factors regulating the population of E. riparium. This method will not

86

be easy to implement on Mt. Merapi as the wet mountain climate retains high moisture (van

Steenis 1972). Light and competition with adult plants were considered to be the main

factors that regulate the population of E. odoratum throughout secondary succession

(Kushwaha et al. 1981; Norgrove et al. 2000; Yadav and Tripathi 1981). Hence the use of

shade, for example by planting native, fast-growing legume tree species such as Calliandra

callothyrsus or Paraserianthes falcataria, then adjusting their density once E. odoratum

has been removed, could be an important action to suppress this species and kick-start the

vegetation succession sequence (Norgrove et al. 2000). So far it has not been possible to

successfully control Eupatorium using a single method or over a short time frame – instead,

a detailed monitoring and research program to find the most suitable long-term result

management options is required (Kunwar 2003).

Anthropogenic disturbance is likely to be one of the factors that may alter a

successional trajectory (Baeza et al. 2007; Ramaharitra 2006). In other areas of Indonesia,

conversion of forest into agricultural land has increased the occurrence of I. cylindrica

(alang-alang grass) as a weed (Murniati 2002; Soerjani et al. 1983). While the results from

this study showed that I. cylindrica does not appear to be a problem in the areas affected by

the nuées ardentes, expanding highland farming activity would likely cause I. cylindrica to

become a potential problem on Mt. Merapi. Imperata can become a fire hazard as it can

easily ignite, either from extreme temperature, lightning or by fire sparked by the falling of

rocks from the upper slope of the volcano (Andrews 1983). It is also adapted to fire as the

fire stimulates flowering (MacDonald 2009). Thus I. cylindrica dominance could

potentially increase fire frequency and reduce biodiversity (Murniati 2002; Soerjani et al.

1983).

87

Therefore, the park authority should also consider these possibilities when thinking

about the future management options. Mt. Merapi may benefit from wildfire-control

management in its strategic planning, as wildfire not caused by nuées ardentes also occurs;

for example, such wildfires burnt 300 hectares of forest in 1992 and 2002 (Heru 2002).

Fuel reduction can be done by removing I. cylindrica from the infested site to reduce the

risk of fire and the construction of firebreaks to prevent the spread of wildfire. It is evident

from this study that the abundance of I. cylindrica decreased when there is a development

in the vegetation structure, therefore shading by plantation forest or agroforestry using a

fast-growing legume tree species may be more effective in suppressing I. cylindrica if this

species becomes problematic in the future (Brook 1989; MacDonald 2009; Murniati 2002;

Soerjani et al. 1983).

5.5. CONCLUSIONS AND FURTHER STUDY

Understanding succession in tropical volcanic ecosystems is important in

developing its conservation and restoration strategies (Quesada et al. 2009; Walker et al.

2007). The primary conclusion of this study is that the ecosystem on Mt. Merapi is largely

resilient to this type of volcanic disturbance, as shown by the significant increase in species

richness and diversity, increase in positive species association and improvement in soil

nutrients in just over a decade since disturbance. However, this study had also raised some

concern regarding the role of invasive alien species in the succession. Hence, intentional

ecological intervention (rehabilitation) in the form of weed management on Mt. Merapi

should be considered.

88

While this study has been based on the chronosequence approach, which assumes

space-for-time substitution, clearly there are limitations caused by non-equivalence of sites

of different ages in terms of other environmental factors. This is a frequent problem with

studies aiming to infer dynamics from spatial patterns, and it is important to consider the

possible impacts of site differences on the patterns observed. Factors such as slope and

landscape context may be important in determining the dynamics of particular sites (e.g. del

Moral and Ellis 2004). It was beyond the scope of this study to examine these other factors

in more detail, but future research should focus on elucidating their effects on the patterns

found here.

There are also a number of additional issues that should be addressed in future

research. Firstly, the role of Eupatorium and Imperata in plant community development is

important to understand, as these species are becoming more widespread in many disturbed

ecosystems and has been observed to change soil properties and cause an arrested state of

succession (Collins and Jose 2009; Lindig-Cisneros et al. 2006; Murniati 2002). Secondly,

observation of interspecific associations in this study showed that positive associations

were more apparent than negative associations - however an experimental study on species

interactions is needed in order to understand the causation behind the associations. Such an

experimental study will benefit from the available data on the observed interspecific

association because it can focus on a relatively few species that have strong associations.

Lastly, even though the results from this study showed that the ecosystem is resilient to the

nuées ardentes disturbances, escalating human activities on Mt. Merapi are likely to be one

of the factors that may alter the successional trajectory. Therefore a long-term ecosystem

dynamics study is needed to examine the successional trajectory, not only under the

influence of volcanic disturbance, but also escalating anthropogenic disturbance. Long-term

89

and experimental studies are also imperative if the process and mechanism of recovery and

establishment in succession is to be investigated.

90

APPENDICES Appendix 1. Mean abundance of species per family in the primary succession study sites. Asterixis indicates exotic species

Deposits Families/species Life form

2006 1998 1997 1994 Amaranthaceae - - - 0.1 Gomphrena celosioides* Herb Apiaceae Centella asiatica Herb - - - 1.5 Asteraceae Adenostema phirsutum Shrub - - 0.1 - Anaphalis javanica Shrub - 3.1 22.9 9.5 Ageratum conyzoides* Herb - 2.3 - - Blumea lacera Herb 1.6 0.1 0 0.4 Crassocephalum crepidioides* Herb 0.3 - 0.1 0.1 Dichrocephala chrysanthemi* Herb - - 0.5 - Emilia sonchifolia* Herb 0 0 0.4 0.3 Erechtites valerianifolia* Herb - 0.1 - - Erigeron sumatrensis 0.1 - - 0.2 Eupatorium odoratum* Shrub - 1 0.3 1.3 Eupatorium riparium* Shrub 0.1 6.1 16.4 5 Galingsoga parviflora Herb - - - 0.1 Gnaphalium japonicum Herb 0.7 - - 0 Laggera alata Herb - - - 0.1 Brassicaceae Nasturtium indicium Herb - - 0.6 - Commelinaceae Aneilema nudiflorum* Herb - - 0.2 - Cyperaceae Cyperus flavidus* Grass - 57.4 - 0.4 Cyperus rotundus* Grass 0.1 - 4.5 0.5 Davalliaceae Humata repens Fern - - 0.1 - Fabaceae Calliandra callothyrsus Tree 0.2 0.3 3.8 0.3 Dalbergia sissoo Tree - - - 0.1 Lespedeza junghuhniana Herb - 0.2 0.1 0.2 Uraria lagopodioides Shrub - - - 0.2 Labiatae Leucas lavandulaefolia Herb - - 1.3 - Lomariopsidaceae Bolbitis sinuata Fern - - - 0.1

91

Deposits

Families/species Life form 2006 1998 1997 1994

Melastomataceae Melastoma affine Shrub 0.7 - - 3.7 Pinaceae Pinus merkusii Tree 3.3 - - - Piperaceae Piper decumanum Climb - 0.2 - - Plantaginaceae Artanema longifolia Herb - 0.4 - - Poaceae Brachiaria mutica* Grass - - - 0.2 Eleusine indica Grass - - - 0.1 Imperata cylindrica Grass - 54.6 10.2 5.5 Panicum reptans Grass - 21.1 - - Paspalum conjugatum Grass 13.9 - - 1.5 Paspalum longifolium Grass 0.1 - - 0.7 Pennisetum purpureum Grass 1.5 - 0.2 - Polyosma ilicifolia Grass - 0.6 6.4 0.3 Polytoca bracteata Grass - - 0.8 0.1 Polytrias amaura Grass - - 17.1 18.6 Setaria sp Grass - - - 2.5 Polygalaceae Polygala paniculata* Herb 2.8 - - 10.5 Polypodiaceae Athyrium macrocarpum Fern 11.6 0.6 3.4 0.5 Rosaceae Rubus flaxinifolius* Shrub - - 0.7 - Rubiaceae Borreria alata Herb - - - 2.9 Pshychotria malayana Shrub - 0.1 - - Theaceae Schima wallichii Tree - - - 0.1 Tiliaceae Grewia sp Shrub 1.6 - - - Urticaceae Debregeasia longifolia Shrub - - 0.1 - Verbenaceae Lantana camara* Shrub - - 0.7 - Stachytarpheta jamaicensis* Herb - 0.1 - -

92

Appendix 2. Mean abundance of species per family in the secondary succession study sites. Asterixis indicates exotic species

Sites Families/species Life

form 2006 1998 1997 1994 Ref. Site Apiaceae Centella asiatica Herb 8.6 - 1.4 7 - Araceae Strobilus asper - - - 0.6 3 Arecaceae Calamus sp Palm - - - - 0.8 Asteraceae Ageratum conyzoides* Herb 5.8 - 0.2 8.8 0.4 Anaphalis javanica Shrub 2.8 0.2 - - - Erigeron sumatrensis - 1.4 - - - Eupatorium odoratum* Shrub - 13.6 11.4 75.6 - Eupatorium riparium* Shrub - 382.2 463 - 51.6 Eupatorium sp* Shrub 7.6 - - - - Balsaminaceae Impatiens sp - 1.4 - - - Cyatheaceae Cyathea contaminan Tree fern - 0.6 0.2 0.2 - Cyperaceae Cyperus rotundus* Grass 1 - 0.8 0.2 0.2 Cyperus sp* Grass - 1 0 0.2 - Euphorbiaceae Homalanthus giganteus Tree - 14.4 - - - Phyllanthus urinaria Shrub - - - 2.6 - Sauropus androgynus Tree 0.2 0.2 0.8 0 0 Fabaceae Acacia decurrens* Tree 20.4 1.2 - 2.2 - Albizia lopantha Tree - 4.8 - - - Calliandra callothyrsus Tree - - 78.6 8 0.4 Erythrina sp Tree - 1 - - - Lespedeza junghuhniana Herb - 1 - - - Leucaena glauca* Tree 0.2 - 1.2 0.2 - Paraserianthes falcataria Tree - - 2.2 1.6 - Parkia sp Tree - 0.2 0.4 - - Mimosa sp Herb 6.4 - - - - Fagaceae Lithocarpus costata Tree - 1 - 2 -

93

Sites Families/species Life

form 2006 1998 1997 1994 Ref. Site Hammamelidaceae Altingia excelsa Tree 1.6 - 0.4 0.6 17.8 Hymenophyllaceae Trichomanes maximum* Fern - - - - 4 Labiatae Leucas lavandulaefolia Herb - - 6.8 1.6 - Pogostemon auricularia Grass - - 1.4 0.4 - Melastomataceae Melastoma malabathricum Shrub - - 7.2 2.6 - Melastoma sp Shrub - 2.6 - - - Meliaceae Dysoxyllum caulostacyhum Tree - - - - 2 Orchidaceae Herminium lanceum* Orchid - 2.8 - - - Papilionaceae Shuteria vestita Shrub - 3.8 - 6.8 - Pinaceae Pinus merkusii Tree 4.4 - - 6.2 3.2 Poaceae Andropogon citratus Grass - 1 - 2 - Brachiaria paspaloides* Grass - 84.4 - - 11.2 Brachiaria reptans* Grass - 3.6 146.8 100 8.6 Dichantium caricosum* Grass - - 44 125 - Digitaria longiflora* Grass - - - 19.4 - Eleusine indica Grass - - 31 68.4 14 Imperata cylindrica Grass 181.2 72 126.6 35.4 - Oryza granulata Grass 0 176.8 37.6 9.8 67.4 Panicum reptans Grass - - 17.6 - - Pennisetum purpureum Grass 6.8 - - - - Polygalaceae Polygala paniculata* Herb 45.8 - 2.2 14 - Polygonaceae Polygonum chinense* Shrub - 5.4 - - - Polygonum paniculata* Shrub - 0.8 - - - Polypodiaceae Athyrium dilatatum Fern - - - - 1.6 Athyrium macrocarpum Fern - - - - 41.6 Nephrolepsis sp Fern - - - 4 9.6 Portulacaceae Talinum sp* Shrub 2.8 - - - -

94

Sites Families/species Life

form 2006 1998 1997 1994 Ref. Site Rubiaceae Coffea sp* Shrub - - 0.4 - 9 Mitracarpus villosus - - 3.8 4.4 - Saprosma arboreum Herb - 3 4.2 21.6 1 Selaginellaceae Selaginella doederleinii Fern - - 1.6 0.2 318 Smilacaeae Smilax sp* Tree - - 2.2 - 3.4 Theaceae Schima wallichii Tree - - 5.2 5.4 5.2 Urticaceae Laportea sinuata Shrub - - - - 1.6 Verbenaceae Lantana camara* Shrub - 0.6 3.6 4.2 -

95

Appendix 3. Sketch of Eupatorium odoratum

96

Appendix 4. Sketch of Eupatorium riparium

97

REFERENCES

Abdurachman E. K., Bourdier J. L. & Voight B. (2000) Nuees ardentes of 22 November

1994 at Merapi Volcano, Java Indonesia. Journal of Volcanology and Geothermal

Research 100, 345-61.

Andrews A. C. (1983) Imperata cylindrica in the highlands of Northern Thailand: Its productivity and status as a weed. Mountain Research and Development 3, 386-8.

Anonym. (1999) Penambangan Pasir di Merapi : Semakin Merusak, Semakin Merugikan. WALHI Yogyakarta, Yogyakarta.

Anonym. (2004) Rencana Pengelolaan Taman Nasional Gunung Merapi Periode 2005-2024. p. 125. Balai Konservasi Sumber Daya Alam Yogyakarta & Pusat Studi Agroekologi Universitas Gadjah Mada, Yogyakarta.

Anonym. (2009) Keadaan Umum Kawasan Taman Nasional Bromo Tengger Semeru. Balai Besar Taman Nasional Bromo Tengger Semeru, Malang.

Antos J. A. & Zobel D. B. (2005) Plant responses in forest of the Tephra-fall zone. In: Ecological responses to the 1980 eruption of Mount St. Helens (eds V. H. Dale, F. J. Swanson and C. M. Crisafulli) p. 47. Springer, New York.

Aplet G. H., Hughes R. F. & Vitousek P. M. (1998) Ecosystem development on Hawaiian lava flows: biomass and species composition. Journal of Vegetation Science 9, 17-26.

Aubert M., Alard D. & Bureau F. (2003) Diversity of plant assemblages in managed temperate forests: a case study in Mormandy (France). Forest Ecol Manage 175, 321–37.

Backer C. A. & van den Brink R. C. B. (1963) Flora of Java. The Rijksherbarium, Leiden.

Baer S. G., Collins S. L., Blair J. M., Knapp A. K. & Fiedler A. K. (2005) Soil heterogeneity effects on tall grass prairie community heterogeneity: an application of ecological theory to restoration ecology. Restoration Ecology 13, 413-24.

Baeza M. J., Valdecantos A., Alloza J. A. & Vallejo V. R. (2007) Human disturbance and environmental factors as drivers of long-term-post-fire regeneration patterns in Mediterranian forests. Journal of vegetation science 18, 243-52.

Baniya C. B., Solhoy T. & Vetaas O. R. (2009) Temporal changes in species diversity and composition in abandoned fields in a trans-Himalayan landscape, Nepal. Plant Ecology 201, 383-99.

Barbour M. G., Burk J. H. & Pitts W. D. (1980) Terrestrial plant ecology. The Benjamin Cummings Publishing Company Inc., California.

Bardintzeff J. M. (1984) Merapi Volcano (Java, Indonesia) and Merapi-type nuees ardentes. Bulletin Volcanology 47.

98

Bautista-Cruz A. & del Castillo R. F. (2005) Soil changes during secondary succession in a tropical montane cloud forest area. Soil Science Society of America Journal 69, 906-14.

Bellingham P. J., Walker L. R. & Wardle D. A. (2001) Differential facilitation by a nitrogen-fixing shrub during primary succession influences relative performance of canopy tree species. Journal of Ecology 89, 861-75.

Belousov A., Voight B. & Belousova M. (2007) Directed blasts and blast-generated pyroclastic density currents: a comparison of the Bezymianny 1956, Mount St Helens 1980, and Soufrière Hills, Montserrat 1997 eruptions and deposits. Bulletin Volcanology 69, 701–40.

Bemmelen W., R. (1970) The geology of Indonesia. Martinus Nijhoff.

Berthommier P. C. & Camus G. (1991) Merapi et ses e´ruptions:importance des me´chanismes phre´atomagmatiques. Bull. Soc.Ge´ol. Fr 5, 635–44.

Bond W. J. & Wilgen B. W. v. (1996) Fire and Plants. Chapman & Hall, London.

Bormann B. T., Spaltenstein H., Mcclellan M. H., Ugolini F. C., Cromack J., K. & Nay S. M. (1995) Rapid soil development after wind throw disturbance in pristine forests Journal of Ecology 83, 747-57.

Brearley F. Q., Prajadinata S., Kidd P. S., Proctor J. & Suriantata. (2004) Structure and floristics of an old secondary rain forest in Central Kalimantan, Indonesia, and a comparison with adjacent primary forest. Forest Ecology and Management 195, 385-97.

Bremmer J. M. & Mulvaney C. S. (1982) Nitrogen total. In: Method of soil analysis. (ed A. L. Page) pp. 595-624. American society of agronomy Inc., Wisconsin.

Brook R. M. (1989) Review of literature on Imperata cylindrica (L) Raeushel. Tropical Pest Management 35, 12-25.

Bush M. B., Whittaker R. J. & Partomohardjo T. (1992) Multiple Krakatau pathways, divergence of types in lowland forests. Geojurnal 28, 99-185.

Callaway R. M. & Walker L. R. (1997) Competition and facilitation: A synthetic approach to interactions in plant communities. Ecology 78, 1958–65.

Camus G., A. G., Mossand-Berthommier P. C. & Vincent P. M. (2000) Merapi (Central Java, Indonesia): An outline of the structural and magmatological evolution, with a special emphasis to the major pyroclastic events. Journal of Volcanology and Geothermal Research 100, 139–63.

Capogna F., Persiani A. M., Maggi O., Dowgiallo G., Puppi G. & Manes F. (2009) Effects of different fire intensities on chemical and biological soil components and related feedbacks on a Mediterranean shrub (Phillyrea angustifolia L.). Plant Ecology 204, 155–71.

99

Chazdon R. L., Letcher S. G., van Breugel M., Martı´nez-Ramos M., Bongers F. & Finegan B. (2007) Rates of change in tree communities of secondary Neotropical forests following major disturbances. Phil. Trans. R. Soc. B 362, 273–89.

Chiu C., Cheung E. & Wok K. (2001) Pelean eruptions and nuees ardentes. Thinkquest.

Clarke K. R. (1993) Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology 18, 117-43.

Clarke K. R. & Ainsworth M. (1993) A method of linking multivariate community structure to environmental variables. Marine Ecology Progress Series 92, 205-19.

Clarke K. R. & Gorley R. N. (2005) PRIMER: Plymouth Routines In Multivariate Ecological Research. PRIMER-E Ltd., Plymouth.

Clearly D., F.R., Priadjati A., Suryokusumo B. K. & Steph B. J. M. (2006) Butterfly, seedling, sapling and tree diversity and composition on a fire-affected Bornean rainforest. Austral Ecology 31, 46-57.

Clements F. E. (1916) Plant Succession. Carnegie Institute Washington Publisher, Washington.

Collins A. R. & Jose S. (2009) Imperata cylindrica, an Exotic Invasive Grass,Changes Soil Chemical Properties of Forest Ecosystems in the Southeastern United States. In: Invasive Plants and Forest Ecosystems (eds R. K. Kohli, S. Jose, H. P. Singh and D. R. Batish) p. 237. CRC Press, London.

Connell J. H. & Slatyer R. O. (1977) Mechanisms of succession in natural communities and their role in community stability and organization. The American Naturalists 111, 1119–44.

Cook W. M., Yao J., Foster B. L., Holt R. D. & Patrick L. B. (2005) Secondary succession in an experimentally fragmented landscape: Community patterns across space and time. Ecology 86, 1267-79.

Crain C. M., Albertson L. K. & Bertness M. D. (2008) Secondary succession dynamics in estuarine marshes across landscape-scale salinity gradients. Ecology 89, 2889–99.

Cramer V. A., Hobbs R. J. & Standish R. J. (2008) What's new about old fields? Land abandonment and ecosystem assembly. Trends in Ecology & Evolution 23, 104-12.

Curtis J. T. & McIntosh R. P. (1950) The interrelations of certain analytic and synthetic phytosociological characters. Ecology 31, 435-55.

Dale V., H., Swanson F., J. & Crisafulli C., M. (2005a) Disturbance, Survival, and Succession: Understanding Ecological Responses to the 1980 Eruption of Mount St. Helens. In: Ecological Responses to the 1980 Eruption of Mount St. Helens (eds V. Dale, H., F. Swanson, J. and C. Crisafulli, M) pp. 3-12. Springer, New York.

Dale V. H., Acevedo J. D. & MacMahon J. (2005b) Effects of Modern Volcanic Eruptions on Vegetation. In: Volcanoes and the Environment (eds J. Marti and G. Ernst) p. 227. Cambridge University Press, New York.

100

Dale V. H., Campbell D. R., Adams W. M., Crisafulli C. M., Dains V. I., Frenzen P. M. & Holland R. F. (2005c) Plant succession on the Mount St. Helens Debris-Avalanche deposit. In: Ecological responses to the 1980 eruption of Mount St. Helens (eds V. H. Dale, F. J. Swanson and C. M. Crisafulli) p. 59. Springer, New York.

Darmawijaya M. I. (1990) Klasifikasi Tanah. Gadjah Mada University Press, Yogyakarta.

Debano L. F. & Conrad D. E. (1978) The effect of fire on nutrients in a Chaparral Ecosystem. Ecology 59, 489-97.

del Moral R. (2000) Succession and local species turnover on Mount St. Helens, Washington. Acta Phytogeogr. Suec. 85, 51-60.

del Moral R. (2007) Limits to convergence of vegetation during early primary succession. Journal of Vegetation Science 18, 479-88.

del Moral R. & Ellis E. E. (2004) Gradients in compositional variation on lahars, Mount St. Helens, Washington, USA. Plant Ecology 175, 273–86.

del Moral R., Saura J. M. & Emenegger J. N. (2010) Primary succession trajectories on a barren plain, Mount St. Helens, Washington. Journal of Vegetation Science 1, 1-11.

del Moral R. & Wood D. M. (1993) Early primary succession on the volcano Mount St. Helens. Journal of Vegetation Science 4, 223-34.

Dinas Kehutanan DIY. (1999) Rencana Umum Pengelolaan Kawasan Lindung Propinsi Daerah Istimewa Yogyakarta. (ed D. K. P. D. I. Yogyakarta).

Durán J., Rodríguez A., Fernández-Palacios J. M. & Gallardo A. (2009) Changes in net N mineralization rates and soil N and P pools in a pine forest wildfire chronosequence. Biol Fertil Soils 45, 781–8.

Dzwonko Z. & Gawrofiski S. (1994) The role of woodland fragments, soil types, and dominant species in secondary succession on the western Carpathian foothills. Vegetatio 111, 149-60.

Eggler F. A. (1954) Vegetation science concept I : Initial floristic composition-a factor in old field vegetation development. Vegetatio 4, 412-7.

Eggler W. A. (1959) Manner of invasion of volcanic deposits by plants, with further evidence from Parricutin and Jorullo Ecological Monographs 29, 267-84.

Endo M., Yamamura Y., Tanaka A., Nakano T. & Yasuda T. (2008) Nurse-plant effects of a dwarf shrub on the establishment of tree seedlings in a volcanic desert on Mt, Fuji, central Japan. Arctic Antarctic and Alpine Research 40, 335-42.

Eussen J. H. H. & Soerjani M. (1975) Problems and control of ‘‘alang-alang’’ [Imperata cylindrica (L.) Beauv.] in Indonesia. In: 5th Annual Conference Asian-Pacific Weed Science Society p. 58.

101

Finnegan B. (1996) Pattern and process in neotropical secondary rain forests: The first 100 years of succession. Trends in Ecology and Evolution 11, 119-24.

Foster B. L. & Tilman D. (2000) Dynamic and static views of succession: Testing the descriptive power of the chronosequence approach. Plant Ecology 146, 1–10.

Franklin J. F., MacMahon J. A., Swanson F. J. & Sedell J. R. (1985) Ecosystem responses to catastrophic disturbances: Lesson from Mount St. Helens. National Geographic Research 1, 198-216.

Fridriksson S. & Magnusson B. (1992) Development of the ecosystem on Surtsey with reference to Anak Krakatau. GeoJournal 28, 287–91.

Gertisser R. & Keller J. (2003) Trace element and Sr, Nd, Pb and O isotope variations in medium-K and high-K volcanic rocks from Merapi Volcano, Central Java, Indonesia: Evidence for the involvement of subducted sediments in Sunda Arc magma genesis. Journal of Petrology 44, 457.

Gomez-Pompa A. & Vazquez-Yanes C. (1981) Successional Studies of a Rain Forest in Mexico. In: Forest Succession: Concepts and Application (eds D. C. West, H. H. Shugart and D. B. Botkin) pp. 246-66. Springer-Verlag, New York.

Gomez C., Lavigne F., Hadmoko D. S., Lespinasse N. & Wassmer P. (2009) Block-and-ash flow deposition: A conceptual model from a GPR survey on pyroclastic-flow deposits at Merapi Volcano, Indonesia. Geomorphology 110, 118–27.

Gonzalez-Tagle M. A., Schwendenmann L., Perez J. J. & Schulz R. (2008) Forest structure and woody plant species composition along a fire chronosequence in mixed pine-oak forest in the Sierra Madre Oriental, Northeast Mexico. Forest Ecology and Management 256, 161-7.

Grashoff J. L. & Beaman J. H. (1970) Studies in Eupatorium (Compositae), III. Apparent Wind Pollination. Brittonia 22, 77-84.

Hardiwinoto S., Pudyatmoko S. & Sabarnurdin S. (1998) Tingkat ketahanan dan proses regenerasi vegetasi setelah letusan Gunung Merapi. Manusia dan Lingkungan 5, 47-59.

Herben T. (1996) Permanent plots as tools for plant community ecology. Journal of Vegetation Science 7, 195-202.

Herrera B. & Finegan B. (1997) Substrate conditions, foliar nutrients and the distributions of two canopy tree species in a Costa Rican secondary rain forest. Plant and Soil 191, 259-67.

Heru C. N. (2002) 300 Hektare Hutan Merapi Dilalap Api. Koran tempo, Yogyakarta.

Heyne K. (1987) Tumbuhan Berguna Indonesia. Yayasan Sarana Wana Jaya, Jakarta.

Hobbs R., J. & Huenneke L. F. (1992) Disturbance, diversity and invasion: Implication for conservation. Conservation Biology 6, 324-36.

102

Hobbs R., J., Jentsch A. & Temperton M., Vicky. (2007) Restoration as a process of assembly and succession mediated by disturbance. In: Linking Restoration and Ecological Succession (eds R. L. Walker, J. Walker and R. Hobbs, J.) pp. 150-67. Springer, New York.

Hobbs R. J. & Norton D. A. (1996) Toward a conceptual framework for restoration ecology. Restoration Ecology 4, 93–110.

Hodkinson I. D., Webb N. R. & Coulson S. J. (2002) Primary community assembly on land - the missing stages: why are the heterotrophic organisms always there first? Journal of Ecology 90, 569-77.

Horn S. P., Kennedy L. M. & Orvis K. H. (2001) Vegetation recovery following a high elevation fire in the Dominican Republic. Biotropica 33, 701-8.

Hubbell S. P., Foster R. B., O'Brien S. T., Harms K. E., Condit R., Wechsler B., Wright S. J. & Loo de Lao S. (1999) Light-Gap Disturbances, Recruitment Limitation, and Tree Diversity in a Neotropical Forest. In: Science, New Series pp. 554-7.

Hughes R. F. & Denslow J. S. (2005) Invasion by a N-2-fixing tree alters function and structure in wet lowland forests of Hawaii. Ecological Applications 15, 1615-28.

Inouye R., S. & Tilman D. (1988) Convergence and divergence of old-field plant communities along experimental nitrogen gradients. Ecology 69, 995-1004.

IPNI. (2008) The International Plant Names Index Databases. The Royal Botanic Garden, Harvard University Herbaria and Australian National Herbarium.

Irawan D. E. & Puradimaja D. J. (2006) The hydrogeology of the volcanic spring belt, east slope of Gunung Ciremai, West Java, Indonesia. The Geological Society of London.

Isango J. A. (2007) Stand Structure and Tree Species Composition of Tanzania Miombo Woodlands: A Case Study from Miombo Woodlands of Community Based Forest Management in Iringa District. In: Management of Indigenous Tree Species for Ecosystem Restoration and Wood Production in Semi-Arid Miombo Woodlands in Eastern Africa pp. 43-56. MITMIOMBO, Tanzania.

Johnson E. A. & Miyanishi K. (2008) Testing the assumptions of chronosequence in succession. Ecology Letters 11, 419–31.

Jonathan J. & Hariadi B. P. J. (1999) Imperata cylindrica (L.) RaeuschelIn. In: Plant Resources of South-East Asia No. 12(1): Medicinal and poisonous plants 1. (eds L. S. de Padua, N. Bunyapraphatsara and R. H. M. J. Lemmens) p. 310. Backhuys Publisher, Leiden, The Netherlands.

Kelfoun K., Legros F. & Gourgaud A. (2000) A statistical study of trees damaged by the 22 November 1994 eruption of Merapi volcano (Java, Indonesia): Relationships between ash-cloud surges and block-and-ash flows. Journal of Volcanology and Geothermal Research 100, 379–93.

103

Kennard D. K. & Gholz H. L. (2001) Effects of high and low intensity fires on soil properties and plant growth in a Bolivian dry forest. Plant and Soil 234, 119–29.

Kent M. & Coker P. (1992) Vegetation Description and Analysis, A practical Approach. John Wiley & Sons, New York.

Kunwar R. M. (2003) Invasive alien plants and Eupatorium: Biodiversity and livelihood. Him J Sci 1, 129-33.

Kushwaha S. P. S., Ramakrishnan P. S. & Tripathi R. S. (1981) Population dynamics of Eupatorium odoratum in successional environments following slash and burn agriculture. Journal of Applied Ecology 18, 529-35.

Kusmana C. (1995) Teknik Pengukuran Keanekaragaman Tumbuhan. In: Pelatihan Teknik Pengukuran dan Monitoring Biodiversity di Hutan Tropika Indonesia. Jurusan Koservasi Sumber Daya Hutan Fakultas Kehutanan Institut Pertanian Bogor Bogor.

Lambers H., Raven J. A., Shaver G. R. & Smith S. E. (2007) Plant nutrient-acquisition strategies change with soil age. Trends in Ecology and Evolution 23, 95-103.

Lavigne F. (1999) Lahar hazard micro-zonation and risk assessment in Yogyakarta City, Indonesia. GeoJournal 49, 173-83.

Lavigne F. & Gunnell Y. (2006) Land cover change and abrupt environmental impacts on Javan volcanoes, Indonesia: a long-term perspective on recent events. Regional Environmental Change 6, 86-100.

Le Brocque A. F. (1995a) Ecology of Plant Communities in Ku-ring-gai Chase National Park, New South Wales: An Examination of Vegetation and Environmental Patterns. In: Dept of Applied Biology p. 242. University of Technology Sydney.

Le Brocque A. F. (1995b) Vegetation and environmental patterns on soils derived from Hawkesbury Sandstone Narrabeen substrata in Ku-ring-gai Chase National Park, New South Wales. Australian Journal of Ecology 20.

Lepš J. (1990) Can underlying mechanisms be deduced from observed patterns. In: Spatial processes in plant communities (eds F. Krahulec., A. D. Q. Agnew, S. Agnew and J. H. Willems) pp. 1-11. SPB Academic Publisher, The Hague.

Lepš J. & Rejmanek M. (1991) Convergence or divergence: What should we expect from vegetation succession? Oikos 62, 261-4.

Li X., Wilson S. D. & Song Y. (1999) Secondary succession in two subtropical forests. Plant Ecology 143, 13-21.

Lindig-Cisneros R. (2009) Alternative Stable States for Planning and Implementing Restoration of Production Systems in Michoacan, Mexico. In: New Models for Ecosystem Dynamics and Restoration (eds R. J. Hobbs and K. Suding) pp. 311-22. Island Press, Washington.

104

Lindig-Cisneros R., Galindo-Vallejo S. & Lara-Cabrera S. (2006) Vegetation of tephra deposits 50 years after the end of the eruption of the Paricutin Volcano, Mexico. Southwestern Naturalist 51, 455-61.

Ludwig J. A. & Reynolds J. H. (1988) Statistical ecology: A primer on methods and computing. John Wiley & Sons, Singapore.

Luken J. O. (1990) Directing ecological succession. Chapman and Hall, London.

MacDonald G. E. (2009) Cogongrass (Imperata cylindrica) — A Comprehensive Review of an Invasive Grass. In: Invasive Plants and Forest Ecosystems (eds R. K. Kohli, S. Jose, H. P. Singh and D. R. Batish) pp. 267-94. CRC Press, London.

MacDonald G. E., Brecke B. J., Colvin D. L. & Shilling D. G. (1994) Chemical and mechanical control of Dogfennel (Eupatorium capillifolium). Weed Technology 8, 483-7.

Magurran A. E. (1988) Ecological diversity and its measurement. Princeton University Press, Princeton New Jersey.

Mahecha M. D., Martinez A., Lange H., Reichstein M. & Beck E. (2009) Identification of characteristic plant co-occurrences in neotropical secondary montane forests. Journal of Plant Ecology-Uk 2, 31-41.

Marrinan M. J., Edwards W. & Landsberg J. (2005) Resprouting of saplings following a tropical rainforest fire in north-east Queensland, Australia. Austral Ecology 30, 817–26.

Marti J. & Ernst G. (2005) Volcanoes and environment. Cambridge University Press, New York.

McClanahan T. R. (1986) The effect of a seed source on primary succession in a forest ecosystem. Vegetatio 65, 175-8.

McLean R. C. (1919) Studies in the ecology of tropical rain forest: With special reference to the forest of South Brazil. The journal of ecology 7, 5-54.

Meyer R. E. & Bovey R. W. (1991) Response of Yankeeweed (Eupatorium compositifolium) and Associated Pasture Plants to Herbicides. Weed Technology 5, 214-7.

Milberg P. (1995) Soil seed bank after eighteen years of succession from grassland to forest. Oikos 72, 3-13.

Mitchell P. J., Veneklaas E. J., Lambers H. & Burgess S. S. O. (2008) Using multiple trait associations to define hydraulic functional types in plant communities of south-western Australia. Oecologia 158, 385–97.

Montagnini F. & Jordan C. F. (2005) Tropical forest ecology: The basis for conservation and management. Springer, Berlin.

Morris W. F. & Wood D. M. (1989) The role of Lupine in succession on Mount St. Helens: Facilitation or inhibition? Ecology 70, 697-703.

105

Muller F. (2005) Ecosystem Indicators for the Integrated Management of Landscape Health and Integrity. In: Ecological indicators for assessment of ecosystem health (eds S. E. Jørgensen, R. Costanza and F.-L. Xu) pp. 277-303. CRC Press, London.

Murniati. (2002) From Imperata cylindrica Grasslands to Productive Agroforestry. In: Tropenbos International p. 194. Wageningen University, Wageningen, Netherland.

MVO. (2006) Prekursor Erupsi Gunung Merapi. Geological Department, Indonesian Ministry of Energy and Mineral Resources, Yogyakarta.

Myster R. W. & Malahy M. P. (2008) Is there a middle way between permanent plots and chronosequences? Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 38, 3133-8.

Myster R. W. & Pickett S. T. A. (1992a) A comparison of rate of succession over 18 yr in 10 contrasting old fields. Ecology 75, 387-92.

Myster R. W. & Pickett S. T. A. (1992b) Dynamics of associations between plants in ten old fields during 31 years of succession. J Ecol 80, 291–303.

Newhall C. G., Bronto S., Alloway B., Banks N. G., Bahar I., del Marmol M. A., Hadisantono R. D., Holcomb R. T., McGeehin J., Miksic J. N., Rubin M., Sayudi S. D., Sukhyar R., Andreastuti S., Tilling R. I., Torley R., Trimble D. & Wirakusumah A. D. (2000) 10,000 Years of explosive eruptions of Merapi Volcano,Central Java: archaeological and modern implications. Journal of Volcanology and Geothermal Research 100, 9-50.

Norgrove L., Hauser S. & Weise S. F. (2000) Response of Chromolaena odorata to timber tree densities in an agrisilvicultural system in Cameroon: aboveground biomass, residue decomposition and nutrient release. Agriculture, Ecosystems and Environment 81, 191–207.

Odum E. P. (1969) The strategy of ecosystem development. Science 104, 262–70.

Palmer B., Macqueen D. J. & Gutteridge R. C. (1994) Calliandra callothyrsus - a Multipurpose Tree Legume for Humid Locations. In: Forage Tree Legumes in Tropical Agriculture (eds R. C. Gutteridge and M. H. Shelton). Tropical Grassland Society Of Australia Inc. , Queensland

Pan D. Y., Bouchard A., Legendre P. & Domon G. (1998) Influence of edaphic factors on the spatial structure of inland halophytic communities: a case study in China. Journal of Vegetation Science 9, 797-804.

Pannekoek A. J. (1949) Outline of the Geomorphology of Java. Tijds. K. Ned. Aard. Gen 66, 270-326.

Peet R. K. (1992) Community structure and ecosystem function. In: Plant succession: Theory and Prediction (eds D. C. Glenn-Lewin, R. K. Peet and T. T. Veblen) pp. 103-51. Chapman & Hall, London.

106

Pena C.-M. (2003) Changes in forest structure and species composition during secondary forest succession in the Bolivian Amazon. Biotropica 35, 450-61.

Pickett S. T. A. (1989) Space-for-time substitution as an alternative to long-term studies. In: Long-Term Studies in Ecology: Approaches and Alternatives (ed G. E. Likens) pp. 110-35. Springer-Verlag, New York.

Powers J. S., Becknell J. M., Irving J. & Perez-Aviles D. (2009) Diversity and structure of regenerating tropical dry forests in Costa Rica: Geographic patterns and environmental drivers. Forest Ecology and Management 258, 959-70.

Prawiradiputra B. (2007) Kirinyuh (Chromolaena odorata (L) R.M. King dan H. Robinson): Gulma Padang Rumput yang Merugikan. Wartazoa 17, 46-52.

PVMBG. (2006) Prekursor Gunung Merapi. Badan Geologi Direktorat Energi dan Sumber Daya Mineral, Yogyakarta.

Quesada M., Sanchez-Azofeifa G. A., Alvarez-Anorve M., Stoner K. E., Avila-Cabadilla L., Calvo-Alvarado J., Castillo A., Espirito-Santo M. M., Fagundes M., Fernandes G. W., Gamon J., Lopezaraiza-Mikel M., Lawrence D., Morellato L. P. C., Powers J. S., Neves F. D., Rosas-Guerrero V., Sayago R. & Sanchez-Montoya G. (2009) Succession and management of tropical dry forests in the Americas: Review and new perspectives. Forest Ecology and Management 258, 1014-24.

Radosevich S. R., Holt J. S. & Ghersa C. M. (2007) Ecology of Weeds and Invasive Plants: Relationship to Agriculture and Natural Resource Management. John Wiley & Sons, Inc., New Jersey.

Raghubanshi A. S. & Tripathi A. (2009) Effect of disturbance, habitat fragmentation and alien invasive plants on floral diversity in dry tropical forests of Vindhyan highland: a review. Tropical Ecology 50, 57-69.

Ramaharitra T. (2006) The Effects of Anthropogenic Disturbances on the Structure and Composition of Rain Forest Vegetation. Tropical Resources Bulletin 25, 32-7.

Reilly M. J., Wimberly M. C. & Newell C. L. (2006) Wildfire effects on beta-diversity and species turnover in a forested landscape. Journal of Vegetation Science 17, 447-54.

Rogers H. M. & Hartemink A. E. (2000) Soil seed bank and growth rates of an invasive species, Piper aduncum, in the lowlands of Papua New Guinea. Journal of Tropical Ecology 16, 243-51.

Ross K. A., Fox B. J. & Fox M. D. (2002) Changes to plant species richness in forest fragments: fragment age, disturbance and fire history may be as important as area. Journal of Biogeography 29, 749-65.

Ruprecht E., Bartha S., Botta-Dukát Z. & Szabó A. (2007) Assembly rules during old-field succession in two contrasting environments. Community Ecology 8, 31-40.

107

Scheffer M., Carpenter S., Foley J., A., Folke C. & Walker B. (2001) Catastrophic shifts in ecosystems. Nature 413, 591-6.

Schmidt F. H. & Fergusson J. H. (1951) Rainfall Type Base on Wet and Dry Period Ratios. Verhandeling, 42.

Scott W. E., Hoblitt R. P., Torres R. C., Self S., Martinez M. M. L. & Nillos T. J. (1996) Pyroclastic Flows of the June 15, 1991, Climactic Eruption of Mount Pinatubo. In: Fire and Mud: Eruptions and Lahars of Mount Pinatubo, Philippines (eds C. G. Newhall and R. S. Punongbayan). Philippine Institute of Volcanology and Seismology Quezon City and University of Washington Press, Seattle.

Setiawan N. N. (2008) Struktur dan Komposisi Vegetasi Pada Lahan Bekas Ladang yang di reforestasi pada Cagar Alam Gunung Papandayan, Jawa Barat. In: Biology. Sekolah Ilmu Teknologi Hayati, Bandung.

Simbolon H., Siregar M., Wakiyama S., Sukigara N., Abe Y. & Shimizu H. (2003) Impacts of dry season and forest fire 1997-1998 episodes on mixed Dipterocarp Forest at Bukit Bangkirai, East Kalimantan. Berita Biologi 6, 737-47.

Simkin T. & Siebert L. (1994) Volcanoes of the world. Smithsonian Institute, Wahington D.C.

Simon H. (1996) Metode Inventore Hutan. Aditya Media, Yogyakarta.

Simon H. (1998) Pengantar Ilmu Kehutanan. Bagian Penerbitan Yayasan Pembina Fakultas Kehutanan UGM, Yogyakarta.

Smith W., G. (1914) Notes on Danish vegetation. The journal of ecology 2, 65-70.

Soerjani M., Eussen J. H. H. & Tjitrosudirdjo S. (1983) Imperata Research and Management in Indonesia. Mountain Research and Development 3, 397-404.

Spencer D. R., Perry J. E. & Silberhorn G. M. (2001) Early secondary succession in Bottomland Hardwood Forests of Southeastern Virginia. Environmental Management 27, 559–70.

Spencer R.-J. & Gregory S. B. (2006) Effects of fire on the structure and composition of open Eucalypt Forest. Austral Ecology 31, 638-46.

Spurr S. H. & Barnes B. V. (1980) Forest Ecology. John Wiley and Sons, New York.

Standish R. J., Cramer V. A. & Yates C. J. (2009) A Revised State-and-Transition Model for the Restoration of Woodlands in Western Australia. In: New Models for Ecosystem Dynamics and Restoration (eds R. J. Hobbs and K. Suding) pp. 169-88. Island Press, Washington.

Suarez D. L. (1996) Magnesium and calcium. In: Methods of Soil Analysis (ed J. M. Bigham) pp. 575–602. Soil Science Society of America, Madison.

108

Suding K., N. & Hobbs R. J. (2009) Models of ecosystem dynamics as frameworks for restoration ecology. In: New models for ecosystem dynamics and restoration (eds R. J. Hobbs and K. Suding, N.) pp. 3-17. Island press, Washington.

Supriyadi & Marsono D. (2001) Petunjuk praktikum ekologi hutan. Laboratorium Ekologi Hutan Jurusan Konservasi Sumber Daya Hutan Fakultas Kehutanan UGM, Yogyakarta.

Sutomo. (2004) Biomasa dan struktur komunitas tumbuhan bawah di hutan lindung Kaliurang Yogyakarta: Studi di petak 7 RPH Kaliurang In: Forestry Faculty p. 66. Universitas Gadjah Mada, Yogyakarta.

Swamy P. S., Sundarapandian S. M., Chandrasekar P. & Chandrasekaran S. (2000) Plant species diversity and tree population structure of a humid tropical forest in Tamil Nadu, India. Biodiversity and Conservation 9, 1643-69.

Takahashi T. & Tsujimoto H. (2000) A mechanical model for Merapi-type pyroclastic flow. Journal of Volcanology and Geothermal Research 98, 91–115.

Tan K. H. (2008) Soils in the Humid Tropics and Monsoon Region of Indonesia. CRC Press, New York.

Taylor B. W. (1957) Plant succession on recent volcanoes in Papua. The Journal of Ecology, 45, 233-43.

Temperton V. M., Hobbs R. J., Nuttle T. & Halle S. (2004) Assembly Rules and Restoration Ecology. Island Press, Washington.

Thornton I. (2007) Island Colonization The Origin and Development of Island Communities. Ecological Reviews, Cambridge University Press, Cambridge.

Thouret J. C. & Lavigne F. (2005) Hazards and Risks at Gunung Merapi, Central Java: A Case Study. In: The Physical Geography of Southeast Asia (ed A. Gupta) pp. 300-24. Oxford University Press, Oxford.

Titus J. H. & del Moral R. (1998) Seedling establishment in different microsites on Mount St. Helens, Washington, USA. Plant Ecology 134, 13-26.

Tsuyuzaki S. (1991) Species turnover and diversity during early stages of vegetation recovery on the volcano Usu, northern Japan. Journal of Vegetation Science 2, 301-6.

Tsuyuzaki S. & Hase A. (2005) Plant community dynamics on the Volcano Mount Koma, northern Japan, after the 1996 eruption. Folia Geobotanica 40, 319-30.

Uhl C. (1990) Deforestation, Fire Susceptibility and Potential Tree Responses to Fire in the Eastern Amazon. Ecology 71, 437-49.

Valessini F. (2009) NBIO528 Multivariate Techniques and Community Ecology: Course Handout. Centre for Fish and Fisheries Research Murdoch University, Perth.

van der Pijl L. (1939) The re-establishment of vegetation on Mt. Goentoer (Java). Ann. Jard. Bot. Buitenz. 48, 129-52.

109

van der Putten W. H., Mortimer S. R., Hedlund K., van Dijk C., Brown V. K., Lepš J., Rodriguez-Barrueco C., Roy J., Diaz Len T. A., Gormsen D., Korthals G. W., Lavorel S., Santa Regina I. & Smilauer P. (2000) Plant species diversity as a driver of early succession in abandoned fields: a multi-site approach. Oecologia 124, 91–9.

van Steenis C. G. G. J. (1972) The Mountain Flora of Java. E.J Brill, Leiden.

Velazquez E. & Gomez-Sal A. (2007) Environmental control of early succession on a large landslide in a tropical dry ecosystem (Casita Volcano, Nicaragua). Biotropica 35, 601–9.

Vitousek P. M., Walker L. R., Whiteaker L. D. & Matson P. A. (1993) Nutrient limitations to plant-growth during primary succession in Hawaii-Volcanos-National-Park. Biogeochemistry 23, 197-215.

Voight B., Sukhyar R. & Wirakusumah A. D. (2000) Introduction to the special issue on Merapi Volcano. Journal of Volcanology and Geothermal Research 100, 1-8.

Walker L. R., Clarkson B. D., Silvester W. B. & Clarkson B. R. (2003) Colonization dynamics and facilitative impacts of a nitrogen-fixing shrub in primary succession. Journal of Vegetation Science 14, 277-90.

Walker L. R. & del Moral R. (2009) Transition Dynamics in Succession: Implications for Rates, Trajectories and Restoration. In: New Models for Ecosystem Dynamics and Restoration. (eds K. Suding and R. J. Hobbs) pp. 33-49. Island Press., Washington.

Walker L. R., Walker J. & del Moral R. (2007) Forging a New Alliance Between Succession and Restoration. In: Linking Restoration and Ecological Succession (eds L. R. Walker, J. Walker and R. J. Hobbs) pp. 1-18. Springer, New York,.

Walker R. L. & del Moral R. (2003) Primary succession and ecosystem rehabilitation. Cambridge University Press

Walker T. W. & Syers J. K. (1976) The fate of phosphorus during pedogenesis. Geoderma 15, 1–19.

Walkey A. & Black A. I. (1934) An examination of the Degtjoreff method for determinating soil organic matter and a proposed modification of the chromic acid titration method. Soil Sci., 37-9.

Wang C. T., Long R. J., Wang Q. L., Jing Z. C. & Shi J. J. (2009) Changes in plant diversity, biomass and soil C, in Alpine Meadows at different degradation stages in the Headwater Region of Three Rivers, China. Land Degradation and Development 20, 187–98.

Watson P. & Wardel-Johnson G. (2004) Fire frequency and time since fire effects on the open-forest and woodland flora of Girraween National Park, south-east Queensland, Australia. Austral Ecology 29, 225-36.

Weill A. (2004) Volcanoes. Saddleback educational publishing, California.

110

Whittaker R. H. (1960) Vegetation of the Siskiyou Mountains, Oregon and California. Ecol. Monogr. 22, 1-44.

Whittaker R. J., Partomihardjo T. & Jones S., H. (1999) Interesting times on Krakatau: Stand dynamics in the 1990s. Philosophical transactions: Biological Sciences 354, 1857-67.

Whitten T., Soeriaatmadja R. E. & Afiff S. A. (1996) The ecology of Indonesia series volume II: The ecology of Java and Bali. Periplus, Hongkong.

Widyatmoko D. & Burgman M. A. (2006) Influences of edaphic factors on the distribution and abundance of a rare palm (Cyrtostachys renda) in a peat swamp forest in eastern Sumatra, Indonesia. Austral Ecology 31, 964–74.

Wills T. J. (2002) Succession in sand heathland at Loch Sport, Victoria: changes in vegetation, soil seed banks and species traits. In: School of Biological Sciences p. 240. Monash University, Melbourne.

Wuragil. (2009) Invasi Spesies Asing, Ancaman Satwa dan Flora Lokal Tempo Interaktif.

Yadav A. S. & Tripathi A. (1985) Effect of soil moisture and sowing density on population growth of Eupatorium adenophorum and E. riparium. Plant and Soil 88, 441-7.

Yadav A. S. & Tripathi R. S. (1981) Population dynamics of the ruderal weed Eupatorium Odoratum and its natural regulation. Oikos 36, 355-61.

Zahawi R. A. & Augspurger C. K. (1999) Early plant succession in abandoned pastures in Ecuador. Biotropica 31, 540-52.

Zhu W. Z., Cheng S., Cai X. H., He F. & Wang J. X. (2009) Changes in plant species diversity along a chronosequence of vegetation restoration in the humid evergreen broad-leaved forest in the Rainy Zone of West China. Ecological Research 24, 315-25.

Zimmerman N., Hughes R. F., Cordell S., Hart P., Chang H. K., Perez D., Like R. K. & Ostertag R. (2008) Patterns of primary succession of native and introduced plants in lowland wet forests in eastern Hawai‘i. Biotropica 40, 277–84.

Zuo X., Zhao X., Zhao H., Zhang T., Guo Y., Li Y. & Huang Y. (2009) Spatial heterogeneity of soil properties and vegetation–soil relationships following vegetation restoration of mobile dunes in Horqin Sandy Land, Northern China. Plant Soil 318, 153–67.